1GCC(1)                                GNU                               GCC(1)
2
3
4

NAME

6       gcc - GNU project C and C++ compiler
7

SYNOPSIS

9       gcc [-c|-S|-E] [-std=standard]
10           [-g] [-pg] [-Olevel]
11           [-Wwarn...] [-Wpedantic]
12           [-Idir...] [-Ldir...]
13           [-Dmacro[=defn]...] [-Umacro]
14           [-foption...] [-mmachine-option...]
15           [-o outfile] [@file] infile...
16
17       Only the most useful options are listed here; see below for the
18       remainder.  g++ accepts mostly the same options as gcc.
19

DESCRIPTION

21       When you invoke GCC, it normally does preprocessing, compilation,
22       assembly and linking.  The "overall options" allow you to stop this
23       process at an intermediate stage.  For example, the -c option says not
24       to run the linker.  Then the output consists of object files output by
25       the assembler.
26
27       Other options are passed on to one or more stages of processing.  Some
28       options control the preprocessor and others the compiler itself.  Yet
29       other options control the assembler and linker; most of these are not
30       documented here, since you rarely need to use any of them.
31
32       Most of the command-line options that you can use with GCC are useful
33       for C programs; when an option is only useful with another language
34       (usually C++), the explanation says so explicitly.  If the description
35       for a particular option does not mention a source language, you can use
36       that option with all supported languages.
37
38       The usual way to run GCC is to run the executable called gcc, or
39       machine-gcc when cross-compiling, or machine-gcc-version to run a
40       specific version of GCC.  When you compile C++ programs, you should
41       invoke GCC as g++ instead.
42
43       The gcc program accepts options and file names as operands.  Many
44       options have multi-letter names; therefore multiple single-letter
45       options may not be grouped: -dv is very different from -d -v.
46
47       You can mix options and other arguments.  For the most part, the order
48       you use doesn't matter.  Order does matter when you use several options
49       of the same kind; for example, if you specify -L more than once, the
50       directories are searched in the order specified.  Also, the placement
51       of the -l option is significant.
52
53       Many options have long names starting with -f or with -W---for example,
54       -fmove-loop-invariants, -Wformat and so on.  Most of these have both
55       positive and negative forms; the negative form of -ffoo is -fno-foo.
56       This manual documents only one of these two forms, whichever one is not
57       the default.
58
59       Some options take one or more arguments typically separated either by a
60       space or by the equals sign (=) from the option name.  Unless
61       documented otherwise, an argument can be either numeric or a string.
62       Numeric arguments must typically be small unsigned decimal or
63       hexadecimal integers.  Hexadecimal arguments must begin with the 0x
64       prefix.  Arguments to options that specify a size threshold of some
65       sort may be arbitrarily large decimal or hexadecimal integers followed
66       by a byte size suffix designating a multiple of bytes such as "kB" and
67       "KiB" for kilobyte and kibibyte, respectively, "MB" and "MiB" for
68       megabyte and mebibyte, "GB" and "GiB" for gigabyte and gigibyte, and so
69       on.  Such arguments are designated by byte-size in the following text.
70       Refer to the NIST, IEC, and other relevant national and international
71       standards for the full listing and explanation of the binary and
72       decimal byte size prefixes.
73

OPTIONS

75   Option Summary
76       Here is a summary of all the options, grouped by type.  Explanations
77       are in the following sections.
78
79       Overall Options
80           -c  -S  -E  -o file  -x language -v  -###  --help[=class[,...]]
81           --target-help  --version -pass-exit-codes  -pipe  -specs=file
82           -wrapper @file  -ffile-prefix-map=old=new -fplugin=file
83           -fplugin-arg-name=arg -fdump-ada-spec[-slim]
84           -fada-spec-parent=unit  -fdump-go-spec=file
85
86       C Language Options
87           -ansi  -std=standard  -fgnu89-inline
88           -fpermitted-flt-eval-methods=standard -aux-info filename
89           -fallow-parameterless-variadic-functions -fno-asm  -fno-builtin
90           -fno-builtin-function  -fgimple -fhosted  -ffreestanding -fopenacc
91           -fopenacc-dim=geom -fopenmp  -fopenmp-simd -fms-extensions
92           -fplan9-extensions  -fsso-struct=endianness
93           -fallow-single-precision  -fcond-mismatch  -flax-vector-conversions
94           -fsigned-bitfields  -fsigned-char -funsigned-bitfields
95           -funsigned-char
96
97       C++ Language Options
98           -fabi-version=n  -fno-access-control -faligned-new=n
99           -fargs-in-order=n  -fchar8_t  -fcheck-new -fconstexpr-depth=n
100           -fconstexpr-cache-depth=n -fconstexpr-loop-limit=n
101           -fconstexpr-ops-limit=n -fno-elide-constructors
102           -fno-enforce-eh-specs -fno-gnu-keywords -fno-implicit-templates
103           -fno-implicit-inline-templates -fno-implement-inlines
104           -fms-extensions -fnew-inheriting-ctors -fnew-ttp-matching
105           -fno-nonansi-builtins  -fnothrow-opt  -fno-operator-names
106           -fno-optional-diags  -fpermissive -fno-pretty-templates -fno-rtti
107           -fsized-deallocation -ftemplate-backtrace-limit=n
108           -ftemplate-depth=n -fno-threadsafe-statics  -fuse-cxa-atexit
109           -fno-weak  -nostdinc++ -fvisibility-inlines-hidden
110           -fvisibility-ms-compat -fext-numeric-literals -Wabi-tag
111           -Wcatch-value  -Wcatch-value=n -Wno-class-conversion
112           -Wclass-memaccess -Wcomma-subscript  -Wconditionally-supported
113           -Wno-conversion-null  -Wctor-dtor-privacy  -Wno-delete-incomplete
114           -Wdelete-non-virtual-dtor  -Wdeprecated-copy
115           -Wdeprecated-copy-dtor -Weffc++  -Wextra-semi
116           -Wno-inaccessible-base -Wno-inherited-variadic-ctor
117           -Wno-init-list-lifetime -Wno-invalid-offsetof  -Wno-literal-suffix
118           -Wmismatched-tags -Wmultiple-inheritance  -Wnamespaces  -Wnarrowing
119           -Wnoexcept  -Wnoexcept-type  -Wnon-virtual-dtor -Wpessimizing-move
120           -Wno-placement-new  -Wplacement-new=n -Wredundant-move
121           -Wredundant-tags -Wreorder  -Wregister -Wstrict-null-sentinel
122           -Wno-subobject-linkage  -Wtemplates -Wno-non-template-friend
123           -Wold-style-cast -Woverloaded-virtual  -Wno-pmf-conversions
124           -Wsign-promo -Wsized-deallocation  -Wsuggest-final-methods
125           -Wsuggest-final-types  -Wsuggest-override -Wno-terminate
126           -Wuseless-cast  -Wvirtual-inheritance -Wno-virtual-move-assign
127           -Wvolatile  -Wzero-as-null-pointer-constant
128
129       Objective-C and Objective-C++ Language Options
130           -fconstant-string-class=class-name -fgnu-runtime  -fnext-runtime
131           -fno-nil-receivers -fobjc-abi-version=n -fobjc-call-cxx-cdtors
132           -fobjc-direct-dispatch -fobjc-exceptions -fobjc-gc -fobjc-nilcheck
133           -fobjc-std=objc1 -fno-local-ivars
134           -fivar-visibility=[public|protected|private|package]
135           -freplace-objc-classes -fzero-link -gen-decls -Wassign-intercept
136           -Wno-property-assign-default -Wno-protocol  -Wselector
137           -Wstrict-selector-match -Wundeclared-selector
138
139       Diagnostic Message Formatting Options
140           -fmessage-length=n -fdiagnostics-show-location=[once|every-line]
141           -fdiagnostics-color=[auto|never|always]
142           -fdiagnostics-urls=[auto|never|always]
143           -fdiagnostics-format=[text|json] -fno-diagnostics-show-option
144           -fno-diagnostics-show-caret -fno-diagnostics-show-labels
145           -fno-diagnostics-show-line-numbers -fno-diagnostics-show-cwe
146           -fdiagnostics-minimum-margin-width=width
147           -fdiagnostics-parseable-fixits  -fdiagnostics-generate-patch
148           -fdiagnostics-show-template-tree  -fno-elide-type
149           -fdiagnostics-path-format=[none|separate-events|inline-events]
150           -fdiagnostics-show-path-depths -fno-show-column
151
152       Warning Options
153           -fsyntax-only  -fmax-errors=n  -Wpedantic -pedantic-errors -w
154           -Wextra  -Wall  -Wabi=n -Waddress  -Wno-address-of-packed-member
155           -Waggregate-return -Walloc-size-larger-than=byte-size  -Walloc-zero
156           -Walloca  -Walloca-larger-than=byte-size
157           -Wno-aggressive-loop-optimizations -Warith-conversion
158           -Warray-bounds  -Warray-bounds=n -Wno-attributes
159           -Wattribute-alias=n -Wno-attribute-alias -Wno-attribute-warning
160           -Wbool-compare  -Wbool-operation -Wno-builtin-declaration-mismatch
161           -Wno-builtin-macro-redefined  -Wc90-c99-compat  -Wc99-c11-compat
162           -Wc11-c2x-compat -Wc++-compat  -Wc++11-compat  -Wc++14-compat
163           -Wc++17-compat -Wc++20-compat -Wcast-align  -Wcast-align=strict
164           -Wcast-function-type  -Wcast-qual -Wchar-subscripts -Wclobbered
165           -Wcomment -Wconversion  -Wno-coverage-mismatch  -Wno-cpp
166           -Wdangling-else  -Wdate-time -Wno-deprecated
167           -Wno-deprecated-declarations  -Wno-designated-init
168           -Wdisabled-optimization -Wno-discarded-array-qualifiers
169           -Wno-discarded-qualifiers -Wno-div-by-zero  -Wdouble-promotion
170           -Wduplicated-branches  -Wduplicated-cond -Wempty-body
171           -Wno-endif-labels  -Wenum-compare  -Wenum-conversion -Werror
172           -Werror=*  -Wexpansion-to-defined  -Wfatal-errors
173           -Wfloat-conversion  -Wfloat-equal  -Wformat  -Wformat=2
174           -Wno-format-contains-nul  -Wno-format-extra-args
175           -Wformat-nonliteral  -Wformat-overflow=n -Wformat-security
176           -Wformat-signedness  -Wformat-truncation=n -Wformat-y2k
177           -Wframe-address -Wframe-larger-than=byte-size
178           -Wno-free-nonheap-object -Wno-hsa  -Wno-if-not-aligned
179           -Wno-ignored-attributes -Wignored-qualifiers
180           -Wno-incompatible-pointer-types -Wimplicit  -Wimplicit-fallthrough
181           -Wimplicit-fallthrough=n -Wno-implicit-function-declaration
182           -Wno-implicit-int -Winit-self  -Winline  -Wno-int-conversion
183           -Wint-in-bool-context -Wno-int-to-pointer-cast
184           -Wno-invalid-memory-model -Winvalid-pch  -Wjump-misses-init
185           -Wlarger-than=byte-size -Wlogical-not-parentheses  -Wlogical-op
186           -Wlong-long -Wno-lto-type-mismatch -Wmain  -Wmaybe-uninitialized
187           -Wmemset-elt-size  -Wmemset-transposed-args
188           -Wmisleading-indentation  -Wmissing-attributes  -Wmissing-braces
189           -Wmissing-field-initializers  -Wmissing-format-attribute
190           -Wmissing-include-dirs  -Wmissing-noreturn  -Wno-missing-profile
191           -Wno-multichar  -Wmultistatement-macros  -Wnonnull
192           -Wnonnull-compare -Wnormalized=[none|id|nfc|nfkc]
193           -Wnull-dereference  -Wno-odr  -Wopenmp-simd -Wno-overflow
194           -Woverlength-strings  -Wno-override-init-side-effects -Wpacked
195           -Wno-packed-bitfield-compat  -Wpacked-not-aligned  -Wpadded
196           -Wparentheses  -Wno-pedantic-ms-format -Wpointer-arith
197           -Wno-pointer-compare  -Wno-pointer-to-int-cast -Wno-pragmas
198           -Wno-prio-ctor-dtor  -Wredundant-decls -Wrestrict
199           -Wno-return-local-addr  -Wreturn-type -Wno-scalar-storage-order
200           -Wsequence-point -Wshadow  -Wshadow=global  -Wshadow=local
201           -Wshadow=compatible-local -Wno-shadow-ivar
202           -Wno-shift-count-negative  -Wno-shift-count-overflow
203           -Wshift-negative-value -Wno-shift-overflow  -Wshift-overflow=n
204           -Wsign-compare  -Wsign-conversion -Wno-sizeof-array-argument
205           -Wsizeof-pointer-div  -Wsizeof-pointer-memaccess -Wstack-protector
206           -Wstack-usage=byte-size  -Wstrict-aliasing -Wstrict-aliasing=n
207           -Wstrict-overflow  -Wstrict-overflow=n -Wstring-compare
208           -Wstringop-overflow=n  -Wno-stringop-truncation
209           -Wsuggest-attribute=[pure|const|noreturn|format|malloc] -Wswitch
210           -Wno-switch-bool  -Wswitch-default  -Wswitch-enum
211           -Wno-switch-outside-range  -Wno-switch-unreachable  -Wsync-nand
212           -Wsystem-headers  -Wtautological-compare  -Wtrampolines
213           -Wtrigraphs -Wtype-limits  -Wundef -Wuninitialized
214           -Wunknown-pragmas -Wunsuffixed-float-constants  -Wunused
215           -Wunused-but-set-parameter  -Wunused-but-set-variable
216           -Wunused-const-variable  -Wunused-const-variable=n
217           -Wunused-function  -Wunused-label  -Wunused-local-typedefs
218           -Wunused-macros -Wunused-parameter  -Wno-unused-result
219           -Wunused-value  -Wunused-variable -Wno-varargs  -Wvariadic-macros
220           -Wvector-operation-performance -Wvla  -Wvla-larger-than=byte-size
221           -Wno-vla-larger-than -Wvolatile-register-var  -Wwrite-strings
222           -Wzero-length-bounds
223
224       Static Analyzer Options
225           -fanalyzer -fanalyzer-call-summaries -fanalyzer-checker=name
226           -fanalyzer-fine-grained -fanalyzer-state-merge
227           -fanalyzer-state-purge -fanalyzer-transitivity
228           -fanalyzer-verbose-edges -fanalyzer-verbose-state-changes
229           -fanalyzer-verbosity=level -fdump-analyzer -fdump-analyzer-stderr
230           -fdump-analyzer-callgraph -fdump-analyzer-exploded-graph
231           -fdump-analyzer-exploded-nodes -fdump-analyzer-exploded-nodes-2
232           -fdump-analyzer-exploded-nodes-3 -fdump-analyzer-state-purge
233           -fdump-analyzer-supergraph -Wno-analyzer-double-fclose
234           -Wno-analyzer-double-free
235           -Wno-analyzer-exposure-through-output-file -Wno-analyzer-file-leak
236           -Wno-analyzer-free-of-non-heap -Wno-analyzer-malloc-leak
237           -Wno-analyzer-null-argument -Wno-analyzer-null-dereference
238           -Wno-analyzer-possible-null-argument
239           -Wno-analyzer-possible-null-dereference
240           -Wno-analyzer-stale-setjmp-buffer -Wno-analyzer-tainted-array-index
241           -Wanalyzer-too-complex
242           -Wno-analyzer-unsafe-call-within-signal-handler
243           -Wno-analyzer-use-after-free
244           -Wno-analyzer-use-of-pointer-in-stale-stack-frame
245           -Wno-analyzer-use-of-uninitialized-value
246
247       C and Objective-C-only Warning Options
248           -Wbad-function-cast  -Wmissing-declarations
249           -Wmissing-parameter-type  -Wmissing-prototypes  -Wnested-externs
250           -Wold-style-declaration  -Wold-style-definition -Wstrict-prototypes
251           -Wtraditional  -Wtraditional-conversion
252           -Wdeclaration-after-statement  -Wpointer-sign
253
254       Debugging Options
255           -g  -glevel  -gdwarf  -gdwarf-version -ggdb  -grecord-gcc-switches
256           -gno-record-gcc-switches -gstabs  -gstabs+  -gstrict-dwarf
257           -gno-strict-dwarf -gas-loc-support  -gno-as-loc-support
258           -gas-locview-support  -gno-as-locview-support -gcolumn-info
259           -gno-column-info -gstatement-frontiers  -gno-statement-frontiers
260           -gvariable-location-views  -gno-variable-location-views
261           -ginternal-reset-location-views  -gno-internal-reset-location-views
262           -ginline-points  -gno-inline-points -gvms  -gxcoff  -gxcoff+
263           -gz[=type] -gsplit-dwarf  -gdescribe-dies  -gno-describe-dies
264           -fdebug-prefix-map=old=new  -fdebug-types-section
265           -fno-eliminate-unused-debug-types -femit-struct-debug-baseonly
266           -femit-struct-debug-reduced -femit-struct-debug-detailed[=spec-
267           list] -fno-eliminate-unused-debug-symbols
268           -femit-class-debug-always -fno-merge-debug-strings
269           -fno-dwarf2-cfi-asm -fvar-tracking  -fvar-tracking-assignments
270
271       Optimization Options
272           -faggressive-loop-optimizations -falign-functions[=n[:m:[n2[:m2]]]]
273           -falign-jumps[=n[:m:[n2[:m2]]]] -falign-labels[=n[:m:[n2[:m2]]]]
274           -falign-loops[=n[:m:[n2[:m2]]]] -fno-allocation-dce
275           -fallow-store-data-races -fassociative-math  -fauto-profile
276           -fauto-profile[=path] -fauto-inc-dec  -fbranch-probabilities
277           -fcaller-saves -fcombine-stack-adjustments  -fconserve-stack
278           -fcompare-elim  -fcprop-registers  -fcrossjumping
279           -fcse-follow-jumps  -fcse-skip-blocks  -fcx-fortran-rules
280           -fcx-limited-range -fdata-sections  -fdce  -fdelayed-branch
281           -fdelete-null-pointer-checks  -fdevirtualize
282           -fdevirtualize-speculatively -fdevirtualize-at-ltrans  -fdse
283           -fearly-inlining  -fipa-sra  -fexpensive-optimizations
284           -ffat-lto-objects -ffast-math  -ffinite-math-only  -ffloat-store
285           -fexcess-precision=style -ffinite-loops -fforward-propagate
286           -ffp-contract=style  -ffunction-sections -fgcse
287           -fgcse-after-reload  -fgcse-las  -fgcse-lm  -fgraphite-identity
288           -fgcse-sm  -fhoist-adjacent-loads  -fif-conversion -fif-conversion2
289           -findirect-inlining -finline-functions
290           -finline-functions-called-once  -finline-limit=n
291           -finline-small-functions  -fipa-cp  -fipa-cp-clone -fipa-bit-cp
292           -fipa-vrp  -fipa-pta  -fipa-profile  -fipa-pure-const
293           -fipa-reference  -fipa-reference-addressable -fipa-stack-alignment
294           -fipa-icf  -fira-algorithm=algorithm -flive-patching=level
295           -fira-region=region  -fira-hoist-pressure -fira-loop-pressure
296           -fno-ira-share-save-slots -fno-ira-share-spill-slots
297           -fisolate-erroneous-paths-dereference
298           -fisolate-erroneous-paths-attribute -fivopts
299           -fkeep-inline-functions  -fkeep-static-functions
300           -fkeep-static-consts  -flimit-function-alignment
301           -flive-range-shrinkage -floop-block  -floop-interchange
302           -floop-strip-mine -floop-unroll-and-jam  -floop-nest-optimize
303           -floop-parallelize-all  -flra-remat  -flto  -flto-compression-level
304           -flto-partition=alg  -fmerge-all-constants -fmerge-constants
305           -fmodulo-sched  -fmodulo-sched-allow-regmoves
306           -fmove-loop-invariants  -fno-branch-count-reg -fno-defer-pop
307           -fno-fp-int-builtin-inexact  -fno-function-cse
308           -fno-guess-branch-probability  -fno-inline  -fno-math-errno
309           -fno-peephole -fno-peephole2  -fno-printf-return-value
310           -fno-sched-interblock -fno-sched-spec  -fno-signed-zeros
311           -fno-toplevel-reorder  -fno-trapping-math
312           -fno-zero-initialized-in-bss -fomit-frame-pointer
313           -foptimize-sibling-calls -fpartial-inlining  -fpeel-loops
314           -fpredictive-commoning -fprefetch-loop-arrays -fprofile-correction
315           -fprofile-use  -fprofile-use=path -fprofile-partial-training
316           -fprofile-values -fprofile-reorder-functions -freciprocal-math
317           -free  -frename-registers  -freorder-blocks
318           -freorder-blocks-algorithm=algorithm -freorder-blocks-and-partition
319           -freorder-functions -frerun-cse-after-loop
320           -freschedule-modulo-scheduled-loops -frounding-math
321           -fsave-optimization-record -fsched2-use-superblocks
322           -fsched-pressure -fsched-spec-load  -fsched-spec-load-dangerous
323           -fsched-stalled-insns-dep[=n]  -fsched-stalled-insns[=n]
324           -fsched-group-heuristic  -fsched-critical-path-heuristic
325           -fsched-spec-insn-heuristic  -fsched-rank-heuristic
326           -fsched-last-insn-heuristic  -fsched-dep-count-heuristic
327           -fschedule-fusion -fschedule-insns  -fschedule-insns2
328           -fsection-anchors -fselective-scheduling  -fselective-scheduling2
329           -fsel-sched-pipelining  -fsel-sched-pipelining-outer-loops
330           -fsemantic-interposition  -fshrink-wrap  -fshrink-wrap-separate
331           -fsignaling-nans -fsingle-precision-constant
332           -fsplit-ivs-in-unroller  -fsplit-loops -fsplit-paths
333           -fsplit-wide-types  -fsplit-wide-types-early  -fssa-backprop
334           -fssa-phiopt -fstdarg-opt  -fstore-merging  -fstrict-aliasing
335           -fthread-jumps  -ftracer  -ftree-bit-ccp -ftree-builtin-call-dce
336           -ftree-ccp  -ftree-ch -ftree-coalesce-vars  -ftree-copy-prop
337           -ftree-dce  -ftree-dominator-opts -ftree-dse  -ftree-forwprop
338           -ftree-fre  -fcode-hoisting -ftree-loop-if-convert  -ftree-loop-im
339           -ftree-phiprop  -ftree-loop-distribution
340           -ftree-loop-distribute-patterns -ftree-loop-ivcanon
341           -ftree-loop-linear  -ftree-loop-optimize -ftree-loop-vectorize
342           -ftree-parallelize-loops=n  -ftree-pre  -ftree-partial-pre
343           -ftree-pta -ftree-reassoc  -ftree-scev-cprop  -ftree-sink
344           -ftree-slsr  -ftree-sra -ftree-switch-conversion  -ftree-tail-merge
345           -ftree-ter  -ftree-vectorize  -ftree-vrp  -funconstrained-commons
346           -funit-at-a-time  -funroll-all-loops  -funroll-loops
347           -funsafe-math-optimizations  -funswitch-loops -fipa-ra
348           -fvariable-expansion-in-unroller  -fvect-cost-model  -fvpt -fweb
349           -fwhole-program  -fwpa  -fuse-linker-plugin --param name=value -O
350           -O0  -O1  -O2  -O3  -Os  -Ofast  -Og
351
352       Program Instrumentation Options
353           -p  -pg  -fprofile-arcs  --coverage  -ftest-coverage
354           -fprofile-abs-path -fprofile-dir=path  -fprofile-generate
355           -fprofile-generate=path -fprofile-note=path
356           -fprofile-prefix-path=path -fprofile-update=method
357           -fprofile-filter-files=regex -fprofile-exclude-files=regex
358           -fprofile-reproducible=[multithreaded|parallel-runs|serial]
359           -fsanitize=style  -fsanitize-recover  -fsanitize-recover=style
360           -fasan-shadow-offset=number  -fsanitize-sections=s1,s2,...
361           -fsanitize-undefined-trap-on-error  -fbounds-check
362           -fcf-protection=[full|branch|return|none|check] -fstack-protector
363           -fstack-protector-all  -fstack-protector-strong
364           -fstack-protector-explicit  -fstack-check
365           -fstack-limit-register=reg  -fstack-limit-symbol=sym
366           -fno-stack-limit  -fsplit-stack -fvtable-verify=[std|preinit|none]
367           -fvtv-counts  -fvtv-debug -finstrument-functions
368           -finstrument-functions-exclude-function-list=sym,sym,...
369           -finstrument-functions-exclude-file-list=file,file,...
370
371       Preprocessor Options
372           -Aquestion=answer -A-question[=answer] -C  -CC  -Dmacro[=defn] -dD
373           -dI  -dM  -dN  -dU -fdebug-cpp  -fdirectives-only
374           -fdollars-in-identifiers -fexec-charset=charset
375           -fextended-identifiers -finput-charset=charset
376           -fmacro-prefix-map=old=new -fmax-include-depth=depth
377           -fno-canonical-system-headers  -fpch-deps  -fpch-preprocess
378           -fpreprocessed  -ftabstop=width  -ftrack-macro-expansion
379           -fwide-exec-charset=charset  -fworking-directory -H  -imacros file
380           -include file -M  -MD  -MF  -MG  -MM  -MMD  -MP  -MQ  -MT
381           -no-integrated-cpp  -P  -pthread  -remap -traditional
382           -traditional-cpp  -trigraphs -Umacro  -undef -Wp,option
383           -Xpreprocessor option
384
385       Assembler Options
386           -Wa,option  -Xassembler option
387
388       Linker Options
389           object-file-name  -fuse-ld=linker  -llibrary -nostartfiles
390           -nodefaultlibs  -nolibc  -nostdlib -e entry  --entry=entry -pie
391           -pthread  -r  -rdynamic -s  -static  -static-pie  -static-libgcc
392           -static-libstdc++ -static-libasan  -static-libtsan  -static-liblsan
393           -static-libubsan -shared  -shared-libgcc  -symbolic -T script
394           -Wl,option  -Xlinker option -u symbol  -z keyword
395
396       Directory Options
397           -Bprefix  -Idir  -I- -idirafter dir -imacros file  -imultilib dir
398           -iplugindir=dir  -iprefix file -iquote dir  -isysroot dir  -isystem
399           dir -iwithprefix dir  -iwithprefixbefore dir -Ldir
400           -no-canonical-prefixes  --no-sysroot-suffix -nostdinc  -nostdinc++
401           --sysroot=dir
402
403       Code Generation Options
404           -fcall-saved-reg  -fcall-used-reg -ffixed-reg  -fexceptions
405           -fnon-call-exceptions  -fdelete-dead-exceptions  -funwind-tables
406           -fasynchronous-unwind-tables -fno-gnu-unique
407           -finhibit-size-directive  -fcommon  -fno-ident -fpcc-struct-return
408           -fpic  -fPIC  -fpie  -fPIE  -fno-plt -fno-jump-tables
409           -frecord-gcc-switches -freg-struct-return  -fshort-enums
410           -fshort-wchar -fverbose-asm  -fpack-struct[=n] -fleading-underscore
411           -ftls-model=model -fstack-reuse=reuse_level -ftrampolines  -ftrapv
412           -fwrapv -fvisibility=[default|internal|hidden|protected]
413           -fstrict-volatile-bitfields  -fsync-libcalls
414
415       Developer Options
416           -dletters  -dumpspecs  -dumpmachine  -dumpversion -dumpfullversion
417           -fcallgraph-info[=su,da] -fchecking  -fchecking=n -fdbg-cnt-list
418           -fdbg-cnt=counter-value-list -fdisable-ipa-pass_name
419           -fdisable-rtl-pass_name -fdisable-rtl-pass-name=range-list
420           -fdisable-tree-pass_name -fdisable-tree-pass-name=range-list
421           -fdump-debug  -fdump-earlydebug -fdump-noaddr  -fdump-unnumbered
422           -fdump-unnumbered-links -fdump-final-insns[=file] -fdump-ipa-all
423           -fdump-ipa-cgraph  -fdump-ipa-inline -fdump-lang-all
424           -fdump-lang-switch -fdump-lang-switch-options
425           -fdump-lang-switch-options=filename -fdump-passes -fdump-rtl-pass
426           -fdump-rtl-pass=filename -fdump-statistics -fdump-tree-all
427           -fdump-tree-switch -fdump-tree-switch-options
428           -fdump-tree-switch-options=filename -fcompare-debug[=opts]
429           -fcompare-debug-second -fenable-kind-pass -fenable-kind-pass=range-
430           list -fira-verbose=n -flto-report  -flto-report-wpa
431           -fmem-report-wpa -fmem-report  -fpre-ipa-mem-report
432           -fpost-ipa-mem-report -fopt-info  -fopt-info-options[=file]
433           -fprofile-report -frandom-seed=string  -fsched-verbose=n
434           -fsel-sched-verbose  -fsel-sched-dump-cfg
435           -fsel-sched-pipelining-verbose -fstats  -fstack-usage
436           -ftime-report  -ftime-report-details
437           -fvar-tracking-assignments-toggle  -gtoggle
438           -print-file-name=library  -print-libgcc-file-name
439           -print-multi-directory  -print-multi-lib  -print-multi-os-directory
440           -print-prog-name=program  -print-search-dirs  -Q -print-sysroot
441           -print-sysroot-headers-suffix -save-temps  -save-temps=cwd
442           -save-temps=obj  -time[=file]
443
444       Machine-Dependent Options
445           AArch64 Options -mabi=name  -mbig-endian  -mlittle-endian
446           -mgeneral-regs-only -mcmodel=tiny  -mcmodel=small  -mcmodel=large
447           -mstrict-align  -mno-strict-align -momit-leaf-frame-pointer
448           -mtls-dialect=desc  -mtls-dialect=traditional -mtls-size=size
449           -mfix-cortex-a53-835769  -mfix-cortex-a53-843419
450           -mlow-precision-recip-sqrt  -mlow-precision-sqrt
451           -mlow-precision-div -mpc-relative-literal-loads
452           -msign-return-address=scope -mbranch-protection=none|standard|pac-
453           ret[+leaf +b-key]|bti -mharden-sls=opts -march=name  -mcpu=name
454           -mtune=name -moverride=string  -mverbose-cost-dump
455           -mstack-protector-guard=guard -mstack-protector-guard-reg=sysreg
456           -mstack-protector-guard-offset=offset -mtrack-speculation
457           -moutline-atomics
458
459           Adapteva Epiphany Options -mhalf-reg-file  -mprefer-short-insn-regs
460           -mbranch-cost=num  -mcmove  -mnops=num  -msoft-cmpsf -msplit-lohi
461           -mpost-inc  -mpost-modify  -mstack-offset=num -mround-nearest
462           -mlong-calls  -mshort-calls  -msmall16 -mfp-mode=mode
463           -mvect-double  -max-vect-align=num -msplit-vecmove-early
464           -m1reg-reg
465
466           AMD GCN Options -march=gpu -mtune=gpu -mstack-size=bytes
467
468           ARC Options -mbarrel-shifter  -mjli-always -mcpu=cpu  -mA6
469           -mARC600  -mA7  -mARC700 -mdpfp  -mdpfp-compact  -mdpfp-fast
470           -mno-dpfp-lrsr -mea  -mno-mpy  -mmul32x16  -mmul64  -matomic -mnorm
471           -mspfp  -mspfp-compact  -mspfp-fast  -msimd  -msoft-float  -mswap
472           -mcrc  -mdsp-packa  -mdvbf  -mlock  -mmac-d16  -mmac-24  -mrtsc
473           -mswape -mtelephony  -mxy  -misize  -mannotate-align  -marclinux
474           -marclinux_prof -mlong-calls  -mmedium-calls  -msdata
475           -mirq-ctrl-saved -mrgf-banked-regs  -mlpc-width=width  -G num
476           -mvolatile-cache  -mtp-regno=regno -malign-call  -mauto-modify-reg
477           -mbbit-peephole  -mno-brcc -mcase-vector-pcrel  -mcompact-casesi
478           -mno-cond-exec  -mearly-cbranchsi -mexpand-adddi  -mindexed-loads
479           -mlra  -mlra-priority-none -mlra-priority-compact mlra-priority-
480           noncompact  -mmillicode -mmixed-code  -mq-class  -mRcq  -mRcw
481           -msize-level=level -mtune=cpu  -mmultcost=num  -mcode-density-frame
482           -munalign-prob-threshold=probability  -mmpy-option=multo -mdiv-rem
483           -mcode-density  -mll64  -mfpu=fpu  -mrf16  -mbranch-index
484
485           ARM Options -mapcs-frame  -mno-apcs-frame -mabi=name
486           -mapcs-stack-check  -mno-apcs-stack-check -mapcs-reentrant
487           -mno-apcs-reentrant -mgeneral-regs-only -msched-prolog
488           -mno-sched-prolog -mlittle-endian  -mbig-endian -mbe8  -mbe32
489           -mfloat-abi=name -mfp16-format=name -mthumb-interwork
490           -mno-thumb-interwork -mcpu=name  -march=name  -mfpu=name
491           -mtune=name  -mprint-tune-info -mstructure-size-boundary=n
492           -mabort-on-noreturn -mlong-calls  -mno-long-calls -msingle-pic-base
493           -mno-single-pic-base -mpic-register=reg -mnop-fun-dllimport
494           -mpoke-function-name -mthumb  -marm  -mflip-thumb -mtpcs-frame
495           -mtpcs-leaf-frame -mcaller-super-interworking
496           -mcallee-super-interworking -mtp=name  -mtls-dialect=dialect
497           -mword-relocations -mfix-cortex-m3-ldrd -munaligned-access
498           -mneon-for-64bits -mslow-flash-data -masm-syntax-unified
499           -mrestrict-it -mverbose-cost-dump -mpure-code -mcmse -mfdpic
500
501           AVR Options -mmcu=mcu  -mabsdata  -maccumulate-args
502           -mbranch-cost=cost -mcall-prologues  -mgas-isr-prologues  -mint8
503           -mdouble=bits -mlong-double=bits -mn_flash=size  -mno-interrupts
504           -mmain-is-OS_task  -mrelax  -mrmw  -mstrict-X  -mtiny-stack
505           -mfract-convert-truncate -mshort-calls  -nodevicelib
506           -nodevicespecs -Waddr-space-convert  -Wmisspelled-isr
507
508           Blackfin Options -mcpu=cpu[-sirevision] -msim
509           -momit-leaf-frame-pointer  -mno-omit-leaf-frame-pointer
510           -mspecld-anomaly  -mno-specld-anomaly  -mcsync-anomaly
511           -mno-csync-anomaly -mlow-64k  -mno-low64k  -mstack-check-l1
512           -mid-shared-library -mno-id-shared-library  -mshared-library-id=n
513           -mleaf-id-shared-library  -mno-leaf-id-shared-library -msep-data
514           -mno-sep-data  -mlong-calls  -mno-long-calls -mfast-fp
515           -minline-plt  -mmulticore  -mcorea  -mcoreb  -msdram -micplb
516
517           C6X Options -mbig-endian  -mlittle-endian  -march=cpu -msim
518           -msdata=sdata-type
519
520           CRIS Options -mcpu=cpu  -march=cpu  -mtune=cpu -mmax-stack-frame=n
521           -melinux-stacksize=n -metrax4  -metrax100  -mpdebug  -mcc-init
522           -mno-side-effects -mstack-align  -mdata-align  -mconst-align
523           -m32-bit  -m16-bit  -m8-bit  -mno-prologue-epilogue  -mno-gotplt
524           -melf  -maout  -melinux  -mlinux  -sim  -sim2 -mmul-bug-workaround
525           -mno-mul-bug-workaround
526
527           CR16 Options -mmac -mcr16cplus  -mcr16c -msim  -mint32  -mbit-ops
528           -mdata-model=model
529
530           C-SKY Options -march=arch  -mcpu=cpu -mbig-endian  -EB
531           -mlittle-endian  -EL -mhard-float  -msoft-float  -mfpu=fpu
532           -mdouble-float  -mfdivdu -melrw  -mistack  -mmp  -mcp  -mcache
533           -msecurity  -mtrust -mdsp  -medsp  -mvdsp -mdiv  -msmart
534           -mhigh-registers  -manchor -mpushpop  -mmultiple-stld  -mconstpool
535           -mstack-size  -mccrt -mbranch-cost=n  -mcse-cc  -msched-prolog
536
537           Darwin Options -all_load  -allowable_client  -arch
538           -arch_errors_fatal -arch_only  -bind_at_load  -bundle
539           -bundle_loader -client_name  -compatibility_version
540           -current_version -dead_strip -dependency-file  -dylib_file
541           -dylinker_install_name -dynamic  -dynamiclib
542           -exported_symbols_list -filelist  -flat_namespace
543           -force_cpusubtype_ALL -force_flat_namespace
544           -headerpad_max_install_names -iframework -image_base  -init
545           -install_name  -keep_private_externs -multi_module
546           -multiply_defined  -multiply_defined_unused -noall_load
547           -no_dead_strip_inits_and_terms -nofixprebinding  -nomultidefs
548           -noprebind  -noseglinkedit -pagezero_size  -prebind
549           -prebind_all_twolevel_modules -private_bundle  -read_only_relocs
550           -sectalign -sectobjectsymbols  -whyload  -seg1addr -sectcreate
551           -sectobjectsymbols  -sectorder -segaddr  -segs_read_only_addr
552           -segs_read_write_addr -seg_addr_table  -seg_addr_table_filename
553           -seglinkedit -segprot  -segs_read_only_addr  -segs_read_write_addr
554           -single_module  -static  -sub_library  -sub_umbrella
555           -twolevel_namespace  -umbrella  -undefined -unexported_symbols_list
556           -weak_reference_mismatches -whatsloaded  -F  -gused  -gfull
557           -mmacosx-version-min=version -mkernel  -mone-byte-bool
558
559           DEC Alpha Options -mno-fp-regs  -msoft-float -mieee
560           -mieee-with-inexact  -mieee-conformant -mfp-trap-mode=mode
561           -mfp-rounding-mode=mode -mtrap-precision=mode  -mbuild-constants
562           -mcpu=cpu-type  -mtune=cpu-type -mbwx  -mmax  -mfix  -mcix
563           -mfloat-vax  -mfloat-ieee -mexplicit-relocs  -msmall-data
564           -mlarge-data -msmall-text  -mlarge-text -mmemory-latency=time
565
566           eBPF Options -mbig-endian -mlittle-endian -mkernel=version
567           -mframe-limit=bytes -mxbpf
568
569           FR30 Options -msmall-model  -mno-lsim
570
571           FT32 Options -msim  -mlra  -mnodiv  -mft32b  -mcompress  -mnopm
572
573           FRV Options -mgpr-32  -mgpr-64  -mfpr-32  -mfpr-64 -mhard-float
574           -msoft-float -malloc-cc  -mfixed-cc  -mdword  -mno-dword -mdouble
575           -mno-double -mmedia  -mno-media  -mmuladd  -mno-muladd -mfdpic
576           -minline-plt  -mgprel-ro  -multilib-library-pic -mlinked-fp
577           -mlong-calls  -malign-labels -mlibrary-pic  -macc-4  -macc-8 -mpack
578           -mno-pack  -mno-eflags  -mcond-move  -mno-cond-move
579           -moptimize-membar  -mno-optimize-membar -mscc  -mno-scc
580           -mcond-exec  -mno-cond-exec -mvliw-branch  -mno-vliw-branch
581           -mmulti-cond-exec  -mno-multi-cond-exec  -mnested-cond-exec
582           -mno-nested-cond-exec  -mtomcat-stats -mTLS  -mtls -mcpu=cpu
583
584           GNU/Linux Options -mglibc  -muclibc  -mmusl  -mbionic  -mandroid
585           -tno-android-cc  -tno-android-ld
586
587           H8/300 Options -mrelax  -mh  -ms  -mn  -mexr  -mno-exr  -mint32
588           -malign-300
589
590           HPPA Options -march=architecture-type -mcaller-copies
591           -mdisable-fpregs  -mdisable-indexing -mfast-indirect-calls  -mgas
592           -mgnu-ld   -mhp-ld -mfixed-range=register-range -mjump-in-delay
593           -mlinker-opt  -mlong-calls -mlong-load-store  -mno-disable-fpregs
594           -mno-disable-indexing  -mno-fast-indirect-calls  -mno-gas
595           -mno-jump-in-delay  -mno-long-load-store -mno-portable-runtime
596           -mno-soft-float -mno-space-regs  -msoft-float  -mpa-risc-1-0
597           -mpa-risc-1-1  -mpa-risc-2-0  -mportable-runtime -mschedule=cpu-
598           type  -mspace-regs  -msio  -mwsio -munix=unix-std  -nolibdld
599           -static  -threads
600
601           IA-64 Options -mbig-endian  -mlittle-endian  -mgnu-as  -mgnu-ld
602           -mno-pic -mvolatile-asm-stop  -mregister-names  -msdata  -mno-sdata
603           -mconstant-gp  -mauto-pic  -mfused-madd
604           -minline-float-divide-min-latency
605           -minline-float-divide-max-throughput -mno-inline-float-divide
606           -minline-int-divide-min-latency -minline-int-divide-max-throughput
607           -mno-inline-int-divide -minline-sqrt-min-latency
608           -minline-sqrt-max-throughput -mno-inline-sqrt -mdwarf2-asm
609           -mearly-stop-bits -mfixed-range=register-range  -mtls-size=tls-size
610           -mtune=cpu-type  -milp32  -mlp64 -msched-br-data-spec
611           -msched-ar-data-spec  -msched-control-spec -msched-br-in-data-spec
612           -msched-ar-in-data-spec  -msched-in-control-spec -msched-spec-ldc
613           -msched-spec-control-ldc -msched-prefer-non-data-spec-insns
614           -msched-prefer-non-control-spec-insns
615           -msched-stop-bits-after-every-cycle
616           -msched-count-spec-in-critical-path
617           -msel-sched-dont-check-control-spec  -msched-fp-mem-deps-zero-cost
618           -msched-max-memory-insns-hard-limit  -msched-max-memory-insns=max-
619           insns
620
621           LM32 Options -mbarrel-shift-enabled  -mdivide-enabled
622           -mmultiply-enabled -msign-extend-enabled  -muser-enabled
623
624           M32R/D Options -m32r2  -m32rx  -m32r -mdebug -malign-loops
625           -mno-align-loops -missue-rate=number -mbranch-cost=number
626           -mmodel=code-size-model-type -msdata=sdata-type -mno-flush-func
627           -mflush-func=name -mno-flush-trap  -mflush-trap=number -G num
628
629           M32C Options -mcpu=cpu  -msim  -memregs=number
630
631           M680x0 Options -march=arch  -mcpu=cpu  -mtune=tune -m68000  -m68020
632           -m68020-40  -m68020-60  -m68030  -m68040 -m68060  -mcpu32  -m5200
633           -m5206e  -m528x  -m5307  -m5407 -mcfv4e  -mbitfield  -mno-bitfield
634           -mc68000  -mc68020 -mnobitfield  -mrtd  -mno-rtd  -mdiv  -mno-div
635           -mshort -mno-short  -mhard-float  -m68881  -msoft-float  -mpcrel
636           -malign-int  -mstrict-align  -msep-data  -mno-sep-data
637           -mshared-library-id=n  -mid-shared-library  -mno-id-shared-library
638           -mxgot  -mno-xgot  -mlong-jump-table-offsets
639
640           MCore Options -mhardlit  -mno-hardlit  -mdiv  -mno-div
641           -mrelax-immediates -mno-relax-immediates  -mwide-bitfields
642           -mno-wide-bitfields -m4byte-functions  -mno-4byte-functions
643           -mcallgraph-data -mno-callgraph-data  -mslow-bytes  -mno-slow-bytes
644           -mno-lsim -mlittle-endian  -mbig-endian  -m210  -m340
645           -mstack-increment
646
647           MeP Options -mabsdiff  -mall-opts  -maverage  -mbased=n  -mbitops
648           -mc=n  -mclip  -mconfig=name  -mcop  -mcop32  -mcop64  -mivc2 -mdc
649           -mdiv  -meb  -mel  -mio-volatile  -ml  -mleadz  -mm  -mminmax
650           -mmult  -mno-opts  -mrepeat  -ms  -msatur  -msdram  -msim
651           -msimnovec  -mtf -mtiny=n
652
653           MicroBlaze Options -msoft-float  -mhard-float  -msmall-divides
654           -mcpu=cpu -mmemcpy  -mxl-soft-mul  -mxl-soft-div  -mxl-barrel-shift
655           -mxl-pattern-compare  -mxl-stack-check  -mxl-gp-opt  -mno-clearbss
656           -mxl-multiply-high  -mxl-float-convert  -mxl-float-sqrt
657           -mbig-endian  -mlittle-endian  -mxl-reorder  -mxl-mode-app-model
658           -mpic-data-is-text-relative
659
660           MIPS Options -EL  -EB  -march=arch  -mtune=arch -mips1  -mips2
661           -mips3  -mips4  -mips32  -mips32r2  -mips32r3  -mips32r5 -mips32r6
662           -mips64  -mips64r2  -mips64r3  -mips64r5  -mips64r6 -mips16
663           -mno-mips16  -mflip-mips16 -minterlink-compressed
664           -mno-interlink-compressed -minterlink-mips16  -mno-interlink-mips16
665           -mabi=abi  -mabicalls  -mno-abicalls -mshared  -mno-shared  -mplt
666           -mno-plt  -mxgot  -mno-xgot -mgp32  -mgp64  -mfp32  -mfpxx  -mfp64
667           -mhard-float  -msoft-float -mno-float  -msingle-float
668           -mdouble-float -modd-spreg  -mno-odd-spreg -mabs=mode
669           -mnan=encoding -mdsp  -mno-dsp  -mdspr2  -mno-dspr2 -mmcu
670           -mmno-mcu -meva  -mno-eva -mvirt  -mno-virt -mxpa  -mno-xpa -mcrc
671           -mno-crc -mginv  -mno-ginv -mmicromips  -mno-micromips -mmsa
672           -mno-msa -mloongson-mmi  -mno-loongson-mmi -mloongson-ext
673           -mno-loongson-ext -mloongson-ext2  -mno-loongson-ext2 -mfpu=fpu-
674           type -msmartmips  -mno-smartmips -mpaired-single
675           -mno-paired-single  -mdmx  -mno-mdmx -mips3d  -mno-mips3d  -mmt
676           -mno-mt  -mllsc  -mno-llsc -mlong64  -mlong32  -msym32  -mno-sym32
677           -Gnum  -mlocal-sdata  -mno-local-sdata -mextern-sdata
678           -mno-extern-sdata  -mgpopt  -mno-gopt -membedded-data
679           -mno-embedded-data -muninit-const-in-rodata
680           -mno-uninit-const-in-rodata -mcode-readable=setting
681           -msplit-addresses  -mno-split-addresses -mexplicit-relocs
682           -mno-explicit-relocs -mcheck-zero-division
683           -mno-check-zero-division -mdivide-traps  -mdivide-breaks
684           -mload-store-pairs  -mno-load-store-pairs -mmemcpy  -mno-memcpy
685           -mlong-calls  -mno-long-calls -mmad  -mno-mad  -mimadd  -mno-imadd
686           -mfused-madd  -mno-fused-madd  -nocpp -mfix-24k  -mno-fix-24k
687           -mfix-r4000  -mno-fix-r4000  -mfix-r4400  -mno-fix-r4400
688           -mfix-r5900  -mno-fix-r5900 -mfix-r10000  -mno-fix-r10000
689           -mfix-rm7000  -mno-fix-rm7000 -mfix-vr4120  -mno-fix-vr4120
690           -mfix-vr4130  -mno-fix-vr4130  -mfix-sb1  -mno-fix-sb1
691           -mflush-func=func  -mno-flush-func -mbranch-cost=num
692           -mbranch-likely  -mno-branch-likely -mcompact-branches=policy
693           -mfp-exceptions  -mno-fp-exceptions -mvr4130-align
694           -mno-vr4130-align  -msynci  -mno-synci -mlxc1-sxc1  -mno-lxc1-sxc1
695           -mmadd4  -mno-madd4 -mrelax-pic-calls  -mno-relax-pic-calls
696           -mmcount-ra-address -mframe-header-opt  -mno-frame-header-opt
697
698           MMIX Options -mlibfuncs  -mno-libfuncs  -mepsilon  -mno-epsilon
699           -mabi=gnu -mabi=mmixware  -mzero-extend  -mknuthdiv
700           -mtoplevel-symbols -melf  -mbranch-predict  -mno-branch-predict
701           -mbase-addresses -mno-base-addresses  -msingle-exit
702           -mno-single-exit
703
704           MN10300 Options -mmult-bug  -mno-mult-bug -mno-am33  -mam33
705           -mam33-2  -mam34 -mtune=cpu-type -mreturn-pointer-on-d0 -mno-crt0
706           -mrelax  -mliw  -msetlb
707
708           Moxie Options -meb  -mel  -mmul.x  -mno-crt0
709
710           MSP430 Options -msim  -masm-hex  -mmcu=  -mcpu=  -mlarge  -msmall
711           -mrelax -mwarn-mcu -mcode-region=  -mdata-region= -msilicon-errata=
712           -msilicon-errata-warn= -mhwmult=  -minrt  -mtiny-printf
713
714           NDS32 Options -mbig-endian  -mlittle-endian -mreduced-regs
715           -mfull-regs -mcmov  -mno-cmov -mext-perf  -mno-ext-perf -mext-perf2
716           -mno-ext-perf2 -mext-string  -mno-ext-string -mv3push  -mno-v3push
717           -m16bit  -mno-16bit -misr-vector-size=num -mcache-block-size=num
718           -march=arch -mcmodel=code-model -mctor-dtor  -mrelax
719
720           Nios II Options -G num  -mgpopt=option  -mgpopt  -mno-gpopt
721           -mgprel-sec=regexp  -mr0rel-sec=regexp -mel  -meb -mno-bypass-cache
722           -mbypass-cache -mno-cache-volatile  -mcache-volatile
723           -mno-fast-sw-div  -mfast-sw-div -mhw-mul  -mno-hw-mul  -mhw-mulx
724           -mno-hw-mulx  -mno-hw-div  -mhw-div -mcustom-insn=N
725           -mno-custom-insn -mcustom-fpu-cfg=name -mhal  -msmallc
726           -msys-crt0=name  -msys-lib=name -march=arch  -mbmx  -mno-bmx  -mcdx
727           -mno-cdx
728
729           Nvidia PTX Options -m32  -m64  -mmainkernel  -moptimize
730
731           OpenRISC Options -mboard=name  -mnewlib  -mhard-mul  -mhard-div
732           -msoft-mul  -msoft-div -msoft-float  -mhard-float  -mdouble-float
733           -munordered-float -mcmov  -mror  -mrori  -msext  -msfimm  -mshftimm
734
735           PDP-11 Options -mfpu  -msoft-float  -mac0  -mno-ac0  -m40  -m45
736           -m10 -mint32  -mno-int16  -mint16  -mno-int32 -msplit  -munix-asm
737           -mdec-asm  -mgnu-asm  -mlra
738
739           picoChip Options -mae=ae_type  -mvliw-lookahead=N
740           -msymbol-as-address  -mno-inefficient-warnings
741
742           PowerPC Options See RS/6000 and PowerPC Options.
743
744           PRU Options -mmcu=mcu  -minrt  -mno-relax  -mloop -mabi=variant
745
746           RISC-V Options -mbranch-cost=N-instruction -mplt  -mno-plt
747           -mabi=ABI-string -mfdiv  -mno-fdiv -mdiv  -mno-div -march=ISA-
748           string -mtune=processor-string -mpreferred-stack-boundary=num
749           -msmall-data-limit=N-bytes -msave-restore  -mno-save-restore
750           -mstrict-align  -mno-strict-align -mcmodel=medlow  -mcmodel=medany
751           -mexplicit-relocs  -mno-explicit-relocs -mrelax  -mno-relax
752           -mriscv-attribute  -mmo-riscv-attribute -malign-data=type
753
754           RL78 Options -msim  -mmul=none  -mmul=g13  -mmul=g14  -mallregs
755           -mcpu=g10  -mcpu=g13  -mcpu=g14  -mg10  -mg13  -mg14
756           -m64bit-doubles  -m32bit-doubles  -msave-mduc-in-interrupts
757
758           RS/6000 and PowerPC Options -mcpu=cpu-type -mtune=cpu-type
759           -mcmodel=code-model -mpowerpc64 -maltivec  -mno-altivec
760           -mpowerpc-gpopt  -mno-powerpc-gpopt -mpowerpc-gfxopt
761           -mno-powerpc-gfxopt -mmfcrf  -mno-mfcrf  -mpopcntb  -mno-popcntb
762           -mpopcntd  -mno-popcntd -mfprnd  -mno-fprnd -mcmpb  -mno-cmpb
763           -mhard-dfp  -mno-hard-dfp -mfull-toc   -mminimal-toc
764           -mno-fp-in-toc  -mno-sum-in-toc -m64  -m32  -mxl-compat
765           -mno-xl-compat  -mpe -malign-power  -malign-natural -msoft-float
766           -mhard-float  -mmultiple  -mno-multiple -mupdate  -mno-update
767           -mavoid-indexed-addresses  -mno-avoid-indexed-addresses
768           -mfused-madd  -mno-fused-madd  -mbit-align  -mno-bit-align
769           -mstrict-align  -mno-strict-align  -mrelocatable -mno-relocatable
770           -mrelocatable-lib  -mno-relocatable-lib -mtoc  -mno-toc  -mlittle
771           -mlittle-endian  -mbig  -mbig-endian -mdynamic-no-pic  -mswdiv
772           -msingle-pic-base -mprioritize-restricted-insns=priority
773           -msched-costly-dep=dependence_type -minsert-sched-nops=scheme
774           -mcall-aixdesc  -mcall-eabi  -mcall-freebsd -mcall-linux
775           -mcall-netbsd  -mcall-openbsd -mcall-sysv  -mcall-sysv-eabi
776           -mcall-sysv-noeabi -mtraceback=traceback_type -maix-struct-return
777           -msvr4-struct-return -mabi=abi-type  -msecure-plt  -mbss-plt
778           -mlongcall  -mno-longcall  -mpltseq  -mno-pltseq
779           -mblock-move-inline-limit=num -mblock-compare-inline-limit=num
780           -mblock-compare-inline-loop-limit=num
781           -mstring-compare-inline-limit=num -misel  -mno-isel -mvrsave
782           -mno-vrsave -mmulhw  -mno-mulhw -mdlmzb  -mno-dlmzb -mprototype
783           -mno-prototype -msim  -mmvme  -mads  -myellowknife  -memb  -msdata
784           -msdata=opt  -mreadonly-in-sdata  -mvxworks  -G num -mrecip
785           -mrecip=opt  -mno-recip  -mrecip-precision -mno-recip-precision
786           -mveclibabi=type  -mfriz  -mno-friz -mpointers-to-nested-functions
787           -mno-pointers-to-nested-functions -msave-toc-indirect
788           -mno-save-toc-indirect -mpower8-fusion  -mno-mpower8-fusion
789           -mpower8-vector  -mno-power8-vector -mcrypto  -mno-crypto  -mhtm
790           -mno-htm -mquad-memory  -mno-quad-memory -mquad-memory-atomic
791           -mno-quad-memory-atomic -mcompat-align-parm  -mno-compat-align-parm
792           -mfloat128  -mno-float128  -mfloat128-hardware
793           -mno-float128-hardware -mgnu-attribute  -mno-gnu-attribute
794           -mstack-protector-guard=guard -mstack-protector-guard-reg=reg
795           -mstack-protector-guard-offset=offset -mprefixed -mno-prefixed
796           -mpcrel -mno-pcrel -mmma -mno-mmma
797
798           RX Options -m64bit-doubles  -m32bit-doubles  -fpu  -nofpu -mcpu=
799           -mbig-endian-data  -mlittle-endian-data -msmall-data -msim
800           -mno-sim -mas100-syntax  -mno-as100-syntax -mrelax
801           -mmax-constant-size= -mint-register= -mpid -mallow-string-insns
802           -mno-allow-string-insns -mjsr -mno-warn-multiple-fast-interrupts
803           -msave-acc-in-interrupts
804
805           S/390 and zSeries Options -mtune=cpu-type  -march=cpu-type
806           -mhard-float  -msoft-float  -mhard-dfp  -mno-hard-dfp
807           -mlong-double-64  -mlong-double-128 -mbackchain  -mno-backchain
808           -mpacked-stack  -mno-packed-stack -msmall-exec  -mno-small-exec
809           -mmvcle  -mno-mvcle -m64  -m31  -mdebug  -mno-debug  -mesa  -mzarch
810           -mhtm  -mvx  -mzvector -mtpf-trace  -mno-tpf-trace
811           -mtpf-trace-skip  -mno-tpf-trace-skip -mfused-madd  -mno-fused-madd
812           -mwarn-framesize  -mwarn-dynamicstack  -mstack-size  -mstack-guard
813           -mhotpatch=halfwords,halfwords
814
815           Score Options -meb  -mel -mnhwloop -muls -mmac -mscore5  -mscore5u
816           -mscore7  -mscore7d
817
818           SH Options -m1  -m2  -m2e -m2a-nofpu  -m2a-single-only  -m2a-single
819           -m2a -m3  -m3e -m4-nofpu  -m4-single-only  -m4-single  -m4
820           -m4a-nofpu  -m4a-single-only  -m4a-single  -m4a  -m4al -mb  -ml
821           -mdalign  -mrelax -mbigtable  -mfmovd  -mrenesas  -mno-renesas
822           -mnomacsave -mieee  -mno-ieee  -mbitops  -misize
823           -minline-ic_invalidate  -mpadstruct -mprefergot  -musermode
824           -multcost=number  -mdiv=strategy -mdivsi3_libfunc=name
825           -mfixed-range=register-range -maccumulate-outgoing-args
826           -matomic-model=atomic-model -mbranch-cost=num  -mzdcbranch
827           -mno-zdcbranch -mcbranch-force-delay-slot -mfused-madd
828           -mno-fused-madd  -mfsca  -mno-fsca  -mfsrra  -mno-fsrra
829           -mpretend-cmove  -mtas
830
831           Solaris 2 Options -mclear-hwcap  -mno-clear-hwcap  -mimpure-text
832           -mno-impure-text -pthreads
833
834           SPARC Options -mcpu=cpu-type -mtune=cpu-type -mcmodel=code-model
835           -mmemory-model=mem-model -m32  -m64  -mapp-regs  -mno-app-regs
836           -mfaster-structs  -mno-faster-structs  -mflat  -mno-flat -mfpu
837           -mno-fpu  -mhard-float  -msoft-float -mhard-quad-float
838           -msoft-quad-float -mstack-bias  -mno-stack-bias -mstd-struct-return
839           -mno-std-struct-return -munaligned-doubles  -mno-unaligned-doubles
840           -muser-mode  -mno-user-mode -mv8plus  -mno-v8plus  -mvis  -mno-vis
841           -mvis2  -mno-vis2  -mvis3  -mno-vis3 -mvis4  -mno-vis4  -mvis4b
842           -mno-vis4b -mcbcond  -mno-cbcond  -mfmaf  -mno-fmaf  -mfsmuld
843           -mno-fsmuld -mpopc  -mno-popc  -msubxc  -mno-subxc -mfix-at697f
844           -mfix-ut699  -mfix-ut700  -mfix-gr712rc -mlra  -mno-lra
845
846           System V Options -Qy  -Qn  -YP,paths  -Ym,dir
847
848           TILE-Gx Options -mcpu=CPU  -m32  -m64  -mbig-endian
849           -mlittle-endian -mcmodel=code-model
850
851           TILEPro Options -mcpu=cpu  -m32
852
853           V850 Options -mlong-calls  -mno-long-calls  -mep  -mno-ep
854           -mprolog-function  -mno-prolog-function  -mspace -mtda=n  -msda=n
855           -mzda=n -mapp-regs  -mno-app-regs -mdisable-callt
856           -mno-disable-callt -mv850e2v3  -mv850e2  -mv850e1  -mv850es -mv850e
857           -mv850  -mv850e3v5 -mloop -mrelax -mlong-jumps -msoft-float
858           -mhard-float -mgcc-abi -mrh850-abi -mbig-switch
859
860           VAX Options -mg  -mgnu  -munix
861
862           Visium Options -mdebug  -msim  -mfpu  -mno-fpu  -mhard-float
863           -msoft-float -mcpu=cpu-type  -mtune=cpu-type  -msv-mode
864           -muser-mode
865
866           VMS Options -mvms-return-codes  -mdebug-main=prefix  -mmalloc64
867           -mpointer-size=size
868
869           VxWorks Options -mrtp  -non-static  -Bstatic  -Bdynamic -Xbind-lazy
870           -Xbind-now
871
872           x86 Options -mtune=cpu-type  -march=cpu-type -mtune-ctrl=feature-
873           list  -mdump-tune-features  -mno-default -mfpmath=unit
874           -masm=dialect  -mno-fancy-math-387 -mno-fp-ret-in-387  -m80387
875           -mhard-float  -msoft-float -mno-wide-multiply  -mrtd
876           -malign-double -mpreferred-stack-boundary=num
877           -mincoming-stack-boundary=num -mcld  -mcx16  -msahf  -mmovbe
878           -mcrc32 -mrecip  -mrecip=opt -mvzeroupper  -mprefer-avx128
879           -mprefer-vector-width=opt -mmmx  -msse  -msse2  -msse3  -mssse3
880           -msse4.1  -msse4.2  -msse4  -mavx -mavx2  -mavx512f  -mavx512pf
881           -mavx512er  -mavx512cd  -mavx512vl -mavx512bw  -mavx512dq
882           -mavx512ifma  -mavx512vbmi  -msha  -maes -mpclmul  -mfsgsbase
883           -mrdrnd  -mf16c  -mfma  -mpconfig  -mwbnoinvd -mptwrite
884           -mprefetchwt1  -mclflushopt  -mclwb  -mxsavec  -mxsaves -msse4a
885           -m3dnow  -m3dnowa  -mpopcnt  -mabm  -mbmi  -mtbm  -mfma4  -mxop
886           -madx  -mlzcnt  -mbmi2  -mfxsr  -mxsave  -mxsaveopt  -mrtm  -mhle
887           -mlwp -mmwaitx  -mclzero  -mpku  -mthreads  -mgfni  -mvaes
888           -mwaitpkg -mshstk -mmanual-endbr -mforce-indirect-call
889           -mavx512vbmi2 -mavx512bf16 -menqcmd -mvpclmulqdq  -mavx512bitalg
890           -mmovdiri  -mmovdir64b  -mavx512vpopcntdq -mavx5124fmaps
891           -mavx512vnni  -mavx5124vnniw  -mprfchw  -mrdpid -mrdseed  -msgx
892           -mavx512vp2intersect -mcldemote  -mms-bitfields
893           -mno-align-stringops  -minline-all-stringops
894           -minline-stringops-dynamically  -mstringop-strategy=alg
895           -mmemcpy-strategy=strategy  -mmemset-strategy=strategy -mpush-args
896           -maccumulate-outgoing-args  -m128bit-long-double
897           -m96bit-long-double  -mlong-double-64  -mlong-double-80
898           -mlong-double-128 -mregparm=num  -msseregparm -mveclibabi=type
899           -mvect8-ret-in-mem -mpc32  -mpc64  -mpc80  -mstackrealign
900           -momit-leaf-frame-pointer  -mno-red-zone  -mno-tls-direct-seg-refs
901           -mcmodel=code-model  -mabi=name  -maddress-mode=mode -m32  -m64
902           -mx32  -m16  -miamcu  -mlarge-data-threshold=num -msse2avx
903           -mfentry  -mrecord-mcount  -mnop-mcount  -m8bit-idiv
904           -minstrument-return=type -mfentry-name=name -mfentry-section=name
905           -mavx256-split-unaligned-load  -mavx256-split-unaligned-store
906           -malign-data=type  -mstack-protector-guard=guard
907           -mstack-protector-guard-reg=reg
908           -mstack-protector-guard-offset=offset
909           -mstack-protector-guard-symbol=symbol -mgeneral-regs-only
910           -mcall-ms2sysv-xlogues -mindirect-branch=choice
911           -mfunction-return=choice -mindirect-branch-register
912
913           x86 Windows Options -mconsole  -mcygwin  -mno-cygwin  -mdll
914           -mnop-fun-dllimport  -mthread -municode  -mwin32  -mwindows
915           -fno-set-stack-executable
916
917           Xstormy16 Options -msim
918
919           Xtensa Options -mconst16  -mno-const16 -mfused-madd
920           -mno-fused-madd -mforce-no-pic -mserialize-volatile
921           -mno-serialize-volatile -mtext-section-literals
922           -mno-text-section-literals -mauto-litpools  -mno-auto-litpools
923           -mtarget-align  -mno-target-align -mlongcalls  -mno-longcalls
924
925           zSeries Options See S/390 and zSeries Options.
926
927   Options Controlling the Kind of Output
928       Compilation can involve up to four stages: preprocessing, compilation
929       proper, assembly and linking, always in that order.  GCC is capable of
930       preprocessing and compiling several files either into several assembler
931       input files, or into one assembler input file; then each assembler
932       input file produces an object file, and linking combines all the object
933       files (those newly compiled, and those specified as input) into an
934       executable file.
935
936       For any given input file, the file name suffix determines what kind of
937       compilation is done:
938
939       file.c
940           C source code that must be preprocessed.
941
942       file.i
943           C source code that should not be preprocessed.
944
945       file.ii
946           C++ source code that should not be preprocessed.
947
948       file.m
949           Objective-C source code.  Note that you must link with the libobjc
950           library to make an Objective-C program work.
951
952       file.mi
953           Objective-C source code that should not be preprocessed.
954
955       file.mm
956       file.M
957           Objective-C++ source code.  Note that you must link with the
958           libobjc library to make an Objective-C++ program work.  Note that
959           .M refers to a literal capital M.
960
961       file.mii
962           Objective-C++ source code that should not be preprocessed.
963
964       file.h
965           C, C++, Objective-C or Objective-C++ header file to be turned into
966           a precompiled header (default), or C, C++ header file to be turned
967           into an Ada spec (via the -fdump-ada-spec switch).
968
969       file.cc
970       file.cp
971       file.cxx
972       file.cpp
973       file.CPP
974       file.c++
975       file.C
976           C++ source code that must be preprocessed.  Note that in .cxx, the
977           last two letters must both be literally x.  Likewise, .C refers to
978           a literal capital C.
979
980       file.mm
981       file.M
982           Objective-C++ source code that must be preprocessed.
983
984       file.mii
985           Objective-C++ source code that should not be preprocessed.
986
987       file.hh
988       file.H
989       file.hp
990       file.hxx
991       file.hpp
992       file.HPP
993       file.h++
994       file.tcc
995           C++ header file to be turned into a precompiled header or Ada spec.
996
997       file.f
998       file.for
999       file.ftn
1000           Fixed form Fortran source code that should not be preprocessed.
1001
1002       file.F
1003       file.FOR
1004       file.fpp
1005       file.FPP
1006       file.FTN
1007           Fixed form Fortran source code that must be preprocessed (with the
1008           traditional preprocessor).
1009
1010       file.f90
1011       file.f95
1012       file.f03
1013       file.f08
1014           Free form Fortran source code that should not be preprocessed.
1015
1016       file.F90
1017       file.F95
1018       file.F03
1019       file.F08
1020           Free form Fortran source code that must be preprocessed (with the
1021           traditional preprocessor).
1022
1023       file.go
1024           Go source code.
1025
1026       file.brig
1027           BRIG files (binary representation of HSAIL).
1028
1029       file.d
1030           D source code.
1031
1032       file.di
1033           D interface file.
1034
1035       file.dd
1036           D documentation code (Ddoc).
1037
1038       file.ads
1039           Ada source code file that contains a library unit declaration (a
1040           declaration of a package, subprogram, or generic, or a generic
1041           instantiation), or a library unit renaming declaration (a package,
1042           generic, or subprogram renaming declaration).  Such files are also
1043           called specs.
1044
1045       file.adb
1046           Ada source code file containing a library unit body (a subprogram
1047           or package body).  Such files are also called bodies.
1048
1049       file.s
1050           Assembler code.
1051
1052       file.S
1053       file.sx
1054           Assembler code that must be preprocessed.
1055
1056       other
1057           An object file to be fed straight into linking.  Any file name with
1058           no recognized suffix is treated this way.
1059
1060       You can specify the input language explicitly with the -x option:
1061
1062       -x language
1063           Specify explicitly the language for the following input files
1064           (rather than letting the compiler choose a default based on the
1065           file name suffix).  This option applies to all following input
1066           files until the next -x option.  Possible values for language are:
1067
1068                   c  c-header  cpp-output
1069                   c++  c++-header  c++-cpp-output
1070                   objective-c  objective-c-header  objective-c-cpp-output
1071                   objective-c++ objective-c++-header objective-c++-cpp-output
1072                   assembler  assembler-with-cpp
1073                   ada
1074                   d
1075                   f77  f77-cpp-input f95  f95-cpp-input
1076                   go
1077                   brig
1078
1079       -x none
1080           Turn off any specification of a language, so that subsequent files
1081           are handled according to their file name suffixes (as they are if
1082           -x has not been used at all).
1083
1084       If you only want some of the stages of compilation, you can use -x (or
1085       filename suffixes) to tell gcc where to start, and one of the options
1086       -c, -S, or -E to say where gcc is to stop.  Note that some combinations
1087       (for example, -x cpp-output -E) instruct gcc to do nothing at all.
1088
1089       -c  Compile or assemble the source files, but do not link.  The linking
1090           stage simply is not done.  The ultimate output is in the form of an
1091           object file for each source file.
1092
1093           By default, the object file name for a source file is made by
1094           replacing the suffix .c, .i, .s, etc., with .o.
1095
1096           Unrecognized input files, not requiring compilation or assembly,
1097           are ignored.
1098
1099       -S  Stop after the stage of compilation proper; do not assemble.  The
1100           output is in the form of an assembler code file for each non-
1101           assembler input file specified.
1102
1103           By default, the assembler file name for a source file is made by
1104           replacing the suffix .c, .i, etc., with .s.
1105
1106           Input files that don't require compilation are ignored.
1107
1108       -E  Stop after the preprocessing stage; do not run the compiler proper.
1109           The output is in the form of preprocessed source code, which is
1110           sent to the standard output.
1111
1112           Input files that don't require preprocessing are ignored.
1113
1114       -o file
1115           Place output in file file.  This applies to whatever sort of output
1116           is being produced, whether it be an executable file, an object
1117           file, an assembler file or preprocessed C code.
1118
1119           If -o is not specified, the default is to put an executable file in
1120           a.out, the object file for source.suffix in source.o, its assembler
1121           file in source.s, a precompiled header file in source.suffix.gch,
1122           and all preprocessed C source on standard output.
1123
1124       -v  Print (on standard error output) the commands executed to run the
1125           stages of compilation.  Also print the version number of the
1126           compiler driver program and of the preprocessor and the compiler
1127           proper.
1128
1129       -###
1130           Like -v except the commands are not executed and arguments are
1131           quoted unless they contain only alphanumeric characters or "./-_".
1132           This is useful for shell scripts to capture the driver-generated
1133           command lines.
1134
1135       --help
1136           Print (on the standard output) a description of the command-line
1137           options understood by gcc.  If the -v option is also specified then
1138           --help is also passed on to the various processes invoked by gcc,
1139           so that they can display the command-line options they accept.  If
1140           the -Wextra option has also been specified (prior to the --help
1141           option), then command-line options that have no documentation
1142           associated with them are also displayed.
1143
1144       --target-help
1145           Print (on the standard output) a description of target-specific
1146           command-line options for each tool.  For some targets extra target-
1147           specific information may also be printed.
1148
1149       --help={class|[^]qualifier}[,...]
1150           Print (on the standard output) a description of the command-line
1151           options understood by the compiler that fit into all specified
1152           classes and qualifiers.  These are the supported classes:
1153
1154           optimizers
1155               Display all of the optimization options supported by the
1156               compiler.
1157
1158           warnings
1159               Display all of the options controlling warning messages
1160               produced by the compiler.
1161
1162           target
1163               Display target-specific options.  Unlike the --target-help
1164               option however, target-specific options of the linker and
1165               assembler are not displayed.  This is because those tools do
1166               not currently support the extended --help= syntax.
1167
1168           params
1169               Display the values recognized by the --param option.
1170
1171           language
1172               Display the options supported for language, where language is
1173               the name of one of the languages supported in this version of
1174               GCC.  If an option is supported by all languages, one needs to
1175               select common class.
1176
1177           common
1178               Display the options that are common to all languages.
1179
1180           These are the supported qualifiers:
1181
1182           undocumented
1183               Display only those options that are undocumented.
1184
1185           joined
1186               Display options taking an argument that appears after an equal
1187               sign in the same continuous piece of text, such as:
1188               --help=target.
1189
1190           separate
1191               Display options taking an argument that appears as a separate
1192               word following the original option, such as: -o output-file.
1193
1194           Thus for example to display all the undocumented target-specific
1195           switches supported by the compiler, use:
1196
1197                   --help=target,undocumented
1198
1199           The sense of a qualifier can be inverted by prefixing it with the ^
1200           character, so for example to display all binary warning options
1201           (i.e., ones that are either on or off and that do not take an
1202           argument) that have a description, use:
1203
1204                   --help=warnings,^joined,^undocumented
1205
1206           The argument to --help= should not consist solely of inverted
1207           qualifiers.
1208
1209           Combining several classes is possible, although this usually
1210           restricts the output so much that there is nothing to display.  One
1211           case where it does work, however, is when one of the classes is
1212           target.  For example, to display all the target-specific
1213           optimization options, use:
1214
1215                   --help=target,optimizers
1216
1217           The --help= option can be repeated on the command line.  Each
1218           successive use displays its requested class of options, skipping
1219           those that have already been displayed.  If --help is also
1220           specified anywhere on the command line then this takes precedence
1221           over any --help= option.
1222
1223           If the -Q option appears on the command line before the --help=
1224           option, then the descriptive text displayed by --help= is changed.
1225           Instead of describing the displayed options, an indication is given
1226           as to whether the option is enabled, disabled or set to a specific
1227           value (assuming that the compiler knows this at the point where the
1228           --help= option is used).
1229
1230           Here is a truncated example from the ARM port of gcc:
1231
1232                     % gcc -Q -mabi=2 --help=target -c
1233                     The following options are target specific:
1234                     -mabi=                                2
1235                     -mabort-on-noreturn                   [disabled]
1236                     -mapcs                                [disabled]
1237
1238           The output is sensitive to the effects of previous command-line
1239           options, so for example it is possible to find out which
1240           optimizations are enabled at -O2 by using:
1241
1242                   -Q -O2 --help=optimizers
1243
1244           Alternatively you can discover which binary optimizations are
1245           enabled by -O3 by using:
1246
1247                   gcc -c -Q -O3 --help=optimizers > /tmp/O3-opts
1248                   gcc -c -Q -O2 --help=optimizers > /tmp/O2-opts
1249                   diff /tmp/O2-opts /tmp/O3-opts | grep enabled
1250
1251       --version
1252           Display the version number and copyrights of the invoked GCC.
1253
1254       -pass-exit-codes
1255           Normally the gcc program exits with the code of 1 if any phase of
1256           the compiler returns a non-success return code.  If you specify
1257           -pass-exit-codes, the gcc program instead returns with the
1258           numerically highest error produced by any phase returning an error
1259           indication.  The C, C++, and Fortran front ends return 4 if an
1260           internal compiler error is encountered.
1261
1262       -pipe
1263           Use pipes rather than temporary files for communication between the
1264           various stages of compilation.  This fails to work on some systems
1265           where the assembler is unable to read from a pipe; but the GNU
1266           assembler has no trouble.
1267
1268       -specs=file
1269           Process file after the compiler reads in the standard specs file,
1270           in order to override the defaults which the gcc driver program uses
1271           when determining what switches to pass to cc1, cc1plus, as, ld,
1272           etc.  More than one -specs=file can be specified on the command
1273           line, and they are processed in order, from left to right.
1274
1275       -wrapper
1276           Invoke all subcommands under a wrapper program.  The name of the
1277           wrapper program and its parameters are passed as a comma separated
1278           list.
1279
1280                   gcc -c t.c -wrapper gdb,--args
1281
1282           This invokes all subprograms of gcc under gdb --args, thus the
1283           invocation of cc1 is gdb --args cc1 ....
1284
1285       -ffile-prefix-map=old=new
1286           When compiling files residing in directory old, record any
1287           references to them in the result of the compilation as if the files
1288           resided in directory new instead.  Specifying this option is
1289           equivalent to specifying all the individual -f*-prefix-map options.
1290           This can be used to make reproducible builds that are location
1291           independent.  See also -fmacro-prefix-map and -fdebug-prefix-map.
1292
1293       -fplugin=name.so
1294           Load the plugin code in file name.so, assumed to be a shared object
1295           to be dlopen'd by the compiler.  The base name of the shared object
1296           file is used to identify the plugin for the purposes of argument
1297           parsing (See -fplugin-arg-name-key=value below).  Each plugin
1298           should define the callback functions specified in the Plugins API.
1299
1300       -fplugin-arg-name-key=value
1301           Define an argument called key with a value of value for the plugin
1302           called name.
1303
1304       -fdump-ada-spec[-slim]
1305           For C and C++ source and include files, generate corresponding Ada
1306           specs.
1307
1308       -fada-spec-parent=unit
1309           In conjunction with -fdump-ada-spec[-slim] above, generate Ada
1310           specs as child units of parent unit.
1311
1312       -fdump-go-spec=file
1313           For input files in any language, generate corresponding Go
1314           declarations in file.  This generates Go "const", "type", "var",
1315           and "func" declarations which may be a useful way to start writing
1316           a Go interface to code written in some other language.
1317
1318       @file
1319           Read command-line options from file.  The options read are inserted
1320           in place of the original @file option.  If file does not exist, or
1321           cannot be read, then the option will be treated literally, and not
1322           removed.
1323
1324           Options in file are separated by whitespace.  A whitespace
1325           character may be included in an option by surrounding the entire
1326           option in either single or double quotes.  Any character (including
1327           a backslash) may be included by prefixing the character to be
1328           included with a backslash.  The file may itself contain additional
1329           @file options; any such options will be processed recursively.
1330
1331   Compiling C++ Programs
1332       C++ source files conventionally use one of the suffixes .C, .cc, .cpp,
1333       .CPP, .c++, .cp, or .cxx; C++ header files often use .hh, .hpp, .H, or
1334       (for shared template code) .tcc; and preprocessed C++ files use the
1335       suffix .ii.  GCC recognizes files with these names and compiles them as
1336       C++ programs even if you call the compiler the same way as for
1337       compiling C programs (usually with the name gcc).
1338
1339       However, the use of gcc does not add the C++ library.  g++ is a program
1340       that calls GCC and automatically specifies linking against the C++
1341       library.  It treats .c, .h and .i files as C++ source files instead of
1342       C source files unless -x is used.  This program is also useful when
1343       precompiling a C header file with a .h extension for use in C++
1344       compilations.  On many systems, g++ is also installed with the name
1345       c++.
1346
1347       When you compile C++ programs, you may specify many of the same
1348       command-line options that you use for compiling programs in any
1349       language; or command-line options meaningful for C and related
1350       languages; or options that are meaningful only for C++ programs.
1351
1352   Options Controlling C Dialect
1353       The following options control the dialect of C (or languages derived
1354       from C, such as C++, Objective-C and Objective-C++) that the compiler
1355       accepts:
1356
1357       -ansi
1358           In C mode, this is equivalent to -std=c90. In C++ mode, it is
1359           equivalent to -std=c++98.
1360
1361           This turns off certain features of GCC that are incompatible with
1362           ISO C90 (when compiling C code), or of standard C++ (when compiling
1363           C++ code), such as the "asm" and "typeof" keywords, and predefined
1364           macros such as "unix" and "vax" that identify the type of system
1365           you are using.  It also enables the undesirable and rarely used ISO
1366           trigraph feature.  For the C compiler, it disables recognition of
1367           C++ style // comments as well as the "inline" keyword.
1368
1369           The alternate keywords "__asm__", "__extension__", "__inline__" and
1370           "__typeof__" continue to work despite -ansi.  You would not want to
1371           use them in an ISO C program, of course, but it is useful to put
1372           them in header files that might be included in compilations done
1373           with -ansi.  Alternate predefined macros such as "__unix__" and
1374           "__vax__" are also available, with or without -ansi.
1375
1376           The -ansi option does not cause non-ISO programs to be rejected
1377           gratuitously.  For that, -Wpedantic is required in addition to
1378           -ansi.
1379
1380           The macro "__STRICT_ANSI__" is predefined when the -ansi option is
1381           used.  Some header files may notice this macro and refrain from
1382           declaring certain functions or defining certain macros that the ISO
1383           standard doesn't call for; this is to avoid interfering with any
1384           programs that might use these names for other things.
1385
1386           Functions that are normally built in but do not have semantics
1387           defined by ISO C (such as "alloca" and "ffs") are not built-in
1388           functions when -ansi is used.
1389
1390       -std=
1391           Determine the language standard.   This option is currently only
1392           supported when compiling C or C++.
1393
1394           The compiler can accept several base standards, such as c90 or
1395           c++98, and GNU dialects of those standards, such as gnu90 or
1396           gnu++98.  When a base standard is specified, the compiler accepts
1397           all programs following that standard plus those using GNU
1398           extensions that do not contradict it.  For example, -std=c90 turns
1399           off certain features of GCC that are incompatible with ISO C90,
1400           such as the "asm" and "typeof" keywords, but not other GNU
1401           extensions that do not have a meaning in ISO C90, such as omitting
1402           the middle term of a "?:" expression. On the other hand, when a GNU
1403           dialect of a standard is specified, all features supported by the
1404           compiler are enabled, even when those features change the meaning
1405           of the base standard.  As a result, some strict-conforming programs
1406           may be rejected.  The particular standard is used by -Wpedantic to
1407           identify which features are GNU extensions given that version of
1408           the standard. For example -std=gnu90 -Wpedantic warns about C++
1409           style // comments, while -std=gnu99 -Wpedantic does not.
1410
1411           A value for this option must be provided; possible values are
1412
1413           c90
1414           c89
1415           iso9899:1990
1416               Support all ISO C90 programs (certain GNU extensions that
1417               conflict with ISO C90 are disabled). Same as -ansi for C code.
1418
1419           iso9899:199409
1420               ISO C90 as modified in amendment 1.
1421
1422           c99
1423           c9x
1424           iso9899:1999
1425           iso9899:199x
1426               ISO C99.  This standard is substantially completely supported,
1427               modulo bugs and floating-point issues (mainly but not entirely
1428               relating to optional C99 features from Annexes F and G).  See
1429               <http://gcc.gnu.org/c99status.html> for more information.  The
1430               names c9x and iso9899:199x are deprecated.
1431
1432           c11
1433           c1x
1434           iso9899:2011
1435               ISO C11, the 2011 revision of the ISO C standard.  This
1436               standard is substantially completely supported, modulo bugs,
1437               floating-point issues (mainly but not entirely relating to
1438               optional C11 features from Annexes F and G) and the optional
1439               Annexes K (Bounds-checking interfaces) and L (Analyzability).
1440               The name c1x is deprecated.
1441
1442           c17
1443           c18
1444           iso9899:2017
1445           iso9899:2018
1446               ISO C17, the 2017 revision of the ISO C standard (published in
1447               2018).  This standard is same as C11 except for corrections of
1448               defects (all of which are also applied with -std=c11) and a new
1449               value of "__STDC_VERSION__", and so is supported to the same
1450               extent as C11.
1451
1452           c2x The next version of the ISO C standard, still under
1453               development.  The support for this version is experimental and
1454               incomplete.
1455
1456           gnu90
1457           gnu89
1458               GNU dialect of ISO C90 (including some C99 features).
1459
1460           gnu99
1461           gnu9x
1462               GNU dialect of ISO C99.  The name gnu9x is deprecated.
1463
1464           gnu11
1465           gnu1x
1466               GNU dialect of ISO C11.  The name gnu1x is deprecated.
1467
1468           gnu17
1469           gnu18
1470               GNU dialect of ISO C17.  This is the default for C code.
1471
1472           gnu2x
1473               The next version of the ISO C standard, still under
1474               development, plus GNU extensions.  The support for this version
1475               is experimental and incomplete.
1476
1477           c++98
1478           c++03
1479               The 1998 ISO C++ standard plus the 2003 technical corrigendum
1480               and some additional defect reports. Same as -ansi for C++ code.
1481
1482           gnu++98
1483           gnu++03
1484               GNU dialect of -std=c++98.
1485
1486           c++11
1487           c++0x
1488               The 2011 ISO C++ standard plus amendments.  The name c++0x is
1489               deprecated.
1490
1491           gnu++11
1492           gnu++0x
1493               GNU dialect of -std=c++11.  The name gnu++0x is deprecated.
1494
1495           c++14
1496           c++1y
1497               The 2014 ISO C++ standard plus amendments.  The name c++1y is
1498               deprecated.
1499
1500           gnu++14
1501           gnu++1y
1502               GNU dialect of -std=c++14.  This is the default for C++ code.
1503               The name gnu++1y is deprecated.
1504
1505           c++17
1506           c++1z
1507               The 2017 ISO C++ standard plus amendments.  The name c++1z is
1508               deprecated.
1509
1510           gnu++17
1511           gnu++1z
1512               GNU dialect of -std=c++17.  The name gnu++1z is deprecated.
1513
1514           c++20
1515           c++2a
1516               The next revision of the ISO C++ standard, planned for 2020.
1517               Support is highly experimental, and will almost certainly
1518               change in incompatible ways in future releases.
1519
1520           gnu++20
1521           gnu++2a
1522               GNU dialect of -std=c++20.  Support is highly experimental, and
1523               will almost certainly change in incompatible ways in future
1524               releases.
1525
1526       -fgnu89-inline
1527           The option -fgnu89-inline tells GCC to use the traditional GNU
1528           semantics for "inline" functions when in C99 mode.
1529
1530           Using this option is roughly equivalent to adding the "gnu_inline"
1531           function attribute to all inline functions.
1532
1533           The option -fno-gnu89-inline explicitly tells GCC to use the C99
1534           semantics for "inline" when in C99 or gnu99 mode (i.e., it
1535           specifies the default behavior).  This option is not supported in
1536           -std=c90 or -std=gnu90 mode.
1537
1538           The preprocessor macros "__GNUC_GNU_INLINE__" and
1539           "__GNUC_STDC_INLINE__" may be used to check which semantics are in
1540           effect for "inline" functions.
1541
1542       -fpermitted-flt-eval-methods=style
1543           ISO/IEC TS 18661-3 defines new permissible values for
1544           "FLT_EVAL_METHOD" that indicate that operations and constants with
1545           a semantic type that is an interchange or extended format should be
1546           evaluated to the precision and range of that type.  These new
1547           values are a superset of those permitted under C99/C11, which does
1548           not specify the meaning of other positive values of
1549           "FLT_EVAL_METHOD".  As such, code conforming to C11 may not have
1550           been written expecting the possibility of the new values.
1551
1552           -fpermitted-flt-eval-methods specifies whether the compiler should
1553           allow only the values of "FLT_EVAL_METHOD" specified in C99/C11, or
1554           the extended set of values specified in ISO/IEC TS 18661-3.
1555
1556           style is either "c11" or "ts-18661-3" as appropriate.
1557
1558           The default when in a standards compliant mode (-std=c11 or
1559           similar) is -fpermitted-flt-eval-methods=c11.  The default when in
1560           a GNU dialect (-std=gnu11 or similar) is
1561           -fpermitted-flt-eval-methods=ts-18661-3.
1562
1563       -aux-info filename
1564           Output to the given filename prototyped declarations for all
1565           functions declared and/or defined in a translation unit, including
1566           those in header files.  This option is silently ignored in any
1567           language other than C.
1568
1569           Besides declarations, the file indicates, in comments, the origin
1570           of each declaration (source file and line), whether the declaration
1571           was implicit, prototyped or unprototyped (I, N for new or O for
1572           old, respectively, in the first character after the line number and
1573           the colon), and whether it came from a declaration or a definition
1574           (C or F, respectively, in the following character).  In the case of
1575           function definitions, a K&R-style list of arguments followed by
1576           their declarations is also provided, inside comments, after the
1577           declaration.
1578
1579       -fallow-parameterless-variadic-functions
1580           Accept variadic functions without named parameters.
1581
1582           Although it is possible to define such a function, this is not very
1583           useful as it is not possible to read the arguments.  This is only
1584           supported for C as this construct is allowed by C++.
1585
1586       -fno-asm
1587           Do not recognize "asm", "inline" or "typeof" as a keyword, so that
1588           code can use these words as identifiers.  You can use the keywords
1589           "__asm__", "__inline__" and "__typeof__" instead.  -ansi implies
1590           -fno-asm.
1591
1592           In C++, this switch only affects the "typeof" keyword, since "asm"
1593           and "inline" are standard keywords.  You may want to use the
1594           -fno-gnu-keywords flag instead, which has the same effect.  In C99
1595           mode (-std=c99 or -std=gnu99), this switch only affects the "asm"
1596           and "typeof" keywords, since "inline" is a standard keyword in ISO
1597           C99.
1598
1599       -fno-builtin
1600       -fno-builtin-function
1601           Don't recognize built-in functions that do not begin with
1602           __builtin_ as prefix.
1603
1604           GCC normally generates special code to handle certain built-in
1605           functions more efficiently; for instance, calls to "alloca" may
1606           become single instructions which adjust the stack directly, and
1607           calls to "memcpy" may become inline copy loops.  The resulting code
1608           is often both smaller and faster, but since the function calls no
1609           longer appear as such, you cannot set a breakpoint on those calls,
1610           nor can you change the behavior of the functions by linking with a
1611           different library.  In addition, when a function is recognized as a
1612           built-in function, GCC may use information about that function to
1613           warn about problems with calls to that function, or to generate
1614           more efficient code, even if the resulting code still contains
1615           calls to that function.  For example, warnings are given with
1616           -Wformat for bad calls to "printf" when "printf" is built in and
1617           "strlen" is known not to modify global memory.
1618
1619           With the -fno-builtin-function option only the built-in function
1620           function is disabled.  function must not begin with __builtin_.  If
1621           a function is named that is not built-in in this version of GCC,
1622           this option is ignored.  There is no corresponding
1623           -fbuiltin-function option; if you wish to enable built-in functions
1624           selectively when using -fno-builtin or -ffreestanding, you may
1625           define macros such as:
1626
1627                   #define abs(n)          __builtin_abs ((n))
1628                   #define strcpy(d, s)    __builtin_strcpy ((d), (s))
1629
1630       -fgimple
1631           Enable parsing of function definitions marked with "__GIMPLE".
1632           This is an experimental feature that allows unit testing of GIMPLE
1633           passes.
1634
1635       -fhosted
1636           Assert that compilation targets a hosted environment.  This implies
1637           -fbuiltin.  A hosted environment is one in which the entire
1638           standard library is available, and in which "main" has a return
1639           type of "int".  Examples are nearly everything except a kernel.
1640           This is equivalent to -fno-freestanding.
1641
1642       -ffreestanding
1643           Assert that compilation targets a freestanding environment.  This
1644           implies -fno-builtin.  A freestanding environment is one in which
1645           the standard library may not exist, and program startup may not
1646           necessarily be at "main".  The most obvious example is an OS
1647           kernel.  This is equivalent to -fno-hosted.
1648
1649       -fopenacc
1650           Enable handling of OpenACC directives "#pragma acc" in C/C++ and
1651           "!$acc" in Fortran.  When -fopenacc is specified, the compiler
1652           generates accelerated code according to the OpenACC Application
1653           Programming Interface v2.6 <https://www.openacc.org>.  This option
1654           implies -pthread, and thus is only supported on targets that have
1655           support for -pthread.
1656
1657       -fopenacc-dim=geom
1658           Specify default compute dimensions for parallel offload regions
1659           that do not explicitly specify.  The geom value is a triple of
1660           ':'-separated sizes, in order 'gang', 'worker' and, 'vector'.  A
1661           size can be omitted, to use a target-specific default value.
1662
1663       -fopenmp
1664           Enable handling of OpenMP directives "#pragma omp" in C/C++ and
1665           "!$omp" in Fortran.  When -fopenmp is specified, the compiler
1666           generates parallel code according to the OpenMP Application Program
1667           Interface v4.5 <https://www.openmp.org>.  This option implies
1668           -pthread, and thus is only supported on targets that have support
1669           for -pthread. -fopenmp implies -fopenmp-simd.
1670
1671       -fopenmp-simd
1672           Enable handling of OpenMP's SIMD directives with "#pragma omp" in
1673           C/C++ and "!$omp" in Fortran. Other OpenMP directives are ignored.
1674
1675       -fgnu-tm
1676           When the option -fgnu-tm is specified, the compiler generates code
1677           for the Linux variant of Intel's current Transactional Memory ABI
1678           specification document (Revision 1.1, May 6 2009).  This is an
1679           experimental feature whose interface may change in future versions
1680           of GCC, as the official specification changes.  Please note that
1681           not all architectures are supported for this feature.
1682
1683           For more information on GCC's support for transactional memory,
1684
1685           Note that the transactional memory feature is not supported with
1686           non-call exceptions (-fnon-call-exceptions).
1687
1688       -fms-extensions
1689           Accept some non-standard constructs used in Microsoft header files.
1690
1691           In C++ code, this allows member names in structures to be similar
1692           to previous types declarations.
1693
1694                   typedef int UOW;
1695                   struct ABC {
1696                     UOW UOW;
1697                   };
1698
1699           Some cases of unnamed fields in structures and unions are only
1700           accepted with this option.
1701
1702           Note that this option is off for all targets except for x86 targets
1703           using ms-abi.
1704
1705       -fplan9-extensions
1706           Accept some non-standard constructs used in Plan 9 code.
1707
1708           This enables -fms-extensions, permits passing pointers to
1709           structures with anonymous fields to functions that expect pointers
1710           to elements of the type of the field, and permits referring to
1711           anonymous fields declared using a typedef.    This is only
1712           supported for C, not C++.
1713
1714       -fcond-mismatch
1715           Allow conditional expressions with mismatched types in the second
1716           and third arguments.  The value of such an expression is void.
1717           This option is not supported for C++.
1718
1719       -flax-vector-conversions
1720           Allow implicit conversions between vectors with differing numbers
1721           of elements and/or incompatible element types.  This option should
1722           not be used for new code.
1723
1724       -funsigned-char
1725           Let the type "char" be unsigned, like "unsigned char".
1726
1727           Each kind of machine has a default for what "char" should be.  It
1728           is either like "unsigned char" by default or like "signed char" by
1729           default.
1730
1731           Ideally, a portable program should always use "signed char" or
1732           "unsigned char" when it depends on the signedness of an object.
1733           But many programs have been written to use plain "char" and expect
1734           it to be signed, or expect it to be unsigned, depending on the
1735           machines they were written for.  This option, and its inverse, let
1736           you make such a program work with the opposite default.
1737
1738           The type "char" is always a distinct type from each of "signed
1739           char" or "unsigned char", even though its behavior is always just
1740           like one of those two.
1741
1742       -fsigned-char
1743           Let the type "char" be signed, like "signed char".
1744
1745           Note that this is equivalent to -fno-unsigned-char, which is the
1746           negative form of -funsigned-char.  Likewise, the option
1747           -fno-signed-char is equivalent to -funsigned-char.
1748
1749       -fsigned-bitfields
1750       -funsigned-bitfields
1751       -fno-signed-bitfields
1752       -fno-unsigned-bitfields
1753           These options control whether a bit-field is signed or unsigned,
1754           when the declaration does not use either "signed" or "unsigned".
1755           By default, such a bit-field is signed, because this is consistent:
1756           the basic integer types such as "int" are signed types.
1757
1758       -fsso-struct=endianness
1759           Set the default scalar storage order of structures and unions to
1760           the specified endianness.  The accepted values are big-endian,
1761           little-endian and native for the native endianness of the target
1762           (the default).  This option is not supported for C++.
1763
1764           Warning: the -fsso-struct switch causes GCC to generate code that
1765           is not binary compatible with code generated without it if the
1766           specified endianness is not the native endianness of the target.
1767
1768   Options Controlling C++ Dialect
1769       This section describes the command-line options that are only
1770       meaningful for C++ programs.  You can also use most of the GNU compiler
1771       options regardless of what language your program is in.  For example,
1772       you might compile a file firstClass.C like this:
1773
1774               g++ -g -fstrict-enums -O -c firstClass.C
1775
1776       In this example, only -fstrict-enums is an option meant only for C++
1777       programs; you can use the other options with any language supported by
1778       GCC.
1779
1780       Some options for compiling C programs, such as -std, are also relevant
1781       for C++ programs.
1782
1783       Here is a list of options that are only for compiling C++ programs:
1784
1785       -fabi-version=n
1786           Use version n of the C++ ABI.  The default is version 0.
1787
1788           Version 0 refers to the version conforming most closely to the C++
1789           ABI specification.  Therefore, the ABI obtained using version 0
1790           will change in different versions of G++ as ABI bugs are fixed.
1791
1792           Version 1 is the version of the C++ ABI that first appeared in G++
1793           3.2.
1794
1795           Version 2 is the version of the C++ ABI that first appeared in G++
1796           3.4, and was the default through G++ 4.9.
1797
1798           Version 3 corrects an error in mangling a constant address as a
1799           template argument.
1800
1801           Version 4, which first appeared in G++ 4.5, implements a standard
1802           mangling for vector types.
1803
1804           Version 5, which first appeared in G++ 4.6, corrects the mangling
1805           of attribute const/volatile on function pointer types, decltype of
1806           a plain decl, and use of a function parameter in the declaration of
1807           another parameter.
1808
1809           Version 6, which first appeared in G++ 4.7, corrects the promotion
1810           behavior of C++11 scoped enums and the mangling of template
1811           argument packs, const/static_cast, prefix ++ and --, and a class
1812           scope function used as a template argument.
1813
1814           Version 7, which first appeared in G++ 4.8, that treats nullptr_t
1815           as a builtin type and corrects the mangling of lambdas in default
1816           argument scope.
1817
1818           Version 8, which first appeared in G++ 4.9, corrects the
1819           substitution behavior of function types with function-cv-
1820           qualifiers.
1821
1822           Version 9, which first appeared in G++ 5.2, corrects the alignment
1823           of "nullptr_t".
1824
1825           Version 10, which first appeared in G++ 6.1, adds mangling of
1826           attributes that affect type identity, such as ia32 calling
1827           convention attributes (e.g. stdcall).
1828
1829           Version 11, which first appeared in G++ 7, corrects the mangling of
1830           sizeof... expressions and operator names.  For multiple entities
1831           with the same name within a function, that are declared in
1832           different scopes, the mangling now changes starting with the
1833           twelfth occurrence.  It also implies -fnew-inheriting-ctors.
1834
1835           Version 12, which first appeared in G++ 8, corrects the calling
1836           conventions for empty classes on the x86_64 target and for classes
1837           with only deleted copy/move constructors.  It accidentally changes
1838           the calling convention for classes with a deleted copy constructor
1839           and a trivial move constructor.
1840
1841           Version 13, which first appeared in G++ 8.2, fixes the accidental
1842           change in version 12.
1843
1844           Version 14, which first appeared in G++ 10, corrects the mangling
1845           of the nullptr expression.
1846
1847           See also -Wabi.
1848
1849       -fabi-compat-version=n
1850           On targets that support strong aliases, G++ works around mangling
1851           changes by creating an alias with the correct mangled name when
1852           defining a symbol with an incorrect mangled name.  This switch
1853           specifies which ABI version to use for the alias.
1854
1855           With -fabi-version=0 (the default), this defaults to 11 (GCC 7
1856           compatibility).  If another ABI version is explicitly selected,
1857           this defaults to 0.  For compatibility with GCC versions 3.2
1858           through 4.9, use -fabi-compat-version=2.
1859
1860           If this option is not provided but -Wabi=n is, that version is used
1861           for compatibility aliases.  If this option is provided along with
1862           -Wabi (without the version), the version from this option is used
1863           for the warning.
1864
1865       -fno-access-control
1866           Turn off all access checking.  This switch is mainly useful for
1867           working around bugs in the access control code.
1868
1869       -faligned-new
1870           Enable support for C++17 "new" of types that require more alignment
1871           than "void* ::operator new(std::size_t)" provides.  A numeric
1872           argument such as "-faligned-new=32" can be used to specify how much
1873           alignment (in bytes) is provided by that function, but few users
1874           will need to override the default of "alignof(std::max_align_t)".
1875
1876           This flag is enabled by default for -std=c++17.
1877
1878       -fchar8_t
1879       -fno-char8_t
1880           Enable support for "char8_t" as adopted for C++2a.  This includes
1881           the addition of a new "char8_t" fundamental type, changes to the
1882           types of UTF-8 string and character literals, new signatures for
1883           user-defined literals, associated standard library updates, and new
1884           "__cpp_char8_t" and "__cpp_lib_char8_t" feature test macros.
1885
1886           This option enables functions to be overloaded for ordinary and
1887           UTF-8 strings:
1888
1889                   int f(const char *);    // #1
1890                   int f(const char8_t *); // #2
1891                   int v1 = f("text");     // Calls #1
1892                   int v2 = f(u8"text");   // Calls #2
1893
1894           and introduces new signatures for user-defined literals:
1895
1896                   int operator""_udl1(char8_t);
1897                   int v3 = u8'x'_udl1;
1898                   int operator""_udl2(const char8_t*, std::size_t);
1899                   int v4 = u8"text"_udl2;
1900                   template<typename T, T...> int operator""_udl3();
1901                   int v5 = u8"text"_udl3;
1902
1903           The change to the types of UTF-8 string and character literals
1904           introduces incompatibilities with ISO C++11 and later standards.
1905           For example, the following code is well-formed under ISO C++11, but
1906           is ill-formed when -fchar8_t is specified.
1907
1908                   char ca[] = u8"xx";     // error: char-array initialized from wide
1909                                           //        string
1910                   const char *cp = u8"xx";// error: invalid conversion from
1911                                           //        `const char8_t*' to `const char*'
1912                   int f(const char*);
1913                   auto v = f(u8"xx");     // error: invalid conversion from
1914                                           //        `const char8_t*' to `const char*'
1915                   std::string s{u8"xx"};  // error: no matching function for call to
1916                                           //        `std::basic_string<char>::basic_string()'
1917                   using namespace std::literals;
1918                   s = u8"xx"s;            // error: conversion from
1919                                           //        `basic_string<char8_t>' to non-scalar
1920                                           //        type `basic_string<char>' requested
1921
1922       -fcheck-new
1923           Check that the pointer returned by "operator new" is non-null
1924           before attempting to modify the storage allocated.  This check is
1925           normally unnecessary because the C++ standard specifies that
1926           "operator new" only returns 0 if it is declared "throw()", in which
1927           case the compiler always checks the return value even without this
1928           option.  In all other cases, when "operator new" has a non-empty
1929           exception specification, memory exhaustion is signalled by throwing
1930           "std::bad_alloc".  See also new (nothrow).
1931
1932       -fconcepts
1933       -fconcepts-ts
1934           Below -std=c++2a, -fconcepts enables support for the C++ Extensions
1935           for Concepts Technical Specification, ISO 19217 (2015).
1936
1937           With -std=c++2a and above, Concepts are part of the language
1938           standard, so -fconcepts defaults to on.  But the standard
1939           specification of Concepts differs significantly from the TS, so
1940           some constructs that were allowed in the TS but didn't make it into
1941           the standard can still be enabled by -fconcepts-ts.
1942
1943       -fconstexpr-depth=n
1944           Set the maximum nested evaluation depth for C++11 constexpr
1945           functions to n.  A limit is needed to detect endless recursion
1946           during constant expression evaluation.  The minimum specified by
1947           the standard is 512.
1948
1949       -fconstexpr-cache-depth=n
1950           Set the maximum level of nested evaluation depth for C++11
1951           constexpr functions that will be cached to n.  This is a heuristic
1952           that trades off compilation speed (when the cache avoids repeated
1953           calculations) against memory consumption (when the cache grows very
1954           large from highly recursive evaluations).  The default is 8.  Very
1955           few users are likely to want to adjust it, but if your code does
1956           heavy constexpr calculations you might want to experiment to find
1957           which value works best for you.
1958
1959       -fconstexpr-loop-limit=n
1960           Set the maximum number of iterations for a loop in C++14 constexpr
1961           functions to n.  A limit is needed to detect infinite loops during
1962           constant expression evaluation.  The default is 262144 (1<<18).
1963
1964       -fconstexpr-ops-limit=n
1965           Set the maximum number of operations during a single constexpr
1966           evaluation.  Even when number of iterations of a single loop is
1967           limited with the above limit, if there are several nested loops and
1968           each of them has many iterations but still smaller than the above
1969           limit, or if in a body of some loop or even outside of a loop too
1970           many expressions need to be evaluated, the resulting constexpr
1971           evaluation might take too long.  The default is 33554432 (1<<25).
1972
1973       -fcoroutines
1974           Enable support for the C++ coroutines extension (experimental).
1975
1976       -fno-elide-constructors
1977           The C++ standard allows an implementation to omit creating a
1978           temporary that is only used to initialize another object of the
1979           same type.  Specifying this option disables that optimization, and
1980           forces G++ to call the copy constructor in all cases.  This option
1981           also causes G++ to call trivial member functions which otherwise
1982           would be expanded inline.
1983
1984           In C++17, the compiler is required to omit these temporaries, but
1985           this option still affects trivial member functions.
1986
1987       -fno-enforce-eh-specs
1988           Don't generate code to check for violation of exception
1989           specifications at run time.  This option violates the C++ standard,
1990           but may be useful for reducing code size in production builds, much
1991           like defining "NDEBUG".  This does not give user code permission to
1992           throw exceptions in violation of the exception specifications; the
1993           compiler still optimizes based on the specifications, so throwing
1994           an unexpected exception results in undefined behavior at run time.
1995
1996       -fextern-tls-init
1997       -fno-extern-tls-init
1998           The C++11 and OpenMP standards allow "thread_local" and
1999           "threadprivate" variables to have dynamic (runtime) initialization.
2000           To support this, any use of such a variable goes through a wrapper
2001           function that performs any necessary initialization.  When the use
2002           and definition of the variable are in the same translation unit,
2003           this overhead can be optimized away, but when the use is in a
2004           different translation unit there is significant overhead even if
2005           the variable doesn't actually need dynamic initialization.  If the
2006           programmer can be sure that no use of the variable in a non-
2007           defining TU needs to trigger dynamic initialization (either because
2008           the variable is statically initialized, or a use of the variable in
2009           the defining TU will be executed before any uses in another TU),
2010           they can avoid this overhead with the -fno-extern-tls-init option.
2011
2012           On targets that support symbol aliases, the default is
2013           -fextern-tls-init.  On targets that do not support symbol aliases,
2014           the default is -fno-extern-tls-init.
2015
2016       -fno-gnu-keywords
2017           Do not recognize "typeof" as a keyword, so that code can use this
2018           word as an identifier.  You can use the keyword "__typeof__"
2019           instead.  This option is implied by the strict ISO C++ dialects:
2020           -ansi, -std=c++98, -std=c++11, etc.
2021
2022       -fno-implicit-templates
2023           Never emit code for non-inline templates that are instantiated
2024           implicitly (i.e. by use); only emit code for explicit
2025           instantiations.  If you use this option, you must take care to
2026           structure your code to include all the necessary explicit
2027           instantiations to avoid getting undefined symbols at link time.
2028
2029       -fno-implicit-inline-templates
2030           Don't emit code for implicit instantiations of inline templates,
2031           either.  The default is to handle inlines differently so that
2032           compiles with and without optimization need the same set of
2033           explicit instantiations.
2034
2035       -fno-implement-inlines
2036           To save space, do not emit out-of-line copies of inline functions
2037           controlled by "#pragma implementation".  This causes linker errors
2038           if these functions are not inlined everywhere they are called.
2039
2040       -fms-extensions
2041           Disable Wpedantic warnings about constructs used in MFC, such as
2042           implicit int and getting a pointer to member function via non-
2043           standard syntax.
2044
2045       -fnew-inheriting-ctors
2046           Enable the P0136 adjustment to the semantics of C++11 constructor
2047           inheritance.  This is part of C++17 but also considered to be a
2048           Defect Report against C++11 and C++14.  This flag is enabled by
2049           default unless -fabi-version=10 or lower is specified.
2050
2051       -fnew-ttp-matching
2052           Enable the P0522 resolution to Core issue 150, template template
2053           parameters and default arguments: this allows a template with
2054           default template arguments as an argument for a template template
2055           parameter with fewer template parameters.  This flag is enabled by
2056           default for -std=c++17.
2057
2058       -fno-nonansi-builtins
2059           Disable built-in declarations of functions that are not mandated by
2060           ANSI/ISO C.  These include "ffs", "alloca", "_exit", "index",
2061           "bzero", "conjf", and other related functions.
2062
2063       -fnothrow-opt
2064           Treat a "throw()" exception specification as if it were a
2065           "noexcept" specification to reduce or eliminate the text size
2066           overhead relative to a function with no exception specification.
2067           If the function has local variables of types with non-trivial
2068           destructors, the exception specification actually makes the
2069           function smaller because the EH cleanups for those variables can be
2070           optimized away.  The semantic effect is that an exception thrown
2071           out of a function with such an exception specification results in a
2072           call to "terminate" rather than "unexpected".
2073
2074       -fno-operator-names
2075           Do not treat the operator name keywords "and", "bitand", "bitor",
2076           "compl", "not", "or" and "xor" as synonyms as keywords.
2077
2078       -fno-optional-diags
2079           Disable diagnostics that the standard says a compiler does not need
2080           to issue.  Currently, the only such diagnostic issued by G++ is the
2081           one for a name having multiple meanings within a class.
2082
2083       -fpermissive
2084           Downgrade some diagnostics about nonconformant code from errors to
2085           warnings.  Thus, using -fpermissive allows some nonconforming code
2086           to compile.
2087
2088       -fno-pretty-templates
2089           When an error message refers to a specialization of a function
2090           template, the compiler normally prints the signature of the
2091           template followed by the template arguments and any typedefs or
2092           typenames in the signature (e.g. "void f(T) [with T = int]" rather
2093           than "void f(int)") so that it's clear which template is involved.
2094           When an error message refers to a specialization of a class
2095           template, the compiler omits any template arguments that match the
2096           default template arguments for that template.  If either of these
2097           behaviors make it harder to understand the error message rather
2098           than easier, you can use -fno-pretty-templates to disable them.
2099
2100       -fno-rtti
2101           Disable generation of information about every class with virtual
2102           functions for use by the C++ run-time type identification features
2103           ("dynamic_cast" and "typeid").  If you don't use those parts of the
2104           language, you can save some space by using this flag.  Note that
2105           exception handling uses the same information, but G++ generates it
2106           as needed. The "dynamic_cast" operator can still be used for casts
2107           that do not require run-time type information, i.e. casts to "void
2108           *" or to unambiguous base classes.
2109
2110           Mixing code compiled with -frtti with that compiled with -fno-rtti
2111           may not work.  For example, programs may fail to link if a class
2112           compiled with -fno-rtti is used as a base for a class compiled with
2113           -frtti.
2114
2115       -fsized-deallocation
2116           Enable the built-in global declarations
2117
2118                   void operator delete (void *, std::size_t) noexcept;
2119                   void operator delete[] (void *, std::size_t) noexcept;
2120
2121           as introduced in C++14.  This is useful for user-defined
2122           replacement deallocation functions that, for example, use the size
2123           of the object to make deallocation faster.  Enabled by default
2124           under -std=c++14 and above.  The flag -Wsized-deallocation warns
2125           about places that might want to add a definition.
2126
2127       -fstrict-enums
2128           Allow the compiler to optimize using the assumption that a value of
2129           enumerated type can only be one of the values of the enumeration
2130           (as defined in the C++ standard; basically, a value that can be
2131           represented in the minimum number of bits needed to represent all
2132           the enumerators).  This assumption may not be valid if the program
2133           uses a cast to convert an arbitrary integer value to the enumerated
2134           type.
2135
2136       -fstrong-eval-order
2137           Evaluate member access, array subscripting, and shift expressions
2138           in left-to-right order, and evaluate assignment in right-to-left
2139           order, as adopted for C++17.  Enabled by default with -std=c++17.
2140           -fstrong-eval-order=some enables just the ordering of member access
2141           and shift expressions, and is the default without -std=c++17.
2142
2143       -ftemplate-backtrace-limit=n
2144           Set the maximum number of template instantiation notes for a single
2145           warning or error to n.  The default value is 10.
2146
2147       -ftemplate-depth=n
2148           Set the maximum instantiation depth for template classes to n.  A
2149           limit on the template instantiation depth is needed to detect
2150           endless recursions during template class instantiation.  ANSI/ISO
2151           C++ conforming programs must not rely on a maximum depth greater
2152           than 17 (changed to 1024 in C++11).  The default value is 900, as
2153           the compiler can run out of stack space before hitting 1024 in some
2154           situations.
2155
2156       -fno-threadsafe-statics
2157           Do not emit the extra code to use the routines specified in the C++
2158           ABI for thread-safe initialization of local statics.  You can use
2159           this option to reduce code size slightly in code that doesn't need
2160           to be thread-safe.
2161
2162       -fuse-cxa-atexit
2163           Register destructors for objects with static storage duration with
2164           the "__cxa_atexit" function rather than the "atexit" function.
2165           This option is required for fully standards-compliant handling of
2166           static destructors, but only works if your C library supports
2167           "__cxa_atexit".
2168
2169       -fno-use-cxa-get-exception-ptr
2170           Don't use the "__cxa_get_exception_ptr" runtime routine.  This
2171           causes "std::uncaught_exception" to be incorrect, but is necessary
2172           if the runtime routine is not available.
2173
2174       -fvisibility-inlines-hidden
2175           This switch declares that the user does not attempt to compare
2176           pointers to inline functions or methods where the addresses of the
2177           two functions are taken in different shared objects.
2178
2179           The effect of this is that GCC may, effectively, mark inline
2180           methods with "__attribute__ ((visibility ("hidden")))" so that they
2181           do not appear in the export table of a DSO and do not require a PLT
2182           indirection when used within the DSO.  Enabling this option can
2183           have a dramatic effect on load and link times of a DSO as it
2184           massively reduces the size of the dynamic export table when the
2185           library makes heavy use of templates.
2186
2187           The behavior of this switch is not quite the same as marking the
2188           methods as hidden directly, because it does not affect static
2189           variables local to the function or cause the compiler to deduce
2190           that the function is defined in only one shared object.
2191
2192           You may mark a method as having a visibility explicitly to negate
2193           the effect of the switch for that method.  For example, if you do
2194           want to compare pointers to a particular inline method, you might
2195           mark it as having default visibility.  Marking the enclosing class
2196           with explicit visibility has no effect.
2197
2198           Explicitly instantiated inline methods are unaffected by this
2199           option as their linkage might otherwise cross a shared library
2200           boundary.
2201
2202       -fvisibility-ms-compat
2203           This flag attempts to use visibility settings to make GCC's C++
2204           linkage model compatible with that of Microsoft Visual Studio.
2205
2206           The flag makes these changes to GCC's linkage model:
2207
2208           1.  It sets the default visibility to "hidden", like
2209               -fvisibility=hidden.
2210
2211           2.  Types, but not their members, are not hidden by default.
2212
2213           3.  The One Definition Rule is relaxed for types without explicit
2214               visibility specifications that are defined in more than one
2215               shared object: those declarations are permitted if they are
2216               permitted when this option is not used.
2217
2218           In new code it is better to use -fvisibility=hidden and export
2219           those classes that are intended to be externally visible.
2220           Unfortunately it is possible for code to rely, perhaps
2221           accidentally, on the Visual Studio behavior.
2222
2223           Among the consequences of these changes are that static data
2224           members of the same type with the same name but defined in
2225           different shared objects are different, so changing one does not
2226           change the other; and that pointers to function members defined in
2227           different shared objects may not compare equal.  When this flag is
2228           given, it is a violation of the ODR to define types with the same
2229           name differently.
2230
2231       -fno-weak
2232           Do not use weak symbol support, even if it is provided by the
2233           linker.  By default, G++ uses weak symbols if they are available.
2234           This option exists only for testing, and should not be used by end-
2235           users; it results in inferior code and has no benefits.  This
2236           option may be removed in a future release of G++.
2237
2238       -fext-numeric-literals (C++ and Objective-C++ only)
2239           Accept imaginary, fixed-point, or machine-defined literal number
2240           suffixes as GNU extensions.  When this option is turned off these
2241           suffixes are treated as C++11 user-defined literal numeric
2242           suffixes.  This is on by default for all pre-C++11 dialects and all
2243           GNU dialects: -std=c++98, -std=gnu++98, -std=gnu++11, -std=gnu++14.
2244           This option is off by default for ISO C++11 onwards (-std=c++11,
2245           ...).
2246
2247       -nostdinc++
2248           Do not search for header files in the standard directories specific
2249           to C++, but do still search the other standard directories.  (This
2250           option is used when building the C++ library.)
2251
2252       In addition, these warning options have meanings only for C++ programs:
2253
2254       -Wabi-tag (C++ and Objective-C++ only)
2255           Warn when a type with an ABI tag is used in a context that does not
2256           have that ABI tag.  See C++ Attributes for more information about
2257           ABI tags.
2258
2259       -Wcomma-subscript (C++ and Objective-C++ only)
2260           Warn about uses of a comma expression within a subscripting
2261           expression.  This usage was deprecated in C++2a.  However, a comma
2262           expression wrapped in "( )" is not deprecated.  Example:
2263
2264                   void f(int *a, int b, int c) {
2265                       a[b,c];     // deprecated
2266                       a[(b,c)];   // OK
2267                   }
2268
2269           Enabled by default with -std=c++2a.
2270
2271       -Wctor-dtor-privacy (C++ and Objective-C++ only)
2272           Warn when a class seems unusable because all the constructors or
2273           destructors in that class are private, and it has neither friends
2274           nor public static member functions.  Also warn if there are no non-
2275           private methods, and there's at least one private member function
2276           that isn't a constructor or destructor.
2277
2278       -Wdelete-non-virtual-dtor (C++ and Objective-C++ only)
2279           Warn when "delete" is used to destroy an instance of a class that
2280           has virtual functions and non-virtual destructor. It is unsafe to
2281           delete an instance of a derived class through a pointer to a base
2282           class if the base class does not have a virtual destructor.  This
2283           warning is enabled by -Wall.
2284
2285       -Wdeprecated-copy (C++ and Objective-C++ only)
2286           Warn that the implicit declaration of a copy constructor or copy
2287           assignment operator is deprecated if the class has a user-provided
2288           copy constructor or copy assignment operator, in C++11 and up.
2289           This warning is enabled by -Wextra.  With -Wdeprecated-copy-dtor,
2290           also deprecate if the class has a user-provided destructor.
2291
2292       -Wno-init-list-lifetime (C++ and Objective-C++ only)
2293           Do not warn about uses of "std::initializer_list" that are likely
2294           to result in dangling pointers.  Since the underlying array for an
2295           "initializer_list" is handled like a normal C++ temporary object,
2296           it is easy to inadvertently keep a pointer to the array past the
2297           end of the array's lifetime.  For example:
2298
2299           *   If a function returns a temporary "initializer_list", or a
2300               local "initializer_list" variable, the array's lifetime ends at
2301               the end of the return statement, so the value returned has a
2302               dangling pointer.
2303
2304           *   If a new-expression creates an "initializer_list", the array
2305               only lives until the end of the enclosing full-expression, so
2306               the "initializer_list" in the heap has a dangling pointer.
2307
2308           *   When an "initializer_list" variable is assigned from a brace-
2309               enclosed initializer list, the temporary array created for the
2310               right side of the assignment only lives until the end of the
2311               full-expression, so at the next statement the
2312               "initializer_list" variable has a dangling pointer.
2313
2314                       // li's initial underlying array lives as long as li
2315                       std::initializer_list<int> li = { 1,2,3 };
2316                       // assignment changes li to point to a temporary array
2317                       li = { 4, 5 };
2318                       // now the temporary is gone and li has a dangling pointer
2319                       int i = li.begin()[0] // undefined behavior
2320
2321           *   When a list constructor stores the "begin" pointer from the
2322               "initializer_list" argument, this doesn't extend the lifetime
2323               of the array, so if a class variable is constructed from a
2324               temporary "initializer_list", the pointer is left dangling by
2325               the end of the variable declaration statement.
2326
2327       -Wno-literal-suffix (C++ and Objective-C++ only)
2328           Do not warn when a string or character literal is followed by a ud-
2329           suffix which does not begin with an underscore.  As a conforming
2330           extension, GCC treats such suffixes as separate preprocessing
2331           tokens in order to maintain backwards compatibility with code that
2332           uses formatting macros from "<inttypes.h>".  For example:
2333
2334                   #define __STDC_FORMAT_MACROS
2335                   #include <inttypes.h>
2336                   #include <stdio.h>
2337
2338                   int main() {
2339                     int64_t i64 = 123;
2340                     printf("My int64: %" PRId64"\n", i64);
2341                   }
2342
2343           In this case, "PRId64" is treated as a separate preprocessing
2344           token.
2345
2346           This option also controls warnings when a user-defined literal
2347           operator is declared with a literal suffix identifier that doesn't
2348           begin with an underscore. Literal suffix identifiers that don't
2349           begin with an underscore are reserved for future standardization.
2350
2351           These warnings are enabled by default.
2352
2353       -Wno-narrowing (C++ and Objective-C++ only)
2354           For C++11 and later standards, narrowing conversions are diagnosed
2355           by default, as required by the standard.  A narrowing conversion
2356           from a constant produces an error, and a narrowing conversion from
2357           a non-constant produces a warning, but -Wno-narrowing suppresses
2358           the diagnostic.  Note that this does not affect the meaning of
2359           well-formed code; narrowing conversions are still considered ill-
2360           formed in SFINAE contexts.
2361
2362           With -Wnarrowing in C++98, warn when a narrowing conversion
2363           prohibited by C++11 occurs within { }, e.g.
2364
2365                   int i = { 2.2 }; // error: narrowing from double to int
2366
2367           This flag is included in -Wall and -Wc++11-compat.
2368
2369       -Wnoexcept (C++ and Objective-C++ only)
2370           Warn when a noexcept-expression evaluates to false because of a
2371           call to a function that does not have a non-throwing exception
2372           specification (i.e. "throw()" or "noexcept") but is known by the
2373           compiler to never throw an exception.
2374
2375       -Wnoexcept-type (C++ and Objective-C++ only)
2376           Warn if the C++17 feature making "noexcept" part of a function type
2377           changes the mangled name of a symbol relative to C++14.  Enabled by
2378           -Wabi and -Wc++17-compat.
2379
2380           As an example:
2381
2382                   template <class T> void f(T t) { t(); };
2383                   void g() noexcept;
2384                   void h() { f(g); }
2385
2386           In C++14, "f" calls "f<void(*)()>", but in C++17 it calls
2387           "f<void(*)()noexcept>".
2388
2389       -Wclass-memaccess (C++ and Objective-C++ only)
2390           Warn when the destination of a call to a raw memory function such
2391           as "memset" or "memcpy" is an object of class type, and when
2392           writing into such an object might bypass the class non-trivial or
2393           deleted constructor or copy assignment, violate const-correctness
2394           or encapsulation, or corrupt virtual table pointers.  Modifying the
2395           representation of such objects may violate invariants maintained by
2396           member functions of the class.  For example, the call to "memset"
2397           below is undefined because it modifies a non-trivial class object
2398           and is, therefore, diagnosed.  The safe way to either initialize or
2399           clear the storage of objects of such types is by using the
2400           appropriate constructor or assignment operator, if one is
2401           available.
2402
2403                   std::string str = "abc";
2404                   memset (&str, 0, sizeof str);
2405
2406           The -Wclass-memaccess option is enabled by -Wall.  Explicitly
2407           casting the pointer to the class object to "void *" or to a type
2408           that can be safely accessed by the raw memory function suppresses
2409           the warning.
2410
2411       -Wnon-virtual-dtor (C++ and Objective-C++ only)
2412           Warn when a class has virtual functions and an accessible non-
2413           virtual destructor itself or in an accessible polymorphic base
2414           class, in which case it is possible but unsafe to delete an
2415           instance of a derived class through a pointer to the class itself
2416           or base class.  This warning is automatically enabled if -Weffc++
2417           is specified.
2418
2419       -Wregister (C++ and Objective-C++ only)
2420           Warn on uses of the "register" storage class specifier, except when
2421           it is part of the GNU Explicit Register Variables extension.  The
2422           use of the "register" keyword as storage class specifier has been
2423           deprecated in C++11 and removed in C++17.  Enabled by default with
2424           -std=c++17.
2425
2426       -Wreorder (C++ and Objective-C++ only)
2427           Warn when the order of member initializers given in the code does
2428           not match the order in which they must be executed.  For instance:
2429
2430                   struct A {
2431                     int i;
2432                     int j;
2433                     A(): j (0), i (1) { }
2434                   };
2435
2436           The compiler rearranges the member initializers for "i" and "j" to
2437           match the declaration order of the members, emitting a warning to
2438           that effect.  This warning is enabled by -Wall.
2439
2440       -Wno-pessimizing-move (C++ and Objective-C++ only)
2441           This warning warns when a call to "std::move" prevents copy
2442           elision.  A typical scenario when copy elision can occur is when
2443           returning in a function with a class return type, when the
2444           expression being returned is the name of a non-volatile automatic
2445           object, and is not a function parameter, and has the same type as
2446           the function return type.
2447
2448                   struct T {
2449                   ...
2450                   };
2451                   T fn()
2452                   {
2453                     T t;
2454                     ...
2455                     return std::move (t);
2456                   }
2457
2458           But in this example, the "std::move" call prevents copy elision.
2459
2460           This warning is enabled by -Wall.
2461
2462       -Wno-redundant-move (C++ and Objective-C++ only)
2463           This warning warns about redundant calls to "std::move"; that is,
2464           when a move operation would have been performed even without the
2465           "std::move" call.  This happens because the compiler is forced to
2466           treat the object as if it were an rvalue in certain situations such
2467           as returning a local variable, where copy elision isn't applicable.
2468           Consider:
2469
2470                   struct T {
2471                   ...
2472                   };
2473                   T fn(T t)
2474                   {
2475                     ...
2476                     return std::move (t);
2477                   }
2478
2479           Here, the "std::move" call is redundant.  Because G++ implements
2480           Core Issue 1579, another example is:
2481
2482                   struct T { // convertible to U
2483                   ...
2484                   };
2485                   struct U {
2486                   ...
2487                   };
2488                   U fn()
2489                   {
2490                     T t;
2491                     ...
2492                     return std::move (t);
2493                   }
2494
2495           In this example, copy elision isn't applicable because the type of
2496           the expression being returned and the function return type differ,
2497           yet G++ treats the return value as if it were designated by an
2498           rvalue.
2499
2500           This warning is enabled by -Wextra.
2501
2502       -Wredundant-tags (C++ and Objective-C++ only)
2503           Warn about redundant class-key and enum-key in references to class
2504           types and enumerated types in contexts where the key can be
2505           eliminated without causing an ambiguity.  For example:
2506
2507                   struct foo;
2508                   struct foo *p;   // warn that keyword struct can be eliminated
2509
2510           On the other hand, in this example there is no warning:
2511
2512                   struct foo;
2513                   void foo ();   // "hides" struct foo
2514                   void bar (struct foo&);  // no warning, keyword struct is necessary
2515
2516       -Wno-subobject-linkage (C++ and Objective-C++ only)
2517           Do not warn if a class type has a base or a field whose type uses
2518           the anonymous namespace or depends on a type with no linkage.  If a
2519           type A depends on a type B with no or internal linkage, defining it
2520           in multiple translation units would be an ODR violation because the
2521           meaning of B is different in each translation unit.  If A only
2522           appears in a single translation unit, the best way to silence the
2523           warning is to give it internal linkage by putting it in an
2524           anonymous namespace as well.  The compiler doesn't give this
2525           warning for types defined in the main .C file, as those are
2526           unlikely to have multiple definitions.  -Wsubobject-linkage is
2527           enabled by default.
2528
2529       -Weffc++ (C++ and Objective-C++ only)
2530           Warn about violations of the following style guidelines from Scott
2531           Meyers' Effective C++ series of books:
2532
2533           *   Define a copy constructor and an assignment operator for
2534               classes with dynamically-allocated memory.
2535
2536           *   Prefer initialization to assignment in constructors.
2537
2538           *   Have "operator=" return a reference to *this.
2539
2540           *   Don't try to return a reference when you must return an object.
2541
2542           *   Distinguish between prefix and postfix forms of increment and
2543               decrement operators.
2544
2545           *   Never overload "&&", "||", or ",".
2546
2547           This option also enables -Wnon-virtual-dtor, which is also one of
2548           the effective C++ recommendations.  However, the check is extended
2549           to warn about the lack of virtual destructor in accessible non-
2550           polymorphic bases classes too.
2551
2552           When selecting this option, be aware that the standard library
2553           headers do not obey all of these guidelines; use grep -v to filter
2554           out those warnings.
2555
2556       -Wstrict-null-sentinel (C++ and Objective-C++ only)
2557           Warn about the use of an uncasted "NULL" as sentinel.  When
2558           compiling only with GCC this is a valid sentinel, as "NULL" is
2559           defined to "__null".  Although it is a null pointer constant rather
2560           than a null pointer, it is guaranteed to be of the same size as a
2561           pointer.  But this use is not portable across different compilers.
2562
2563       -Wno-non-template-friend (C++ and Objective-C++ only)
2564           Disable warnings when non-template friend functions are declared
2565           within a template.  In very old versions of GCC that predate
2566           implementation of the ISO standard, declarations such as friend int
2567           foo(int), where the name of the friend is an unqualified-id, could
2568           be interpreted as a particular specialization of a template
2569           function; the warning exists to diagnose compatibility problems,
2570           and is enabled by default.
2571
2572       -Wold-style-cast (C++ and Objective-C++ only)
2573           Warn if an old-style (C-style) cast to a non-void type is used
2574           within a C++ program.  The new-style casts ("dynamic_cast",
2575           "static_cast", "reinterpret_cast", and "const_cast") are less
2576           vulnerable to unintended effects and much easier to search for.
2577
2578       -Woverloaded-virtual (C++ and Objective-C++ only)
2579           Warn when a function declaration hides virtual functions from a
2580           base class.  For example, in:
2581
2582                   struct A {
2583                     virtual void f();
2584                   };
2585
2586                   struct B: public A {
2587                     void f(int);
2588                   };
2589
2590           the "A" class version of "f" is hidden in "B", and code like:
2591
2592                   B* b;
2593                   b->f();
2594
2595           fails to compile.
2596
2597       -Wno-pmf-conversions (C++ and Objective-C++ only)
2598           Disable the diagnostic for converting a bound pointer to member
2599           function to a plain pointer.
2600
2601       -Wsign-promo (C++ and Objective-C++ only)
2602           Warn when overload resolution chooses a promotion from unsigned or
2603           enumerated type to a signed type, over a conversion to an unsigned
2604           type of the same size.  Previous versions of G++ tried to preserve
2605           unsignedness, but the standard mandates the current behavior.
2606
2607       -Wtemplates (C++ and Objective-C++ only)
2608           Warn when a primary template declaration is encountered.  Some
2609           coding rules disallow templates, and this may be used to enforce
2610           that rule.  The warning is inactive inside a system header file,
2611           such as the STL, so one can still use the STL.  One may also
2612           instantiate or specialize templates.
2613
2614       -Wmismatched-tags (C++ and Objective-C++ only)
2615           Warn for declarations of structs, classes, and class templates and
2616           their specializations with a class-key that does not match either
2617           the definition or the first declaration if no definition is
2618           provided.
2619
2620           For example, the declaration of "struct Object" in the argument
2621           list of "draw" triggers the warning.  To avoid it, either remove
2622           the redundant class-key "struct" or replace it with "class" to
2623           match its definition.
2624
2625                   class Object {
2626                   public:
2627                     virtual ~Object () = 0;
2628                   };
2629                   void draw (struct Object*);
2630
2631           It is not wrong to declare a class with the class-key "struct" as
2632           the example above shows.  The -Wmismatched-tags option is intended
2633           to help achieve a consistent style of class declarations.  In code
2634           that is intended to be portable to Windows-based compilers the
2635           warning helps prevent unresolved references due to the difference
2636           in the mangling of symbols declared with different class-keys.  The
2637           option can be used either on its own or in conjunction with
2638           -Wredundant-tags.
2639
2640       -Wmultiple-inheritance (C++ and Objective-C++ only)
2641           Warn when a class is defined with multiple direct base classes.
2642           Some coding rules disallow multiple inheritance, and this may be
2643           used to enforce that rule.  The warning is inactive inside a system
2644           header file, such as the STL, so one can still use the STL.  One
2645           may also define classes that indirectly use multiple inheritance.
2646
2647       -Wvirtual-inheritance
2648           Warn when a class is defined with a virtual direct base class.
2649           Some coding rules disallow multiple inheritance, and this may be
2650           used to enforce that rule.  The warning is inactive inside a system
2651           header file, such as the STL, so one can still use the STL.  One
2652           may also define classes that indirectly use virtual inheritance.
2653
2654       -Wno-virtual-move-assign
2655           Suppress warnings about inheriting from a virtual base with a non-
2656           trivial C++11 move assignment operator.  This is dangerous because
2657           if the virtual base is reachable along more than one path, it is
2658           moved multiple times, which can mean both objects end up in the
2659           moved-from state.  If the move assignment operator is written to
2660           avoid moving from a moved-from object, this warning can be
2661           disabled.
2662
2663       -Wnamespaces
2664           Warn when a namespace definition is opened.  Some coding rules
2665           disallow namespaces, and this may be used to enforce that rule.
2666           The warning is inactive inside a system header file, such as the
2667           STL, so one can still use the STL.  One may also use using
2668           directives and qualified names.
2669
2670       -Wno-terminate (C++ and Objective-C++ only)
2671           Disable the warning about a throw-expression that will immediately
2672           result in a call to "terminate".
2673
2674       -Wno-class-conversion (C++ and Objective-C++ only)
2675           Do not warn when a conversion function converts an object to the
2676           same type, to a base class of that type, or to void; such a
2677           conversion function will never be called.
2678
2679       -Wvolatile (C++ and Objective-C++ only)
2680           Warn about deprecated uses of the "volatile" qualifier.  This
2681           includes postfix and prefix "++" and "--" expressions of
2682           "volatile"-qualified types, using simple assignments where the left
2683           operand is a "volatile"-qualified non-class type for their value,
2684           compound assignments where the left operand is a
2685           "volatile"-qualified non-class type, "volatile"-qualified function
2686           return type, "volatile"-qualified parameter type, and structured
2687           bindings of a "volatile"-qualified type.  This usage was deprecated
2688           in C++20.
2689
2690           Enabled by default with -std=c++2a.
2691
2692       -Wzero-as-null-pointer-constant (C++ and Objective-C++ only)
2693           Warn when a literal 0 is used as null pointer constant.  This can
2694           be useful to facilitate the conversion to "nullptr" in C++11.
2695
2696       -Waligned-new
2697           Warn about a new-expression of a type that requires greater
2698           alignment than the "alignof(std::max_align_t)" but uses an
2699           allocation function without an explicit alignment parameter. This
2700           option is enabled by -Wall.
2701
2702           Normally this only warns about global allocation functions, but
2703           -Waligned-new=all also warns about class member allocation
2704           functions.
2705
2706       -Wno-placement-new
2707       -Wplacement-new=n
2708           Warn about placement new expressions with undefined behavior, such
2709           as constructing an object in a buffer that is smaller than the type
2710           of the object.  For example, the placement new expression below is
2711           diagnosed because it attempts to construct an array of 64 integers
2712           in a buffer only 64 bytes large.
2713
2714                   char buf [64];
2715                   new (buf) int[64];
2716
2717           This warning is enabled by default.
2718
2719           -Wplacement-new=1
2720               This is the default warning level of -Wplacement-new.  At this
2721               level the warning is not issued for some strictly undefined
2722               constructs that GCC allows as extensions for compatibility with
2723               legacy code.  For example, the following "new" expression is
2724               not diagnosed at this level even though it has undefined
2725               behavior according to the C++ standard because it writes past
2726               the end of the one-element array.
2727
2728                       struct S { int n, a[1]; };
2729                       S *s = (S *)malloc (sizeof *s + 31 * sizeof s->a[0]);
2730                       new (s->a)int [32]();
2731
2732           -Wplacement-new=2
2733               At this level, in addition to diagnosing all the same
2734               constructs as at level 1, a diagnostic is also issued for
2735               placement new expressions that construct an object in the last
2736               member of structure whose type is an array of a single element
2737               and whose size is less than the size of the object being
2738               constructed.  While the previous example would be diagnosed,
2739               the following construct makes use of the flexible member array
2740               extension to avoid the warning at level 2.
2741
2742                       struct S { int n, a[]; };
2743                       S *s = (S *)malloc (sizeof *s + 32 * sizeof s->a[0]);
2744                       new (s->a)int [32]();
2745
2746       -Wcatch-value
2747       -Wcatch-value=n (C++ and Objective-C++ only)
2748           Warn about catch handlers that do not catch via reference.  With
2749           -Wcatch-value=1 (or -Wcatch-value for short) warn about polymorphic
2750           class types that are caught by value.  With -Wcatch-value=2 warn
2751           about all class types that are caught by value. With
2752           -Wcatch-value=3 warn about all types that are not caught by
2753           reference. -Wcatch-value is enabled by -Wall.
2754
2755       -Wconditionally-supported (C++ and Objective-C++ only)
2756           Warn for conditionally-supported (C++11 [intro.defs]) constructs.
2757
2758       -Wno-delete-incomplete (C++ and Objective-C++ only)
2759           Do not warn when deleting a pointer to incomplete type, which may
2760           cause undefined behavior at runtime.  This warning is enabled by
2761           default.
2762
2763       -Wextra-semi (C++, Objective-C++ only)
2764           Warn about redundant semicolons after in-class function
2765           definitions.
2766
2767       -Wno-inaccessible-base (C++, Objective-C++ only)
2768           This option controls warnings when a base class is inaccessible in
2769           a class derived from it due to ambiguity.  The warning is enabled
2770           by default.  Note that the warning for ambiguous virtual bases is
2771           enabled by the -Wextra option.
2772
2773                   struct A { int a; };
2774
2775                   struct B : A { };
2776
2777                   struct C : B, A { };
2778
2779       -Wno-inherited-variadic-ctor
2780           Suppress warnings about use of C++11 inheriting constructors when
2781           the base class inherited from has a C variadic constructor; the
2782           warning is on by default because the ellipsis is not inherited.
2783
2784       -Wno-invalid-offsetof (C++ and Objective-C++ only)
2785           Suppress warnings from applying the "offsetof" macro to a non-POD
2786           type.  According to the 2014 ISO C++ standard, applying "offsetof"
2787           to a non-standard-layout type is undefined.  In existing C++
2788           implementations, however, "offsetof" typically gives meaningful
2789           results.  This flag is for users who are aware that they are
2790           writing nonportable code and who have deliberately chosen to ignore
2791           the warning about it.
2792
2793           The restrictions on "offsetof" may be relaxed in a future version
2794           of the C++ standard.
2795
2796       -Wsized-deallocation (C++ and Objective-C++ only)
2797           Warn about a definition of an unsized deallocation function
2798
2799                   void operator delete (void *) noexcept;
2800                   void operator delete[] (void *) noexcept;
2801
2802           without a definition of the corresponding sized deallocation
2803           function
2804
2805                   void operator delete (void *, std::size_t) noexcept;
2806                   void operator delete[] (void *, std::size_t) noexcept;
2807
2808           or vice versa.  Enabled by -Wextra along with -fsized-deallocation.
2809
2810       -Wsuggest-final-types
2811           Warn about types with virtual methods where code quality would be
2812           improved if the type were declared with the C++11 "final"
2813           specifier, or, if possible, declared in an anonymous namespace.
2814           This allows GCC to more aggressively devirtualize the polymorphic
2815           calls. This warning is more effective with link-time optimization,
2816           where the information about the class hierarchy graph is more
2817           complete.
2818
2819       -Wsuggest-final-methods
2820           Warn about virtual methods where code quality would be improved if
2821           the method were declared with the C++11 "final" specifier, or, if
2822           possible, its type were declared in an anonymous namespace or with
2823           the "final" specifier.  This warning is more effective with link-
2824           time optimization, where the information about the class hierarchy
2825           graph is more complete. It is recommended to first consider
2826           suggestions of -Wsuggest-final-types and then rebuild with new
2827           annotations.
2828
2829       -Wsuggest-override
2830           Warn about overriding virtual functions that are not marked with
2831           the "override" keyword.
2832
2833       -Wuseless-cast (C++ and Objective-C++ only)
2834           Warn when an expression is casted to its own type.
2835
2836       -Wno-conversion-null (C++ and Objective-C++ only)
2837           Do not warn for conversions between "NULL" and non-pointer types.
2838           -Wconversion-null is enabled by default.
2839
2840   Options Controlling Objective-C and Objective-C++ Dialects
2841       (NOTE: This manual does not describe the Objective-C and Objective-C++
2842       languages themselves.
2843
2844       This section describes the command-line options that are only
2845       meaningful for Objective-C and Objective-C++ programs.  You can also
2846       use most of the language-independent GNU compiler options.  For
2847       example, you might compile a file some_class.m like this:
2848
2849               gcc -g -fgnu-runtime -O -c some_class.m
2850
2851       In this example, -fgnu-runtime is an option meant only for Objective-C
2852       and Objective-C++ programs; you can use the other options with any
2853       language supported by GCC.
2854
2855       Note that since Objective-C is an extension of the C language,
2856       Objective-C compilations may also use options specific to the C front-
2857       end (e.g., -Wtraditional).  Similarly, Objective-C++ compilations may
2858       use C++-specific options (e.g., -Wabi).
2859
2860       Here is a list of options that are only for compiling Objective-C and
2861       Objective-C++ programs:
2862
2863       -fconstant-string-class=class-name
2864           Use class-name as the name of the class to instantiate for each
2865           literal string specified with the syntax "@"..."".  The default
2866           class name is "NXConstantString" if the GNU runtime is being used,
2867           and "NSConstantString" if the NeXT runtime is being used (see
2868           below).  The -fconstant-cfstrings option, if also present,
2869           overrides the -fconstant-string-class setting and cause "@"...""
2870           literals to be laid out as constant CoreFoundation strings.
2871
2872       -fgnu-runtime
2873           Generate object code compatible with the standard GNU Objective-C
2874           runtime.  This is the default for most types of systems.
2875
2876       -fnext-runtime
2877           Generate output compatible with the NeXT runtime.  This is the
2878           default for NeXT-based systems, including Darwin and Mac OS X.  The
2879           macro "__NEXT_RUNTIME__" is predefined if (and only if) this option
2880           is used.
2881
2882       -fno-nil-receivers
2883           Assume that all Objective-C message dispatches ("[receiver
2884           message:arg]") in this translation unit ensure that the receiver is
2885           not "nil".  This allows for more efficient entry points in the
2886           runtime to be used.  This option is only available in conjunction
2887           with the NeXT runtime and ABI version 0 or 1.
2888
2889       -fobjc-abi-version=n
2890           Use version n of the Objective-C ABI for the selected runtime.
2891           This option is currently supported only for the NeXT runtime.  In
2892           that case, Version 0 is the traditional (32-bit) ABI without
2893           support for properties and other Objective-C 2.0 additions.
2894           Version 1 is the traditional (32-bit) ABI with support for
2895           properties and other Objective-C 2.0 additions.  Version 2 is the
2896           modern (64-bit) ABI.  If nothing is specified, the default is
2897           Version 0 on 32-bit target machines, and Version 2 on 64-bit target
2898           machines.
2899
2900       -fobjc-call-cxx-cdtors
2901           For each Objective-C class, check if any of its instance variables
2902           is a C++ object with a non-trivial default constructor.  If so,
2903           synthesize a special "- (id) .cxx_construct" instance method which
2904           runs non-trivial default constructors on any such instance
2905           variables, in order, and then return "self".  Similarly, check if
2906           any instance variable is a C++ object with a non-trivial
2907           destructor, and if so, synthesize a special "- (void)
2908           .cxx_destruct" method which runs all such default destructors, in
2909           reverse order.
2910
2911           The "- (id) .cxx_construct" and "- (void) .cxx_destruct" methods
2912           thusly generated only operate on instance variables declared in the
2913           current Objective-C class, and not those inherited from
2914           superclasses.  It is the responsibility of the Objective-C runtime
2915           to invoke all such methods in an object's inheritance hierarchy.
2916           The "- (id) .cxx_construct" methods are invoked by the runtime
2917           immediately after a new object instance is allocated; the "- (void)
2918           .cxx_destruct" methods are invoked immediately before the runtime
2919           deallocates an object instance.
2920
2921           As of this writing, only the NeXT runtime on Mac OS X 10.4 and
2922           later has support for invoking the "- (id) .cxx_construct" and "-
2923           (void) .cxx_destruct" methods.
2924
2925       -fobjc-direct-dispatch
2926           Allow fast jumps to the message dispatcher.  On Darwin this is
2927           accomplished via the comm page.
2928
2929       -fobjc-exceptions
2930           Enable syntactic support for structured exception handling in
2931           Objective-C, similar to what is offered by C++.  This option is
2932           required to use the Objective-C keywords @try, @throw, @catch,
2933           @finally and @synchronized.  This option is available with both the
2934           GNU runtime and the NeXT runtime (but not available in conjunction
2935           with the NeXT runtime on Mac OS X 10.2 and earlier).
2936
2937       -fobjc-gc
2938           Enable garbage collection (GC) in Objective-C and Objective-C++
2939           programs.  This option is only available with the NeXT runtime; the
2940           GNU runtime has a different garbage collection implementation that
2941           does not require special compiler flags.
2942
2943       -fobjc-nilcheck
2944           For the NeXT runtime with version 2 of the ABI, check for a nil
2945           receiver in method invocations before doing the actual method call.
2946           This is the default and can be disabled using -fno-objc-nilcheck.
2947           Class methods and super calls are never checked for nil in this way
2948           no matter what this flag is set to.  Currently this flag does
2949           nothing when the GNU runtime, or an older version of the NeXT
2950           runtime ABI, is used.
2951
2952       -fobjc-std=objc1
2953           Conform to the language syntax of Objective-C 1.0, the language
2954           recognized by GCC 4.0.  This only affects the Objective-C additions
2955           to the C/C++ language; it does not affect conformance to C/C++
2956           standards, which is controlled by the separate C/C++ dialect option
2957           flags.  When this option is used with the Objective-C or
2958           Objective-C++ compiler, any Objective-C syntax that is not
2959           recognized by GCC 4.0 is rejected.  This is useful if you need to
2960           make sure that your Objective-C code can be compiled with older
2961           versions of GCC.
2962
2963       -freplace-objc-classes
2964           Emit a special marker instructing ld(1) not to statically link in
2965           the resulting object file, and allow dyld(1) to load it in at run
2966           time instead.  This is used in conjunction with the Fix-and-
2967           Continue debugging mode, where the object file in question may be
2968           recompiled and dynamically reloaded in the course of program
2969           execution, without the need to restart the program itself.
2970           Currently, Fix-and-Continue functionality is only available in
2971           conjunction with the NeXT runtime on Mac OS X 10.3 and later.
2972
2973       -fzero-link
2974           When compiling for the NeXT runtime, the compiler ordinarily
2975           replaces calls to "objc_getClass("...")" (when the name of the
2976           class is known at compile time) with static class references that
2977           get initialized at load time, which improves run-time performance.
2978           Specifying the -fzero-link flag suppresses this behavior and causes
2979           calls to "objc_getClass("...")"  to be retained.  This is useful in
2980           Zero-Link debugging mode, since it allows for individual class
2981           implementations to be modified during program execution.  The GNU
2982           runtime currently always retains calls to "objc_get_class("...")"
2983           regardless of command-line options.
2984
2985       -fno-local-ivars
2986           By default instance variables in Objective-C can be accessed as if
2987           they were local variables from within the methods of the class
2988           they're declared in.  This can lead to shadowing between instance
2989           variables and other variables declared either locally inside a
2990           class method or globally with the same name.  Specifying the
2991           -fno-local-ivars flag disables this behavior thus avoiding variable
2992           shadowing issues.
2993
2994       -fivar-visibility=[public|protected|private|package]
2995           Set the default instance variable visibility to the specified
2996           option so that instance variables declared outside the scope of any
2997           access modifier directives default to the specified visibility.
2998
2999       -gen-decls
3000           Dump interface declarations for all classes seen in the source file
3001           to a file named sourcename.decl.
3002
3003       -Wassign-intercept (Objective-C and Objective-C++ only)
3004           Warn whenever an Objective-C assignment is being intercepted by the
3005           garbage collector.
3006
3007       -Wno-property-assign-default (Objective-C and Objective-C++ only)
3008           Do not warn if a property for an Objective-C object has no assign
3009           semantics specified.
3010
3011       -Wno-protocol (Objective-C and Objective-C++ only)
3012           If a class is declared to implement a protocol, a warning is issued
3013           for every method in the protocol that is not implemented by the
3014           class.  The default behavior is to issue a warning for every method
3015           not explicitly implemented in the class, even if a method
3016           implementation is inherited from the superclass.  If you use the
3017           -Wno-protocol option, then methods inherited from the superclass
3018           are considered to be implemented, and no warning is issued for
3019           them.
3020
3021       -Wselector (Objective-C and Objective-C++ only)
3022           Warn if multiple methods of different types for the same selector
3023           are found during compilation.  The check is performed on the list
3024           of methods in the final stage of compilation.  Additionally, a
3025           check is performed for each selector appearing in a
3026           "@selector(...)"  expression, and a corresponding method for that
3027           selector has been found during compilation.  Because these checks
3028           scan the method table only at the end of compilation, these
3029           warnings are not produced if the final stage of compilation is not
3030           reached, for example because an error is found during compilation,
3031           or because the -fsyntax-only option is being used.
3032
3033       -Wstrict-selector-match (Objective-C and Objective-C++ only)
3034           Warn if multiple methods with differing argument and/or return
3035           types are found for a given selector when attempting to send a
3036           message using this selector to a receiver of type "id" or "Class".
3037           When this flag is off (which is the default behavior), the compiler
3038           omits such warnings if any differences found are confined to types
3039           that share the same size and alignment.
3040
3041       -Wundeclared-selector (Objective-C and Objective-C++ only)
3042           Warn if a "@selector(...)" expression referring to an undeclared
3043           selector is found.  A selector is considered undeclared if no
3044           method with that name has been declared before the "@selector(...)"
3045           expression, either explicitly in an @interface or @protocol
3046           declaration, or implicitly in an @implementation section.  This
3047           option always performs its checks as soon as a "@selector(...)"
3048           expression is found, while -Wselector only performs its checks in
3049           the final stage of compilation.  This also enforces the coding
3050           style convention that methods and selectors must be declared before
3051           being used.
3052
3053       -print-objc-runtime-info
3054           Generate C header describing the largest structure that is passed
3055           by value, if any.
3056
3057   Options to Control Diagnostic Messages Formatting
3058       Traditionally, diagnostic messages have been formatted irrespective of
3059       the output device's aspect (e.g. its width, ...).  You can use the
3060       options described below to control the formatting algorithm for
3061       diagnostic messages, e.g. how many characters per line, how often
3062       source location information should be reported.  Note that some
3063       language front ends may not honor these options.
3064
3065       -fmessage-length=n
3066           Try to format error messages so that they fit on lines of about n
3067           characters.  If n is zero, then no line-wrapping is done; each
3068           error message appears on a single line.  This is the default for
3069           all front ends.
3070
3071           Note - this option also affects the display of the #error and
3072           #warning pre-processor directives, and the deprecated
3073           function/type/variable attribute.  It does not however affect the
3074           pragma GCC warning and pragma GCC error pragmas.
3075
3076       -fdiagnostics-show-location=once
3077           Only meaningful in line-wrapping mode.  Instructs the diagnostic
3078           messages reporter to emit source location information once; that
3079           is, in case the message is too long to fit on a single physical
3080           line and has to be wrapped, the source location won't be emitted
3081           (as prefix) again, over and over, in subsequent continuation lines.
3082           This is the default behavior.
3083
3084       -fdiagnostics-show-location=every-line
3085           Only meaningful in line-wrapping mode.  Instructs the diagnostic
3086           messages reporter to emit the same source location information (as
3087           prefix) for physical lines that result from the process of breaking
3088           a message which is too long to fit on a single line.
3089
3090       -fdiagnostics-color[=WHEN]
3091       -fno-diagnostics-color
3092           Use color in diagnostics.  WHEN is never, always, or auto.  The
3093           default depends on how the compiler has been configured, it can be
3094           any of the above WHEN options or also never if GCC_COLORS
3095           environment variable isn't present in the environment, and auto
3096           otherwise.  auto makes GCC use color only when the standard error
3097           is a terminal, and when not executing in an emacs shell.  The forms
3098           -fdiagnostics-color and -fno-diagnostics-color are aliases for
3099           -fdiagnostics-color=always and -fdiagnostics-color=never,
3100           respectively.
3101
3102           The colors are defined by the environment variable GCC_COLORS.  Its
3103           value is a colon-separated list of capabilities and Select Graphic
3104           Rendition (SGR) substrings. SGR commands are interpreted by the
3105           terminal or terminal emulator.  (See the section in the
3106           documentation of your text terminal for permitted values and their
3107           meanings as character attributes.)  These substring values are
3108           integers in decimal representation and can be concatenated with
3109           semicolons.  Common values to concatenate include 1 for bold, 4 for
3110           underline, 5 for blink, 7 for inverse, 39 for default foreground
3111           color, 30 to 37 for foreground colors, 90 to 97 for 16-color mode
3112           foreground colors, 38;5;0 to 38;5;255 for 88-color and 256-color
3113           modes foreground colors, 49 for default background color, 40 to 47
3114           for background colors, 100 to 107 for 16-color mode background
3115           colors, and 48;5;0 to 48;5;255 for 88-color and 256-color modes
3116           background colors.
3117
3118           The default GCC_COLORS is
3119
3120                   error=01;31:warning=01;35:note=01;36:range1=32:range2=34:locus=01:\
3121                   quote=01:path=01;36:fixit-insert=32:fixit-delete=31:\
3122                   diff-filename=01:diff-hunk=32:diff-delete=31:diff-insert=32:\
3123                   type-diff=01;32
3124
3125           where 01;31 is bold red, 01;35 is bold magenta, 01;36 is bold cyan,
3126           32 is green, 34 is blue, 01 is bold, and 31 is red.  Setting
3127           GCC_COLORS to the empty string disables colors.  Supported
3128           capabilities are as follows.
3129
3130           "error="
3131               SGR substring for error: markers.
3132
3133           "warning="
3134               SGR substring for warning: markers.
3135
3136           "note="
3137               SGR substring for note: markers.
3138
3139           "path="
3140               SGR substring for colorizing paths of control-flow events as
3141               printed via -fdiagnostics-path-format=, such as the identifiers
3142               of individual events and lines indicating interprocedural calls
3143               and returns.
3144
3145           "range1="
3146               SGR substring for first additional range.
3147
3148           "range2="
3149               SGR substring for second additional range.
3150
3151           "locus="
3152               SGR substring for location information, file:line or
3153               file:line:column etc.
3154
3155           "quote="
3156               SGR substring for information printed within quotes.
3157
3158           "fixit-insert="
3159               SGR substring for fix-it hints suggesting text to be inserted
3160               or replaced.
3161
3162           "fixit-delete="
3163               SGR substring for fix-it hints suggesting text to be deleted.
3164
3165           "diff-filename="
3166               SGR substring for filename headers within generated patches.
3167
3168           "diff-hunk="
3169               SGR substring for the starts of hunks within generated patches.
3170
3171           "diff-delete="
3172               SGR substring for deleted lines within generated patches.
3173
3174           "diff-insert="
3175               SGR substring for inserted lines within generated patches.
3176
3177           "type-diff="
3178               SGR substring for highlighting mismatching types within
3179               template arguments in the C++ frontend.
3180
3181       -fdiagnostics-urls[=WHEN]
3182           Use escape sequences to embed URLs in diagnostics.  For example,
3183           when -fdiagnostics-show-option emits text showing the command-line
3184           option controlling a diagnostic, embed a URL for documentation of
3185           that option.
3186
3187           WHEN is never, always, or auto.  auto makes GCC use URL escape
3188           sequences only when the standard error is a terminal, and when not
3189           executing in an emacs shell or any graphical terminal which is
3190           known to be incompatible with this feature, see below.
3191
3192           The default depends on how the compiler has been configured.  It
3193           can be any of the above WHEN options.
3194
3195           GCC can also be configured (via the
3196           --with-diagnostics-urls=auto-if-env configure-time option) so that
3197           the default is affected by environment variables.  Under such a
3198           configuration, GCC defaults to using auto if either GCC_URLS or
3199           TERM_URLS environment variables are present and non-empty in the
3200           environment of the compiler, or never if neither are.
3201
3202           However, even with -fdiagnostics-urls=always the behavior is
3203           dependent on those environment variables: If GCC_URLS is set to
3204           empty or no, do not embed URLs in diagnostics.  If set to st, URLs
3205           use ST escape sequences.  If set to bel, the default, URLs use BEL
3206           escape sequences.  Any other non-empty value enables the feature.
3207           If GCC_URLS is not set, use TERM_URLS as a fallback.  Note: ST is
3208           an ANSI escape sequence, string terminator ESC \, BEL is an ASCII
3209           character, CTRL-G that usually sounds like a beep.
3210
3211           At this time GCC tries to detect also a few terminals that are
3212           known to not implement the URL feature, and have bugs or at least
3213           had bugs in some versions that are still in use, where the URL
3214           escapes are likely to misbehave, i.e. print garbage on the screen.
3215           That list is currently xfce4-terminal, certain known to be buggy
3216           gnome-terminal versions, the linux console, and mingw.  This check
3217           can be skipped with the -fdiagnostics-urls=always.
3218
3219       -fno-diagnostics-show-option
3220           By default, each diagnostic emitted includes text indicating the
3221           command-line option that directly controls the diagnostic (if such
3222           an option is known to the diagnostic machinery).  Specifying the
3223           -fno-diagnostics-show-option flag suppresses that behavior.
3224
3225       -fno-diagnostics-show-caret
3226           By default, each diagnostic emitted includes the original source
3227           line and a caret ^ indicating the column.  This option suppresses
3228           this information.  The source line is truncated to n characters, if
3229           the -fmessage-length=n option is given.  When the output is done to
3230           the terminal, the width is limited to the width given by the
3231           COLUMNS environment variable or, if not set, to the terminal width.
3232
3233       -fno-diagnostics-show-labels
3234           By default, when printing source code (via
3235           -fdiagnostics-show-caret), diagnostics can label ranges of source
3236           code with pertinent information, such as the types of expressions:
3237
3238                       printf ("foo %s bar", long_i + long_j);
3239                                    ~^       ~~~~~~~~~~~~~~~
3240                                     |              |
3241                                     char *         long int
3242
3243           This option suppresses the printing of these labels (in the example
3244           above, the vertical bars and the "char *" and "long int" text).
3245
3246       -fno-diagnostics-show-cwe
3247           Diagnostic messages can optionally have an associated
3248           @url{https://cwe.mitre.org/index.html, CWE} identifier.  GCC itself
3249           only provides such metadata for some of the -fanalyzer diagnostics.
3250           GCC plugins may also provide diagnostics with such metadata.  By
3251           default, if this information is present, it will be printed with
3252           the diagnostic.  This option suppresses the printing of this
3253           metadata.
3254
3255       -fno-diagnostics-show-line-numbers
3256           By default, when printing source code (via
3257           -fdiagnostics-show-caret), a left margin is printed, showing line
3258           numbers.  This option suppresses this left margin.
3259
3260       -fdiagnostics-minimum-margin-width=width
3261           This option controls the minimum width of the left margin printed
3262           by -fdiagnostics-show-line-numbers.  It defaults to 6.
3263
3264       -fdiagnostics-parseable-fixits
3265           Emit fix-it hints in a machine-parseable format, suitable for
3266           consumption by IDEs.  For each fix-it, a line will be printed after
3267           the relevant diagnostic, starting with the string "fix-it:".  For
3268           example:
3269
3270                   fix-it:"test.c":{45:3-45:21}:"gtk_widget_show_all"
3271
3272           The location is expressed as a half-open range, expressed as a
3273           count of bytes, starting at byte 1 for the initial column.  In the
3274           above example, bytes 3 through 20 of line 45 of "test.c" are to be
3275           replaced with the given string:
3276
3277                   00000000011111111112222222222
3278                   12345678901234567890123456789
3279                     gtk_widget_showall (dlg);
3280                     ^^^^^^^^^^^^^^^^^^
3281                     gtk_widget_show_all
3282
3283           The filename and replacement string escape backslash as "\\", tab
3284           as "\t", newline as "\n", double quotes as "\"", non-printable
3285           characters as octal (e.g. vertical tab as "\013").
3286
3287           An empty replacement string indicates that the given range is to be
3288           removed.  An empty range (e.g. "45:3-45:3") indicates that the
3289           string is to be inserted at the given position.
3290
3291       -fdiagnostics-generate-patch
3292           Print fix-it hints to stderr in unified diff format, after any
3293           diagnostics are printed.  For example:
3294
3295                   --- test.c
3296                   +++ test.c
3297                   @ -42,5 +42,5 @
3298
3299                    void show_cb(GtkDialog *dlg)
3300                    {
3301                   -  gtk_widget_showall(dlg);
3302                   +  gtk_widget_show_all(dlg);
3303                    }
3304
3305           The diff may or may not be colorized, following the same rules as
3306           for diagnostics (see -fdiagnostics-color).
3307
3308       -fdiagnostics-show-template-tree
3309           In the C++ frontend, when printing diagnostics showing mismatching
3310           template types, such as:
3311
3312                     could not convert 'std::map<int, std::vector<double> >()'
3313                       from 'map<[...],vector<double>>' to 'map<[...],vector<float>>
3314
3315           the -fdiagnostics-show-template-tree flag enables printing a tree-
3316           like structure showing the common and differing parts of the types,
3317           such as:
3318
3319                     map<
3320                       [...],
3321                       vector<
3322                         [double != float]>>
3323
3324           The parts that differ are highlighted with color ("double" and
3325           "float" in this case).
3326
3327       -fno-elide-type
3328           By default when the C++ frontend prints diagnostics showing
3329           mismatching template types, common parts of the types are printed
3330           as "[...]" to simplify the error message.  For example:
3331
3332                     could not convert 'std::map<int, std::vector<double> >()'
3333                       from 'map<[...],vector<double>>' to 'map<[...],vector<float>>
3334
3335           Specifying the -fno-elide-type flag suppresses that behavior.  This
3336           flag also affects the output of the
3337           -fdiagnostics-show-template-tree flag.
3338
3339       -fdiagnostics-path-format=KIND
3340           Specify how to print paths of control-flow events for diagnostics
3341           that have such a path associated with them.
3342
3343           KIND is none, separate-events, or inline-events, the default.
3344
3345           none means to not print diagnostic paths.
3346
3347           separate-events means to print a separate "note" diagnostic for
3348           each event within the diagnostic.  For example:
3349
3350                   test.c:29:5: error: passing NULL as argument 1 to 'PyList_Append' which requires a non-NULL parameter
3351                   test.c:25:10: note: (1) when 'PyList_New' fails, returning NULL
3352                   test.c:27:3: note: (2) when 'i < count'
3353                   test.c:29:5: note: (3) when calling 'PyList_Append', passing NULL from (1) as argument 1
3354
3355           inline-events means to print the events "inline" within the source
3356           code.  This view attempts to consolidate the events into runs of
3357           sufficiently-close events, printing them as labelled ranges within
3358           the source.
3359
3360           For example, the same events as above might be printed as:
3361
3362                     'test': events 1-3
3363                       |
3364                       |   25 |   list = PyList_New(0);
3365                       |      |          ^~~~~~~~~~~~~
3366                       |      |          |
3367                       |      |          (1) when 'PyList_New' fails, returning NULL
3368                       |   26 |
3369                       |   27 |   for (i = 0; i < count; i++) {
3370                       |      |   ~~~
3371                       |      |   |
3372                       |      |   (2) when 'i < count'
3373                       |   28 |     item = PyLong_FromLong(random());
3374                       |   29 |     PyList_Append(list, item);
3375                       |      |     ~~~~~~~~~~~~~~~~~~~~~~~~~
3376                       |      |     |
3377                       |      |     (3) when calling 'PyList_Append', passing NULL from (1) as argument 1
3378                       |
3379
3380           Interprocedural control flow is shown by grouping the events by
3381           stack frame, and using indentation to show how stack frames are
3382           nested, pushed, and popped.
3383
3384           For example:
3385
3386                     'test': events 1-2
3387                       |
3388                       |  133 | {
3389                       |      | ^
3390                       |      | |
3391                       |      | (1) entering 'test'
3392                       |  134 |   boxed_int *obj = make_boxed_int (i);
3393                       |      |                    ~~~~~~~~~~~~~~~~~~
3394                       |      |                    |
3395                       |      |                    (2) calling 'make_boxed_int'
3396                       |
3397                       +--> 'make_boxed_int': events 3-4
3398                              |
3399                              |  120 | {
3400                              |      | ^
3401                              |      | |
3402                              |      | (3) entering 'make_boxed_int'
3403                              |  121 |   boxed_int *result = (boxed_int *)wrapped_malloc (sizeof (boxed_int));
3404                              |      |                                    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3405                              |      |                                    |
3406                              |      |                                    (4) calling 'wrapped_malloc'
3407                              |
3408                              +--> 'wrapped_malloc': events 5-6
3409                                     |
3410                                     |    7 | {
3411                                     |      | ^
3412                                     |      | |
3413                                     |      | (5) entering 'wrapped_malloc'
3414                                     |    8 |   return malloc (size);
3415                                     |      |          ~~~~~~~~~~~~~
3416                                     |      |          |
3417                                     |      |          (6) calling 'malloc'
3418                                     |
3419                       <-------------+
3420                       |
3421                    'test': event 7
3422                       |
3423                       |  138 |   free_boxed_int (obj);
3424                       |      |   ^~~~~~~~~~~~~~~~~~~~
3425                       |      |   |
3426                       |      |   (7) calling 'free_boxed_int'
3427                       |
3428                   (etc)
3429
3430       -fdiagnostics-show-path-depths
3431           This option provides additional information when printing control-
3432           flow paths associated with a diagnostic.
3433
3434           If this is option is provided then the stack depth will be printed
3435           for each run of events within
3436           -fdiagnostics-path-format=separate-events.
3437
3438           This is intended for use by GCC developers and plugin developers
3439           when debugging diagnostics that report interprocedural control
3440           flow.
3441
3442       -fno-show-column
3443           Do not print column numbers in diagnostics.  This may be necessary
3444           if diagnostics are being scanned by a program that does not
3445           understand the column numbers, such as dejagnu.
3446
3447       -fdiagnostics-format=FORMAT
3448           Select a different format for printing diagnostics.  FORMAT is text
3449           or json.  The default is text.
3450
3451           The json format consists of a top-level JSON array containing JSON
3452           objects representing the diagnostics.
3453
3454           The JSON is emitted as one line, without formatting; the examples
3455           below have been formatted for clarity.
3456
3457           Diagnostics can have child diagnostics.  For example, this error
3458           and note:
3459
3460                   misleading-indentation.c:15:3: warning: this 'if' clause does not
3461                     guard... [-Wmisleading-indentation]
3462                      15 |   if (flag)
3463                         |   ^~
3464                   misleading-indentation.c:17:5: note: ...this statement, but the latter
3465                     is misleadingly indented as if it were guarded by the 'if'
3466                      17 |     y = 2;
3467                         |     ^
3468
3469           might be printed in JSON form (after formatting) like this:
3470
3471                   [
3472                       {
3473                           "kind": "warning",
3474                           "locations": [
3475                               {
3476                                   "caret": {
3477                                       "column": 3,
3478                                       "file": "misleading-indentation.c",
3479                                       "line": 15
3480                                   },
3481                                   "finish": {
3482                                       "column": 4,
3483                                       "file": "misleading-indentation.c",
3484                                       "line": 15
3485                                   }
3486                               }
3487                           ],
3488                           "message": "this \u2018if\u2019 clause does not guard...",
3489                           "option": "-Wmisleading-indentation",
3490                           "option_url": "https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html#index-Wmisleading-indentation",
3491                           "children": [
3492                               {
3493                                   "kind": "note",
3494                                   "locations": [
3495                                       {
3496                                           "caret": {
3497                                               "column": 5,
3498                                               "file": "misleading-indentation.c",
3499                                               "line": 17
3500                                           }
3501                                       }
3502                                   ],
3503                                   "message": "...this statement, but the latter is ..."
3504                               }
3505                           ]
3506                       },
3507                       ...
3508                   ]
3509
3510           where the "note" is a child of the "warning".
3511
3512           A diagnostic has a "kind".  If this is "warning", then there is an
3513           "option" key describing the command-line option controlling the
3514           warning.
3515
3516           A diagnostic can contain zero or more locations.  Each location has
3517           up to three positions within it: a "caret" position and optional
3518           "start" and "finish" positions.  A location can also have an
3519           optional "label" string.  For example, this error:
3520
3521                   bad-binary-ops.c:64:23: error: invalid operands to binary + (have 'S' {aka
3522                      'struct s'} and 'T' {aka 'struct t'})
3523                      64 |   return callee_4a () + callee_4b ();
3524                         |          ~~~~~~~~~~~~ ^ ~~~~~~~~~~~~
3525                         |          |              |
3526                         |          |              T {aka struct t}
3527                         |          S {aka struct s}
3528
3529           has three locations.  Its primary location is at the "+" token at
3530           column 23.  It has two secondary locations, describing the left and
3531           right-hand sides of the expression, which have labels.  It might be
3532           printed in JSON form as:
3533
3534                       {
3535                           "children": [],
3536                           "kind": "error",
3537                           "locations": [
3538                               {
3539                                   "caret": {
3540                                       "column": 23, "file": "bad-binary-ops.c", "line": 64
3541                                   }
3542                               },
3543                               {
3544                                   "caret": {
3545                                       "column": 10, "file": "bad-binary-ops.c", "line": 64
3546                                   },
3547                                   "finish": {
3548                                       "column": 21, "file": "bad-binary-ops.c", "line": 64
3549                                   },
3550                                   "label": "S {aka struct s}"
3551                               },
3552                               {
3553                                   "caret": {
3554                                       "column": 25, "file": "bad-binary-ops.c", "line": 64
3555                                   },
3556                                   "finish": {
3557                                       "column": 36, "file": "bad-binary-ops.c", "line": 64
3558                                   },
3559                                   "label": "T {aka struct t}"
3560                               }
3561                           ],
3562                           "message": "invalid operands to binary + ..."
3563                       }
3564
3565           If a diagnostic contains fix-it hints, it has a "fixits" array,
3566           consisting of half-open intervals, similar to the output of
3567           -fdiagnostics-parseable-fixits.  For example, this diagnostic with
3568           a replacement fix-it hint:
3569
3570                   demo.c:8:15: error: 'struct s' has no member named 'colour'; did you
3571                     mean 'color'?
3572                       8 |   return ptr->colour;
3573                         |               ^~~~~~
3574                         |               color
3575
3576           might be printed in JSON form as:
3577
3578                       {
3579                           "children": [],
3580                           "fixits": [
3581                               {
3582                                   "next": {
3583                                       "column": 21,
3584                                       "file": "demo.c",
3585                                       "line": 8
3586                                   },
3587                                   "start": {
3588                                       "column": 15,
3589                                       "file": "demo.c",
3590                                       "line": 8
3591                                   },
3592                                   "string": "color"
3593                               }
3594                           ],
3595                           "kind": "error",
3596                           "locations": [
3597                               {
3598                                   "caret": {
3599                                       "column": 15,
3600                                       "file": "demo.c",
3601                                       "line": 8
3602                                   },
3603                                   "finish": {
3604                                       "column": 20,
3605                                       "file": "demo.c",
3606                                       "line": 8
3607                                   }
3608                               }
3609                           ],
3610                           "message": "\u2018struct s\u2019 has no member named ..."
3611                       }
3612
3613           where the fix-it hint suggests replacing the text from "start" up
3614           to but not including "next" with "string"'s value.  Deletions are
3615           expressed via an empty value for "string", insertions by having
3616           "start" equal "next".
3617
3618           If the diagnostic has a path of control-flow events associated with
3619           it, it has a "path" array of objects representing the events.  Each
3620           event object has a "description" string, a "location" object, along
3621           with a "function" string and a "depth" number for representing
3622           interprocedural paths.  The "function" represents the current
3623           function at that event, and the "depth" represents the stack depth
3624           relative to some baseline: the higher, the more frames are within
3625           the stack.
3626
3627           For example, the intraprocedural example shown for
3628           -fdiagnostics-path-format= might have this JSON for its path:
3629
3630                       "path": [
3631                           {
3632                               "depth": 0,
3633                               "description": "when 'PyList_New' fails, returning NULL",
3634                               "function": "test",
3635                               "location": {
3636                                   "column": 10,
3637                                   "file": "test.c",
3638                                   "line": 25
3639                               }
3640                           },
3641                           {
3642                               "depth": 0,
3643                               "description": "when 'i < count'",
3644                               "function": "test",
3645                               "location": {
3646                                   "column": 3,
3647                                   "file": "test.c",
3648                                   "line": 27
3649                               }
3650                           },
3651                           {
3652                               "depth": 0,
3653                               "description": "when calling 'PyList_Append', passing NULL from (1) as argument 1",
3654                               "function": "test",
3655                               "location": {
3656                                   "column": 5,
3657                                   "file": "test.c",
3658                                   "line": 29
3659                               }
3660                           }
3661                       ]
3662
3663   Options to Request or Suppress Warnings
3664       Warnings are diagnostic messages that report constructions that are not
3665       inherently erroneous but that are risky or suggest there may have been
3666       an error.
3667
3668       The following language-independent options do not enable specific
3669       warnings but control the kinds of diagnostics produced by GCC.
3670
3671       -fsyntax-only
3672           Check the code for syntax errors, but don't do anything beyond
3673           that.
3674
3675       -fmax-errors=n
3676           Limits the maximum number of error messages to n, at which point
3677           GCC bails out rather than attempting to continue processing the
3678           source code.  If n is 0 (the default), there is no limit on the
3679           number of error messages produced.  If -Wfatal-errors is also
3680           specified, then -Wfatal-errors takes precedence over this option.
3681
3682       -w  Inhibit all warning messages.
3683
3684       -Werror
3685           Make all warnings into errors.
3686
3687       -Werror=
3688           Make the specified warning into an error.  The specifier for a
3689           warning is appended; for example -Werror=switch turns the warnings
3690           controlled by -Wswitch into errors.  This switch takes a negative
3691           form, to be used to negate -Werror for specific warnings; for
3692           example -Wno-error=switch makes -Wswitch warnings not be errors,
3693           even when -Werror is in effect.
3694
3695           The warning message for each controllable warning includes the
3696           option that controls the warning.  That option can then be used
3697           with -Werror= and -Wno-error= as described above.  (Printing of the
3698           option in the warning message can be disabled using the
3699           -fno-diagnostics-show-option flag.)
3700
3701           Note that specifying -Werror=foo automatically implies -Wfoo.
3702           However, -Wno-error=foo does not imply anything.
3703
3704       -Wfatal-errors
3705           This option causes the compiler to abort compilation on the first
3706           error occurred rather than trying to keep going and printing
3707           further error messages.
3708
3709       You can request many specific warnings with options beginning with -W,
3710       for example -Wimplicit to request warnings on implicit declarations.
3711       Each of these specific warning options also has a negative form
3712       beginning -Wno- to turn off warnings; for example, -Wno-implicit.  This
3713       manual lists only one of the two forms, whichever is not the default.
3714       For further language-specific options also refer to C++ Dialect Options
3715       and Objective-C and Objective-C++ Dialect Options.  Additional warnings
3716       can be produced by enabling the static analyzer;
3717
3718       Some options, such as -Wall and -Wextra, turn on other options, such as
3719       -Wunused, which may turn on further options, such as -Wunused-value.
3720       The combined effect of positive and negative forms is that more
3721       specific options have priority over less specific ones, independently
3722       of their position in the command-line. For options of the same
3723       specificity, the last one takes effect. Options enabled or disabled via
3724       pragmas take effect as if they appeared at the end of the command-line.
3725
3726       When an unrecognized warning option is requested (e.g.,
3727       -Wunknown-warning), GCC emits a diagnostic stating that the option is
3728       not recognized.  However, if the -Wno- form is used, the behavior is
3729       slightly different: no diagnostic is produced for -Wno-unknown-warning
3730       unless other diagnostics are being produced.  This allows the use of
3731       new -Wno- options with old compilers, but if something goes wrong, the
3732       compiler warns that an unrecognized option is present.
3733
3734       The effectiveness of some warnings depends on optimizations also being
3735       enabled. For example -Wsuggest-final-types is more effective with link-
3736       time optimization and -Wmaybe-uninitialized does not warn at all unless
3737       optimization is enabled.
3738
3739       -Wpedantic
3740       -pedantic
3741           Issue all the warnings demanded by strict ISO C and ISO C++; reject
3742           all programs that use forbidden extensions, and some other programs
3743           that do not follow ISO C and ISO C++.  For ISO C, follows the
3744           version of the ISO C standard specified by any -std option used.
3745
3746           Valid ISO C and ISO C++ programs should compile properly with or
3747           without this option (though a rare few require -ansi or a -std
3748           option specifying the required version of ISO C).  However, without
3749           this option, certain GNU extensions and traditional C and C++
3750           features are supported as well.  With this option, they are
3751           rejected.
3752
3753           -Wpedantic does not cause warning messages for use of the alternate
3754           keywords whose names begin and end with __.  This alternate format
3755           can also be used to disable warnings for non-ISO __intN types, i.e.
3756           __intN__.  Pedantic warnings are also disabled in the expression
3757           that follows "__extension__".  However, only system header files
3758           should use these escape routes; application programs should avoid
3759           them.
3760
3761           Some users try to use -Wpedantic to check programs for strict ISO C
3762           conformance.  They soon find that it does not do quite what they
3763           want: it finds some non-ISO practices, but not all---only those for
3764           which ISO C requires a diagnostic, and some others for which
3765           diagnostics have been added.
3766
3767           A feature to report any failure to conform to ISO C might be useful
3768           in some instances, but would require considerable additional work
3769           and would be quite different from -Wpedantic.  We don't have plans
3770           to support such a feature in the near future.
3771
3772           Where the standard specified with -std represents a GNU extended
3773           dialect of C, such as gnu90 or gnu99, there is a corresponding base
3774           standard, the version of ISO C on which the GNU extended dialect is
3775           based.  Warnings from -Wpedantic are given where they are required
3776           by the base standard.  (It does not make sense for such warnings to
3777           be given only for features not in the specified GNU C dialect,
3778           since by definition the GNU dialects of C include all features the
3779           compiler supports with the given option, and there would be nothing
3780           to warn about.)
3781
3782       -pedantic-errors
3783           Give an error whenever the base standard (see -Wpedantic) requires
3784           a diagnostic, in some cases where there is undefined behavior at
3785           compile-time and in some other cases that do not prevent
3786           compilation of programs that are valid according to the standard.
3787           This is not equivalent to -Werror=pedantic, since there are errors
3788           enabled by this option and not enabled by the latter and vice
3789           versa.
3790
3791       -Wall
3792           This enables all the warnings about constructions that some users
3793           consider questionable, and that are easy to avoid (or modify to
3794           prevent the warning), even in conjunction with macros.  This also
3795           enables some language-specific warnings described in C++ Dialect
3796           Options and Objective-C and Objective-C++ Dialect Options.
3797
3798           -Wall turns on the following warning flags:
3799
3800           -Waddress -Warray-bounds=1 (only with -O2) -Wbool-compare
3801           -Wbool-operation -Wc++11-compat  -Wc++14-compat -Wcatch-value (C++
3802           and Objective-C++ only) -Wchar-subscripts -Wcomment
3803           -Wduplicate-decl-specifier (C and Objective-C only) -Wenum-compare
3804           (in C/ObjC; this is on by default in C++) -Wenum-conversion in
3805           C/ObjC; -Wformat -Wformat-overflow -Wformat-truncation
3806           -Wint-in-bool-context -Wimplicit (C and Objective-C only)
3807           -Wimplicit-int (C and Objective-C only)
3808           -Wimplicit-function-declaration (C and Objective-C only)
3809           -Winit-self (only for C++) -Wlogical-not-parentheses -Wmain (only
3810           for C/ObjC and unless -ffreestanding) -Wmaybe-uninitialized
3811           -Wmemset-elt-size -Wmemset-transposed-args -Wmisleading-indentation
3812           (only for C/C++) -Wmissing-attributes -Wmissing-braces (only for
3813           C/ObjC) -Wmultistatement-macros -Wnarrowing (only for C++)
3814           -Wnonnull -Wnonnull-compare -Wopenmp-simd -Wparentheses
3815           -Wpessimizing-move (only for C++) -Wpointer-sign -Wreorder
3816           -Wrestrict -Wreturn-type -Wsequence-point -Wsign-compare (only in
3817           C++) -Wsizeof-pointer-div -Wsizeof-pointer-memaccess
3818           -Wstrict-aliasing -Wstrict-overflow=1 -Wswitch
3819           -Wtautological-compare -Wtrigraphs -Wuninitialized
3820           -Wunknown-pragmas -Wunused-function -Wunused-label -Wunused-value
3821           -Wunused-variable -Wvolatile-register-var -Wzero-length-bounds
3822
3823           Note that some warning flags are not implied by -Wall.  Some of
3824           them warn about constructions that users generally do not consider
3825           questionable, but which occasionally you might wish to check for;
3826           others warn about constructions that are necessary or hard to avoid
3827           in some cases, and there is no simple way to modify the code to
3828           suppress the warning. Some of them are enabled by -Wextra but many
3829           of them must be enabled individually.
3830
3831       -Wextra
3832           This enables some extra warning flags that are not enabled by
3833           -Wall. (This option used to be called -W.  The older name is still
3834           supported, but the newer name is more descriptive.)
3835
3836           -Wclobbered -Wcast-function-type -Wdeprecated-copy (C++ only)
3837           -Wempty-body -Wignored-qualifiers -Wimplicit-fallthrough=3
3838           -Wmissing-field-initializers -Wmissing-parameter-type (C only)
3839           -Wold-style-declaration (C only) -Woverride-init -Wsign-compare (C
3840           only) -Wstring-compare -Wredundant-move (only for C++)
3841           -Wtype-limits -Wuninitialized -Wshift-negative-value (in C++03 and
3842           in C99 and newer) -Wunused-parameter (only with -Wunused or -Wall)
3843           -Wunused-but-set-parameter (only with -Wunused or -Wall)
3844
3845           The option -Wextra also prints warning messages for the following
3846           cases:
3847
3848           *   A pointer is compared against integer zero with "<", "<=", ">",
3849               or ">=".
3850
3851           *   (C++ only) An enumerator and a non-enumerator both appear in a
3852               conditional expression.
3853
3854           *   (C++ only) Ambiguous virtual bases.
3855
3856           *   (C++ only) Subscripting an array that has been declared
3857               "register".
3858
3859           *   (C++ only) Taking the address of a variable that has been
3860               declared "register".
3861
3862           *   (C++ only) A base class is not initialized in the copy
3863               constructor of a derived class.
3864
3865       -Wabi (C, Objective-C, C++ and Objective-C++ only)
3866           Warn about code affected by ABI changes.  This includes code that
3867           may not be compatible with the vendor-neutral C++ ABI as well as
3868           the psABI for the particular target.
3869
3870           Since G++ now defaults to updating the ABI with each major release,
3871           normally -Wabi warns only about C++ ABI compatibility problems if
3872           there is a check added later in a release series for an ABI issue
3873           discovered since the initial release.  -Wabi warns about more
3874           things if an older ABI version is selected (with -fabi-version=n).
3875
3876           -Wabi can also be used with an explicit version number to warn
3877           about C++ ABI compatibility with a particular -fabi-version level,
3878           e.g. -Wabi=2 to warn about changes relative to -fabi-version=2.
3879
3880           If an explicit version number is provided and -fabi-compat-version
3881           is not specified, the version number from this option is used for
3882           compatibility aliases.  If no explicit version number is provided
3883           with this option, but -fabi-compat-version is specified, that
3884           version number is used for C++ ABI warnings.
3885
3886           Although an effort has been made to warn about all such cases,
3887           there are probably some cases that are not warned about, even
3888           though G++ is generating incompatible code.  There may also be
3889           cases where warnings are emitted even though the code that is
3890           generated is compatible.
3891
3892           You should rewrite your code to avoid these warnings if you are
3893           concerned about the fact that code generated by G++ may not be
3894           binary compatible with code generated by other compilers.
3895
3896           Known incompatibilities in -fabi-version=2 (which was the default
3897           from GCC 3.4 to 4.9) include:
3898
3899           *   A template with a non-type template parameter of reference type
3900               was mangled incorrectly:
3901
3902                       extern int N;
3903                       template <int &> struct S {};
3904                       void n (S<N>) {2}
3905
3906               This was fixed in -fabi-version=3.
3907
3908           *   SIMD vector types declared using "__attribute ((vector_size))"
3909               were mangled in a non-standard way that does not allow for
3910               overloading of functions taking vectors of different sizes.
3911
3912               The mangling was changed in -fabi-version=4.
3913
3914           *   "__attribute ((const))" and "noreturn" were mangled as type
3915               qualifiers, and "decltype" of a plain declaration was folded
3916               away.
3917
3918               These mangling issues were fixed in -fabi-version=5.
3919
3920           *   Scoped enumerators passed as arguments to a variadic function
3921               are promoted like unscoped enumerators, causing "va_arg" to
3922               complain.  On most targets this does not actually affect the
3923               parameter passing ABI, as there is no way to pass an argument
3924               smaller than "int".
3925
3926               Also, the ABI changed the mangling of template argument packs,
3927               "const_cast", "static_cast", prefix increment/decrement, and a
3928               class scope function used as a template argument.
3929
3930               These issues were corrected in -fabi-version=6.
3931
3932           *   Lambdas in default argument scope were mangled incorrectly, and
3933               the ABI changed the mangling of "nullptr_t".
3934
3935               These issues were corrected in -fabi-version=7.
3936
3937           *   When mangling a function type with function-cv-qualifiers, the
3938               un-qualified function type was incorrectly treated as a
3939               substitution candidate.
3940
3941               This was fixed in -fabi-version=8, the default for GCC 5.1.
3942
3943           *   "decltype(nullptr)" incorrectly had an alignment of 1, leading
3944               to unaligned accesses.  Note that this did not affect the ABI
3945               of a function with a "nullptr_t" parameter, as parameters have
3946               a minimum alignment.
3947
3948               This was fixed in -fabi-version=9, the default for GCC 5.2.
3949
3950           *   Target-specific attributes that affect the identity of a type,
3951               such as ia32 calling conventions on a function type (stdcall,
3952               regparm, etc.), did not affect the mangled name, leading to
3953               name collisions when function pointers were used as template
3954               arguments.
3955
3956               This was fixed in -fabi-version=10, the default for GCC 6.1.
3957
3958           This option also enables warnings about psABI-related changes.  The
3959           known psABI changes at this point include:
3960
3961           *   For SysV/x86-64, unions with "long double" members are passed
3962               in memory as specified in psABI.  Prior to GCC 4.4, this was
3963               not the case.  For example:
3964
3965                       union U {
3966                         long double ld;
3967                         int i;
3968                       };
3969
3970               "union U" is now always passed in memory.
3971
3972       -Wchar-subscripts
3973           Warn if an array subscript has type "char".  This is a common cause
3974           of error, as programmers often forget that this type is signed on
3975           some machines.  This warning is enabled by -Wall.
3976
3977       -Wno-coverage-mismatch
3978           Warn if feedback profiles do not match when using the -fprofile-use
3979           option.  If a source file is changed between compiling with
3980           -fprofile-generate and with -fprofile-use, the files with the
3981           profile feedback can fail to match the source file and GCC cannot
3982           use the profile feedback information.  By default, this warning is
3983           enabled and is treated as an error.  -Wno-coverage-mismatch can be
3984           used to disable the warning or -Wno-error=coverage-mismatch can be
3985           used to disable the error.  Disabling the error for this warning
3986           can result in poorly optimized code and is useful only in the case
3987           of very minor changes such as bug fixes to an existing code-base.
3988           Completely disabling the warning is not recommended.
3989
3990       -Wno-cpp
3991           (C, Objective-C, C++, Objective-C++ and Fortran only) Suppress
3992           warning messages emitted by "#warning" directives.
3993
3994       -Wdouble-promotion (C, C++, Objective-C and Objective-C++ only)
3995           Give a warning when a value of type "float" is implicitly promoted
3996           to "double".  CPUs with a 32-bit "single-precision" floating-point
3997           unit implement "float" in hardware, but emulate "double" in
3998           software.  On such a machine, doing computations using "double"
3999           values is much more expensive because of the overhead required for
4000           software emulation.
4001
4002           It is easy to accidentally do computations with "double" because
4003           floating-point literals are implicitly of type "double".  For
4004           example, in:
4005
4006                   float area(float radius)
4007                   {
4008                      return 3.14159 * radius * radius;
4009                   }
4010
4011           the compiler performs the entire computation with "double" because
4012           the floating-point literal is a "double".
4013
4014       -Wduplicate-decl-specifier (C and Objective-C only)
4015           Warn if a declaration has duplicate "const", "volatile", "restrict"
4016           or "_Atomic" specifier.  This warning is enabled by -Wall.
4017
4018       -Wformat
4019       -Wformat=n
4020           Check calls to "printf" and "scanf", etc., to make sure that the
4021           arguments supplied have types appropriate to the format string
4022           specified, and that the conversions specified in the format string
4023           make sense.  This includes standard functions, and others specified
4024           by format attributes, in the "printf", "scanf", "strftime" and
4025           "strfmon" (an X/Open extension, not in the C standard) families (or
4026           other target-specific families).  Which functions are checked
4027           without format attributes having been specified depends on the
4028           standard version selected, and such checks of functions without the
4029           attribute specified are disabled by -ffreestanding or -fno-builtin.
4030
4031           The formats are checked against the format features supported by
4032           GNU libc version 2.2.  These include all ISO C90 and C99 features,
4033           as well as features from the Single Unix Specification and some BSD
4034           and GNU extensions.  Other library implementations may not support
4035           all these features; GCC does not support warning about features
4036           that go beyond a particular library's limitations.  However, if
4037           -Wpedantic is used with -Wformat, warnings are given about format
4038           features not in the selected standard version (but not for
4039           "strfmon" formats, since those are not in any version of the C
4040           standard).
4041
4042           -Wformat=1
4043           -Wformat
4044               Option -Wformat is equivalent to -Wformat=1, and -Wno-format is
4045               equivalent to -Wformat=0.  Since -Wformat also checks for null
4046               format arguments for several functions, -Wformat also implies
4047               -Wnonnull.  Some aspects of this level of format checking can
4048               be disabled by the options: -Wno-format-contains-nul,
4049               -Wno-format-extra-args, and -Wno-format-zero-length.  -Wformat
4050               is enabled by -Wall.
4051
4052           -Wformat=2
4053               Enable -Wformat plus additional format checks.  Currently
4054               equivalent to -Wformat -Wformat-nonliteral -Wformat-security
4055               -Wformat-y2k.
4056
4057       -Wno-format-contains-nul
4058           If -Wformat is specified, do not warn about format strings that
4059           contain NUL bytes.
4060
4061       -Wno-format-extra-args
4062           If -Wformat is specified, do not warn about excess arguments to a
4063           "printf" or "scanf" format function.  The C standard specifies that
4064           such arguments are ignored.
4065
4066           Where the unused arguments lie between used arguments that are
4067           specified with $ operand number specifications, normally warnings
4068           are still given, since the implementation could not know what type
4069           to pass to "va_arg" to skip the unused arguments.  However, in the
4070           case of "scanf" formats, this option suppresses the warning if the
4071           unused arguments are all pointers, since the Single Unix
4072           Specification says that such unused arguments are allowed.
4073
4074       -Wformat-overflow
4075       -Wformat-overflow=level
4076           Warn about calls to formatted input/output functions such as
4077           "sprintf" and "vsprintf" that might overflow the destination
4078           buffer.  When the exact number of bytes written by a format
4079           directive cannot be determined at compile-time it is estimated
4080           based on heuristics that depend on the level argument and on
4081           optimization.  While enabling optimization will in most cases
4082           improve the accuracy of the warning, it may also result in false
4083           positives.
4084
4085           -Wformat-overflow
4086           -Wformat-overflow=1
4087               Level 1 of -Wformat-overflow enabled by -Wformat employs a
4088               conservative approach that warns only about calls that most
4089               likely overflow the buffer.  At this level, numeric arguments
4090               to format directives with unknown values are assumed to have
4091               the value of one, and strings of unknown length to be empty.
4092               Numeric arguments that are known to be bounded to a subrange of
4093               their type, or string arguments whose output is bounded either
4094               by their directive's precision or by a finite set of string
4095               literals, are assumed to take on the value within the range
4096               that results in the most bytes on output.  For example, the
4097               call to "sprintf" below is diagnosed because even with both a
4098               and b equal to zero, the terminating NUL character ('\0')
4099               appended by the function to the destination buffer will be
4100               written past its end.  Increasing the size of the buffer by a
4101               single byte is sufficient to avoid the warning, though it may
4102               not be sufficient to avoid the overflow.
4103
4104                       void f (int a, int b)
4105                       {
4106                         char buf [13];
4107                         sprintf (buf, "a = %i, b = %i\n", a, b);
4108                       }
4109
4110           -Wformat-overflow=2
4111               Level 2 warns also about calls that might overflow the
4112               destination buffer given an argument of sufficient length or
4113               magnitude.  At level 2, unknown numeric arguments are assumed
4114               to have the minimum representable value for signed types with a
4115               precision greater than 1, and the maximum representable value
4116               otherwise.  Unknown string arguments whose length cannot be
4117               assumed to be bounded either by the directive's precision, or
4118               by a finite set of string literals they may evaluate to, or the
4119               character array they may point to, are assumed to be 1
4120               character long.
4121
4122               At level 2, the call in the example above is again diagnosed,
4123               but this time because with a equal to a 32-bit "INT_MIN" the
4124               first %i directive will write some of its digits beyond the end
4125               of the destination buffer.  To make the call safe regardless of
4126               the values of the two variables, the size of the destination
4127               buffer must be increased to at least 34 bytes.  GCC includes
4128               the minimum size of the buffer in an informational note
4129               following the warning.
4130
4131               An alternative to increasing the size of the destination buffer
4132               is to constrain the range of formatted values.  The maximum
4133               length of string arguments can be bounded by specifying the
4134               precision in the format directive.  When numeric arguments of
4135               format directives can be assumed to be bounded by less than the
4136               precision of their type, choosing an appropriate length
4137               modifier to the format specifier will reduce the required
4138               buffer size.  For example, if a and b in the example above can
4139               be assumed to be within the precision of the "short int" type
4140               then using either the %hi format directive or casting the
4141               argument to "short" reduces the maximum required size of the
4142               buffer to 24 bytes.
4143
4144                       void f (int a, int b)
4145                       {
4146                         char buf [23];
4147                         sprintf (buf, "a = %hi, b = %i\n", a, (short)b);
4148                       }
4149
4150       -Wno-format-zero-length
4151           If -Wformat is specified, do not warn about zero-length formats.
4152           The C standard specifies that zero-length formats are allowed.
4153
4154       -Wformat-nonliteral
4155           If -Wformat is specified, also warn if the format string is not a
4156           string literal and so cannot be checked, unless the format function
4157           takes its format arguments as a "va_list".
4158
4159       -Wformat-security
4160           If -Wformat is specified, also warn about uses of format functions
4161           that represent possible security problems.  At present, this warns
4162           about calls to "printf" and "scanf" functions where the format
4163           string is not a string literal and there are no format arguments,
4164           as in "printf (foo);".  This may be a security hole if the format
4165           string came from untrusted input and contains %n.  (This is
4166           currently a subset of what -Wformat-nonliteral warns about, but in
4167           future warnings may be added to -Wformat-security that are not
4168           included in -Wformat-nonliteral.)
4169
4170       -Wformat-signedness
4171           If -Wformat is specified, also warn if the format string requires
4172           an unsigned argument and the argument is signed and vice versa.
4173
4174       -Wformat-truncation
4175       -Wformat-truncation=level
4176           Warn about calls to formatted input/output functions such as
4177           "snprintf" and "vsnprintf" that might result in output truncation.
4178           When the exact number of bytes written by a format directive cannot
4179           be determined at compile-time it is estimated based on heuristics
4180           that depend on the level argument and on optimization.  While
4181           enabling optimization will in most cases improve the accuracy of
4182           the warning, it may also result in false positives.  Except as
4183           noted otherwise, the option uses the same logic -Wformat-overflow.
4184
4185           -Wformat-truncation
4186           -Wformat-truncation=1
4187               Level 1 of -Wformat-truncation enabled by -Wformat employs a
4188               conservative approach that warns only about calls to bounded
4189               functions whose return value is unused and that will most
4190               likely result in output truncation.
4191
4192           -Wformat-truncation=2
4193               Level 2 warns also about calls to bounded functions whose
4194               return value is used and that might result in truncation given
4195               an argument of sufficient length or magnitude.
4196
4197       -Wformat-y2k
4198           If -Wformat is specified, also warn about "strftime" formats that
4199           may yield only a two-digit year.
4200
4201       -Wnonnull
4202           Warn about passing a null pointer for arguments marked as requiring
4203           a non-null value by the "nonnull" function attribute.
4204
4205           -Wnonnull is included in -Wall and -Wformat.  It can be disabled
4206           with the -Wno-nonnull option.
4207
4208       -Wnonnull-compare
4209           Warn when comparing an argument marked with the "nonnull" function
4210           attribute against null inside the function.
4211
4212           -Wnonnull-compare is included in -Wall.  It can be disabled with
4213           the -Wno-nonnull-compare option.
4214
4215       -Wnull-dereference
4216           Warn if the compiler detects paths that trigger erroneous or
4217           undefined behavior due to dereferencing a null pointer.  This
4218           option is only active when -fdelete-null-pointer-checks is active,
4219           which is enabled by optimizations in most targets.  The precision
4220           of the warnings depends on the optimization options used.
4221
4222       -Winit-self (C, C++, Objective-C and Objective-C++ only)
4223           Warn about uninitialized variables that are initialized with
4224           themselves.  Note this option can only be used with the
4225           -Wuninitialized option.
4226
4227           For example, GCC warns about "i" being uninitialized in the
4228           following snippet only when -Winit-self has been specified:
4229
4230                   int f()
4231                   {
4232                     int i = i;
4233                     return i;
4234                   }
4235
4236           This warning is enabled by -Wall in C++.
4237
4238       -Wno-implicit-int (C and Objective-C only)
4239           This option controls warnings when a declaration does not specify a
4240           type.  This warning is enabled by default in C99 and later dialects
4241           of C, and also by -Wall.
4242
4243       -Wno-implicit-function-declaration (C and Objective-C only)
4244           This option controls warnings when a function is used before being
4245           declared.  This warning is enabled by default in C99 and later
4246           dialects of C, and also by -Wall.  The warning is made into an
4247           error by -pedantic-errors.
4248
4249       -Wimplicit (C and Objective-C only)
4250           Same as -Wimplicit-int and -Wimplicit-function-declaration.  This
4251           warning is enabled by -Wall.
4252
4253       -Wimplicit-fallthrough
4254           -Wimplicit-fallthrough is the same as -Wimplicit-fallthrough=3 and
4255           -Wno-implicit-fallthrough is the same as -Wimplicit-fallthrough=0.
4256
4257       -Wimplicit-fallthrough=n
4258           Warn when a switch case falls through.  For example:
4259
4260                   switch (cond)
4261                     {
4262                     case 1:
4263                       a = 1;
4264                       break;
4265                     case 2:
4266                       a = 2;
4267                     case 3:
4268                       a = 3;
4269                       break;
4270                     }
4271
4272           This warning does not warn when the last statement of a case cannot
4273           fall through, e.g. when there is a return statement or a call to
4274           function declared with the noreturn attribute.
4275           -Wimplicit-fallthrough= also takes into account control flow
4276           statements, such as ifs, and only warns when appropriate.  E.g.
4277
4278                   switch (cond)
4279                     {
4280                     case 1:
4281                       if (i > 3) {
4282                         bar (5);
4283                         break;
4284                       } else if (i < 1) {
4285                         bar (0);
4286                       } else
4287                         return;
4288                     default:
4289                       ...
4290                     }
4291
4292           Since there are occasions where a switch case fall through is
4293           desirable, GCC provides an attribute, "__attribute__
4294           ((fallthrough))", that is to be used along with a null statement to
4295           suppress this warning that would normally occur:
4296
4297                   switch (cond)
4298                     {
4299                     case 1:
4300                       bar (0);
4301                       __attribute__ ((fallthrough));
4302                     default:
4303                       ...
4304                     }
4305
4306           C++17 provides a standard way to suppress the
4307           -Wimplicit-fallthrough warning using "[[fallthrough]];" instead of
4308           the GNU attribute.  In C++11 or C++14 users can use
4309           "[[gnu::fallthrough]];", which is a GNU extension.  Instead of
4310           these attributes, it is also possible to add a fallthrough comment
4311           to silence the warning.  The whole body of the C or C++ style
4312           comment should match the given regular expressions listed below.
4313           The option argument n specifies what kind of comments are accepted:
4314
4315           *<-Wimplicit-fallthrough=0 disables the warning altogether.>
4316           *<-Wimplicit-fallthrough=1 matches ".*" regular>
4317               expression, any comment is used as fallthrough comment.
4318
4319           *<-Wimplicit-fallthrough=2 case insensitively matches>
4320               ".*falls?[ \t-]*thr(ough|u).*" regular expression.
4321
4322           *<-Wimplicit-fallthrough=3 case sensitively matches one of the>
4323               following regular expressions:
4324
4325               *<"-fallthrough">
4326               *<"@fallthrough@">
4327               *<"lint -fallthrough[ \t]*">
4328               *<"[ \t.!]*(ELSE,? |INTENTIONAL(LY)? )?FALL(S |
4329               |-)?THR(OUGH|U)[ \t.!]*(-[^\n\r]*)?">
4330               *<"[ \t.!]*(Else,? |Intentional(ly)? )?Fall((s |
4331               |-)[Tt]|t)hr(ough|u)[ \t.!]*(-[^\n\r]*)?">
4332               *<"[ \t.!]*([Ee]lse,? |[Ii]ntentional(ly)? )?fall(s |
4333               |-)?thr(ough|u)[ \t.!]*(-[^\n\r]*)?">
4334           *<-Wimplicit-fallthrough=4 case sensitively matches one of the>
4335               following regular expressions:
4336
4337               *<"-fallthrough">
4338               *<"@fallthrough@">
4339               *<"lint -fallthrough[ \t]*">
4340               *<"[ \t]*FALLTHR(OUGH|U)[ \t]*">
4341           *<-Wimplicit-fallthrough=5 doesn't recognize any comments as>
4342               fallthrough comments, only attributes disable the warning.
4343
4344           The comment needs to be followed after optional whitespace and
4345           other comments by "case" or "default" keywords or by a user label
4346           that precedes some "case" or "default" label.
4347
4348                   switch (cond)
4349                     {
4350                     case 1:
4351                       bar (0);
4352                       /* FALLTHRU */
4353                     default:
4354                       ...
4355                     }
4356
4357           The -Wimplicit-fallthrough=3 warning is enabled by -Wextra.
4358
4359       -Wno-if-not-aligned (C, C++, Objective-C and Objective-C++ only)
4360           Control if warnings triggered by the "warn_if_not_aligned"
4361           attribute should be issued.  These warnings are enabled by default.
4362
4363       -Wignored-qualifiers (C and C++ only)
4364           Warn if the return type of a function has a type qualifier such as
4365           "const".  For ISO C such a type qualifier has no effect, since the
4366           value returned by a function is not an lvalue.  For C++, the
4367           warning is only emitted for scalar types or "void".  ISO C
4368           prohibits qualified "void" return types on function definitions, so
4369           such return types always receive a warning even without this
4370           option.
4371
4372           This warning is also enabled by -Wextra.
4373
4374       -Wno-ignored-attributes (C and C++ only)
4375           This option controls warnings when an attribute is ignored.  This
4376           is different from the -Wattributes option in that it warns whenever
4377           the compiler decides to drop an attribute, not that the attribute
4378           is either unknown, used in a wrong place, etc.  This warning is
4379           enabled by default.
4380
4381       -Wmain
4382           Warn if the type of "main" is suspicious.  "main" should be a
4383           function with external linkage, returning int, taking either zero
4384           arguments, two, or three arguments of appropriate types.  This
4385           warning is enabled by default in C++ and is enabled by either -Wall
4386           or -Wpedantic.
4387
4388       -Wmisleading-indentation (C and C++ only)
4389           Warn when the indentation of the code does not reflect the block
4390           structure.  Specifically, a warning is issued for "if", "else",
4391           "while", and "for" clauses with a guarded statement that does not
4392           use braces, followed by an unguarded statement with the same
4393           indentation.
4394
4395           In the following example, the call to "bar" is misleadingly
4396           indented as if it were guarded by the "if" conditional.
4397
4398                     if (some_condition ())
4399                       foo ();
4400                       bar ();  /* Gotcha: this is not guarded by the "if".  */
4401
4402           In the case of mixed tabs and spaces, the warning uses the
4403           -ftabstop= option to determine if the statements line up
4404           (defaulting to 8).
4405
4406           The warning is not issued for code involving multiline preprocessor
4407           logic such as the following example.
4408
4409                     if (flagA)
4410                       foo (0);
4411                   #if SOME_CONDITION_THAT_DOES_NOT_HOLD
4412                     if (flagB)
4413                   #endif
4414                       foo (1);
4415
4416           The warning is not issued after a "#line" directive, since this
4417           typically indicates autogenerated code, and no assumptions can be
4418           made about the layout of the file that the directive references.
4419
4420           This warning is enabled by -Wall in C and C++.
4421
4422       -Wmissing-attributes
4423           Warn when a declaration of a function is missing one or more
4424           attributes that a related function is declared with and whose
4425           absence may adversely affect the correctness or efficiency of
4426           generated code.  For example, the warning is issued for
4427           declarations of aliases that use attributes to specify less
4428           restrictive requirements than those of their targets.  This
4429           typically represents a potential optimization opportunity.  By
4430           contrast, the -Wattribute-alias=2 option controls warnings issued
4431           when the alias is more restrictive than the target, which could
4432           lead to incorrect code generation.  Attributes considered include
4433           "alloc_align", "alloc_size", "cold", "const", "hot", "leaf",
4434           "malloc", "nonnull", "noreturn", "nothrow", "pure",
4435           "returns_nonnull", and "returns_twice".
4436
4437           In C++, the warning is issued when an explicit specialization of a
4438           primary template declared with attribute "alloc_align",
4439           "alloc_size", "assume_aligned", "format", "format_arg", "malloc",
4440           or "nonnull" is declared without it.  Attributes "deprecated",
4441           "error", and "warning" suppress the warning..
4442
4443           You can use the "copy" attribute to apply the same set of
4444           attributes to a declaration as that on another declaration without
4445           explicitly enumerating the attributes. This attribute can be
4446           applied to declarations of functions, variables, or types.
4447
4448           -Wmissing-attributes is enabled by -Wall.
4449
4450           For example, since the declaration of the primary function template
4451           below makes use of both attribute "malloc" and "alloc_size" the
4452           declaration of the explicit specialization of the template is
4453           diagnosed because it is missing one of the attributes.
4454
4455                   template <class T>
4456                   T* __attribute__ ((malloc, alloc_size (1)))
4457                   allocate (size_t);
4458
4459                   template <>
4460                   void* __attribute__ ((malloc))   // missing alloc_size
4461                   allocate<void> (size_t);
4462
4463       -Wmissing-braces
4464           Warn if an aggregate or union initializer is not fully bracketed.
4465           In the following example, the initializer for "a" is not fully
4466           bracketed, but that for "b" is fully bracketed.
4467
4468                   int a[2][2] = { 0, 1, 2, 3 };
4469                   int b[2][2] = { { 0, 1 }, { 2, 3 } };
4470
4471           This warning is enabled by -Wall.
4472
4473       -Wmissing-include-dirs (C, C++, Objective-C and Objective-C++ only)
4474           Warn if a user-supplied include directory does not exist.
4475
4476       -Wno-missing-profile
4477           This option controls warnings if feedback profiles are missing when
4478           using the -fprofile-use option.  This option diagnoses those cases
4479           where a new function or a new file is added between compiling with
4480           -fprofile-generate and with -fprofile-use, without regenerating the
4481           profiles.  In these cases, the profile feedback data files do not
4482           contain any profile feedback information for the newly added
4483           function or file respectively.  Also, in the case when profile
4484           count data (.gcda) files are removed, GCC cannot use any profile
4485           feedback information.  In all these cases, warnings are issued to
4486           inform you that a profile generation step is due.  Ignoring the
4487           warning can result in poorly optimized code.  -Wno-missing-profile
4488           can be used to disable the warning, but this is not recommended and
4489           should be done only when non-existent profile data is justified.
4490
4491       -Wmultistatement-macros
4492           Warn about unsafe multiple statement macros that appear to be
4493           guarded by a clause such as "if", "else", "for", "switch", or
4494           "while", in which only the first statement is actually guarded
4495           after the macro is expanded.
4496
4497           For example:
4498
4499                   #define DOIT x++; y++
4500                   if (c)
4501                     DOIT;
4502
4503           will increment "y" unconditionally, not just when "c" holds.  The
4504           can usually be fixed by wrapping the macro in a do-while loop:
4505
4506                   #define DOIT do { x++; y++; } while (0)
4507                   if (c)
4508                     DOIT;
4509
4510           This warning is enabled by -Wall in C and C++.
4511
4512       -Wparentheses
4513           Warn if parentheses are omitted in certain contexts, such as when
4514           there is an assignment in a context where a truth value is
4515           expected, or when operators are nested whose precedence people
4516           often get confused about.
4517
4518           Also warn if a comparison like "x<=y<=z" appears; this is
4519           equivalent to "(x<=y ? 1 : 0) <= z", which is a different
4520           interpretation from that of ordinary mathematical notation.
4521
4522           Also warn for dangerous uses of the GNU extension to "?:" with
4523           omitted middle operand. When the condition in the "?": operator is
4524           a boolean expression, the omitted value is always 1.  Often
4525           programmers expect it to be a value computed inside the conditional
4526           expression instead.
4527
4528           For C++ this also warns for some cases of unnecessary parentheses
4529           in declarations, which can indicate an attempt at a function call
4530           instead of a declaration:
4531
4532                   {
4533                     // Declares a local variable called mymutex.
4534                     std::unique_lock<std::mutex> (mymutex);
4535                     // User meant std::unique_lock<std::mutex> lock (mymutex);
4536                   }
4537
4538           This warning is enabled by -Wall.
4539
4540       -Wsequence-point
4541           Warn about code that may have undefined semantics because of
4542           violations of sequence point rules in the C and C++ standards.
4543
4544           The C and C++ standards define the order in which expressions in a
4545           C/C++ program are evaluated in terms of sequence points, which
4546           represent a partial ordering between the execution of parts of the
4547           program: those executed before the sequence point, and those
4548           executed after it.  These occur after the evaluation of a full
4549           expression (one which is not part of a larger expression), after
4550           the evaluation of the first operand of a "&&", "||", "? :" or ","
4551           (comma) operator, before a function is called (but after the
4552           evaluation of its arguments and the expression denoting the called
4553           function), and in certain other places.  Other than as expressed by
4554           the sequence point rules, the order of evaluation of subexpressions
4555           of an expression is not specified.  All these rules describe only a
4556           partial order rather than a total order, since, for example, if two
4557           functions are called within one expression with no sequence point
4558           between them, the order in which the functions are called is not
4559           specified.  However, the standards committee have ruled that
4560           function calls do not overlap.
4561
4562           It is not specified when between sequence points modifications to
4563           the values of objects take effect.  Programs whose behavior depends
4564           on this have undefined behavior; the C and C++ standards specify
4565           that "Between the previous and next sequence point an object shall
4566           have its stored value modified at most once by the evaluation of an
4567           expression.  Furthermore, the prior value shall be read only to
4568           determine the value to be stored.".  If a program breaks these
4569           rules, the results on any particular implementation are entirely
4570           unpredictable.
4571
4572           Examples of code with undefined behavior are "a = a++;", "a[n] =
4573           b[n++]" and "a[i++] = i;".  Some more complicated cases are not
4574           diagnosed by this option, and it may give an occasional false
4575           positive result, but in general it has been found fairly effective
4576           at detecting this sort of problem in programs.
4577
4578           The C++17 standard will define the order of evaluation of operands
4579           in more cases: in particular it requires that the right-hand side
4580           of an assignment be evaluated before the left-hand side, so the
4581           above examples are no longer undefined.  But this option will still
4582           warn about them, to help people avoid writing code that is
4583           undefined in C and earlier revisions of C++.
4584
4585           The standard is worded confusingly, therefore there is some debate
4586           over the precise meaning of the sequence point rules in subtle
4587           cases.  Links to discussions of the problem, including proposed
4588           formal definitions, may be found on the GCC readings page, at
4589           <http://gcc.gnu.org/readings.html>.
4590
4591           This warning is enabled by -Wall for C and C++.
4592
4593       -Wno-return-local-addr
4594           Do not warn about returning a pointer (or in C++, a reference) to a
4595           variable that goes out of scope after the function returns.
4596
4597       -Wreturn-type
4598           Warn whenever a function is defined with a return type that
4599           defaults to "int".  Also warn about any "return" statement with no
4600           return value in a function whose return type is not "void" (falling
4601           off the end of the function body is considered returning without a
4602           value).
4603
4604           For C only, warn about a "return" statement with an expression in a
4605           function whose return type is "void", unless the expression type is
4606           also "void".  As a GNU extension, the latter case is accepted
4607           without a warning unless -Wpedantic is used.  Attempting to use the
4608           return value of a non-"void" function other than "main" that flows
4609           off the end by reaching the closing curly brace that terminates the
4610           function is undefined.
4611
4612           Unlike in C, in C++, flowing off the end of a non-"void" function
4613           other than "main" results in undefined behavior even when the value
4614           of the function is not used.
4615
4616           This warning is enabled by default in C++ and by -Wall otherwise.
4617
4618       -Wno-shift-count-negative
4619           Controls warnings if a shift count is negative.  This warning is
4620           enabled by default.
4621
4622       -Wno-shift-count-overflow
4623           Controls warnings if a shift count is greater than or equal to the
4624           bit width of the type.  This warning is enabled by default.
4625
4626       -Wshift-negative-value
4627           Warn if left shifting a negative value.  This warning is enabled by
4628           -Wextra in C99 and C++11 modes (and newer).
4629
4630       -Wno-shift-overflow
4631       -Wshift-overflow=n
4632           These options control warnings about left shift overflows.
4633
4634           -Wshift-overflow=1
4635               This is the warning level of -Wshift-overflow and is enabled by
4636               default in C99 and C++11 modes (and newer).  This warning level
4637               does not warn about left-shifting 1 into the sign bit.
4638               (However, in C, such an overflow is still rejected in contexts
4639               where an integer constant expression is required.)  No warning
4640               is emitted in C++2A mode (and newer), as signed left shifts
4641               always wrap.
4642
4643           -Wshift-overflow=2
4644               This warning level also warns about left-shifting 1 into the
4645               sign bit, unless C++14 mode (or newer) is active.
4646
4647       -Wswitch
4648           Warn whenever a "switch" statement has an index of enumerated type
4649           and lacks a "case" for one or more of the named codes of that
4650           enumeration.  (The presence of a "default" label prevents this
4651           warning.)  "case" labels outside the enumeration range also provoke
4652           warnings when this option is used (even if there is a "default"
4653           label).  This warning is enabled by -Wall.
4654
4655       -Wswitch-default
4656           Warn whenever a "switch" statement does not have a "default" case.
4657
4658       -Wswitch-enum
4659           Warn whenever a "switch" statement has an index of enumerated type
4660           and lacks a "case" for one or more of the named codes of that
4661           enumeration.  "case" labels outside the enumeration range also
4662           provoke warnings when this option is used.  The only difference
4663           between -Wswitch and this option is that this option gives a
4664           warning about an omitted enumeration code even if there is a
4665           "default" label.
4666
4667       -Wno-switch-bool
4668           Do not warn when a "switch" statement has an index of boolean type
4669           and the case values are outside the range of a boolean type.  It is
4670           possible to suppress this warning by casting the controlling
4671           expression to a type other than "bool".  For example:
4672
4673                   switch ((int) (a == 4))
4674                     {
4675                     ...
4676                     }
4677
4678           This warning is enabled by default for C and C++ programs.
4679
4680       -Wno-switch-outside-range
4681           This option controls warnings when a "switch" case has a value that
4682           is outside of its respective type range.  This warning is enabled
4683           by default for C and C++ programs.
4684
4685       -Wno-switch-unreachable
4686           Do not warn when a "switch" statement contains statements between
4687           the controlling expression and the first case label, which will
4688           never be executed.  For example:
4689
4690                   switch (cond)
4691                     {
4692                      i = 15;
4693                     ...
4694                      case 5:
4695                     ...
4696                     }
4697
4698           -Wswitch-unreachable does not warn if the statement between the
4699           controlling expression and the first case label is just a
4700           declaration:
4701
4702                   switch (cond)
4703                     {
4704                      int i;
4705                     ...
4706                      case 5:
4707                      i = 5;
4708                     ...
4709                     }
4710
4711           This warning is enabled by default for C and C++ programs.
4712
4713       -Wsync-nand (C and C++ only)
4714           Warn when "__sync_fetch_and_nand" and "__sync_nand_and_fetch"
4715           built-in functions are used.  These functions changed semantics in
4716           GCC 4.4.
4717
4718       -Wunused-but-set-parameter
4719           Warn whenever a function parameter is assigned to, but otherwise
4720           unused (aside from its declaration).
4721
4722           To suppress this warning use the "unused" attribute.
4723
4724           This warning is also enabled by -Wunused together with -Wextra.
4725
4726       -Wunused-but-set-variable
4727           Warn whenever a local variable is assigned to, but otherwise unused
4728           (aside from its declaration).  This warning is enabled by -Wall.
4729
4730           To suppress this warning use the "unused" attribute.
4731
4732           This warning is also enabled by -Wunused, which is enabled by
4733           -Wall.
4734
4735       -Wunused-function
4736           Warn whenever a static function is declared but not defined or a
4737           non-inline static function is unused.  This warning is enabled by
4738           -Wall.
4739
4740       -Wunused-label
4741           Warn whenever a label is declared but not used.  This warning is
4742           enabled by -Wall.
4743
4744           To suppress this warning use the "unused" attribute.
4745
4746       -Wunused-local-typedefs (C, Objective-C, C++ and Objective-C++ only)
4747           Warn when a typedef locally defined in a function is not used.
4748           This warning is enabled by -Wall.
4749
4750       -Wunused-parameter
4751           Warn whenever a function parameter is unused aside from its
4752           declaration.
4753
4754           To suppress this warning use the "unused" attribute.
4755
4756       -Wno-unused-result
4757           Do not warn if a caller of a function marked with attribute
4758           "warn_unused_result" does not use its return value. The default is
4759           -Wunused-result.
4760
4761       -Wunused-variable
4762           Warn whenever a local or static variable is unused aside from its
4763           declaration. This option implies -Wunused-const-variable=1 for C,
4764           but not for C++. This warning is enabled by -Wall.
4765
4766           To suppress this warning use the "unused" attribute.
4767
4768       -Wunused-const-variable
4769       -Wunused-const-variable=n
4770           Warn whenever a constant static variable is unused aside from its
4771           declaration.  -Wunused-const-variable=1 is enabled by
4772           -Wunused-variable for C, but not for C++. In C this declares
4773           variable storage, but in C++ this is not an error since const
4774           variables take the place of "#define"s.
4775
4776           To suppress this warning use the "unused" attribute.
4777
4778           -Wunused-const-variable=1
4779               This is the warning level that is enabled by -Wunused-variable
4780               for C.  It warns only about unused static const variables
4781               defined in the main compilation unit, but not about static
4782               const variables declared in any header included.
4783
4784           -Wunused-const-variable=2
4785               This warning level also warns for unused constant static
4786               variables in headers (excluding system headers).  This is the
4787               warning level of -Wunused-const-variable and must be explicitly
4788               requested since in C++ this isn't an error and in C it might be
4789               harder to clean up all headers included.
4790
4791       -Wunused-value
4792           Warn whenever a statement computes a result that is explicitly not
4793           used. To suppress this warning cast the unused expression to
4794           "void". This includes an expression-statement or the left-hand side
4795           of a comma expression that contains no side effects. For example,
4796           an expression such as "x[i,j]" causes a warning, while
4797           "x[(void)i,j]" does not.
4798
4799           This warning is enabled by -Wall.
4800
4801       -Wunused
4802           All the above -Wunused options combined.
4803
4804           In order to get a warning about an unused function parameter, you
4805           must either specify -Wextra -Wunused (note that -Wall implies
4806           -Wunused), or separately specify -Wunused-parameter.
4807
4808       -Wuninitialized
4809           Warn if an automatic variable is used without first being
4810           initialized.  In C++, warn if a non-static reference or non-static
4811           "const" member appears in a class without constructors.
4812
4813           If you want to warn about code that uses the uninitialized value of
4814           the variable in its own initializer, use the -Winit-self option.
4815
4816           These warnings occur for individual uninitialized elements of
4817           structure, union or array variables as well as for variables that
4818           are uninitialized as a whole.  They do not occur for variables or
4819           elements declared "volatile".  Because these warnings depend on
4820           optimization, the exact variables or elements for which there are
4821           warnings depend on the precise optimization options and version of
4822           GCC used.
4823
4824           Note that there may be no warning about a variable that is used
4825           only to compute a value that itself is never used, because such
4826           computations may be deleted by data flow analysis before the
4827           warnings are printed.
4828
4829       -Wno-invalid-memory-model
4830           This option controls warnings for invocations of __atomic Builtins,
4831           __sync Builtins, and the C11 atomic generic functions with a memory
4832           consistency argument that is either invalid for the operation or
4833           outside the range of values of the "memory_order" enumeration.  For
4834           example, since the "__atomic_store" and "__atomic_store_n" built-
4835           ins are only defined for the relaxed, release, and sequentially
4836           consistent memory orders the following code is diagnosed:
4837
4838                   void store (int *i)
4839                   {
4840                     __atomic_store_n (i, 0, memory_order_consume);
4841                   }
4842
4843           -Winvalid-memory-model is enabled by default.
4844
4845       -Wmaybe-uninitialized
4846           For an automatic (i.e. local) variable, if there exists a path from
4847           the function entry to a use of the variable that is initialized,
4848           but there exist some other paths for which the variable is not
4849           initialized, the compiler emits a warning if it cannot prove the
4850           uninitialized paths are not executed at run time.
4851
4852           These warnings are only possible in optimizing compilation, because
4853           otherwise GCC does not keep track of the state of variables.
4854
4855           These warnings are made optional because GCC may not be able to
4856           determine when the code is correct in spite of appearing to have an
4857           error.  Here is one example of how this can happen:
4858
4859                   {
4860                     int x;
4861                     switch (y)
4862                       {
4863                       case 1: x = 1;
4864                         break;
4865                       case 2: x = 4;
4866                         break;
4867                       case 3: x = 5;
4868                       }
4869                     foo (x);
4870                   }
4871
4872           If the value of "y" is always 1, 2 or 3, then "x" is always
4873           initialized, but GCC doesn't know this. To suppress the warning,
4874           you need to provide a default case with assert(0) or similar code.
4875
4876           This option also warns when a non-volatile automatic variable might
4877           be changed by a call to "longjmp".  The compiler sees only the
4878           calls to "setjmp".  It cannot know where "longjmp" will be called;
4879           in fact, a signal handler could call it at any point in the code.
4880           As a result, you may get a warning even when there is in fact no
4881           problem because "longjmp" cannot in fact be called at the place
4882           that would cause a problem.
4883
4884           Some spurious warnings can be avoided if you declare all the
4885           functions you use that never return as "noreturn".
4886
4887           This warning is enabled by -Wall or -Wextra.
4888
4889       -Wunknown-pragmas
4890           Warn when a "#pragma" directive is encountered that is not
4891           understood by GCC.  If this command-line option is used, warnings
4892           are even issued for unknown pragmas in system header files.  This
4893           is not the case if the warnings are only enabled by the -Wall
4894           command-line option.
4895
4896       -Wno-pragmas
4897           Do not warn about misuses of pragmas, such as incorrect parameters,
4898           invalid syntax, or conflicts between pragmas.  See also
4899           -Wunknown-pragmas.
4900
4901       -Wno-prio-ctor-dtor
4902           Do not warn if a priority from 0 to 100 is used for constructor or
4903           destructor.  The use of constructor and destructor attributes allow
4904           you to assign a priority to the constructor/destructor to control
4905           its order of execution before "main" is called or after it returns.
4906           The priority values must be greater than 100 as the compiler
4907           reserves priority values between 0--100 for the implementation.
4908
4909       -Wstrict-aliasing
4910           This option is only active when -fstrict-aliasing is active.  It
4911           warns about code that might break the strict aliasing rules that
4912           the compiler is using for optimization.  The warning does not catch
4913           all cases, but does attempt to catch the more common pitfalls.  It
4914           is included in -Wall.  It is equivalent to -Wstrict-aliasing=3
4915
4916       -Wstrict-aliasing=n
4917           This option is only active when -fstrict-aliasing is active.  It
4918           warns about code that might break the strict aliasing rules that
4919           the compiler is using for optimization.  Higher levels correspond
4920           to higher accuracy (fewer false positives).  Higher levels also
4921           correspond to more effort, similar to the way -O works.
4922           -Wstrict-aliasing is equivalent to -Wstrict-aliasing=3.
4923
4924           Level 1: Most aggressive, quick, least accurate.  Possibly useful
4925           when higher levels do not warn but -fstrict-aliasing still breaks
4926           the code, as it has very few false negatives.  However, it has many
4927           false positives.  Warns for all pointer conversions between
4928           possibly incompatible types, even if never dereferenced.  Runs in
4929           the front end only.
4930
4931           Level 2: Aggressive, quick, not too precise.  May still have many
4932           false positives (not as many as level 1 though), and few false
4933           negatives (but possibly more than level 1).  Unlike level 1, it
4934           only warns when an address is taken.  Warns about incomplete types.
4935           Runs in the front end only.
4936
4937           Level 3 (default for -Wstrict-aliasing): Should have very few false
4938           positives and few false negatives.  Slightly slower than levels 1
4939           or 2 when optimization is enabled.  Takes care of the common
4940           pun+dereference pattern in the front end: "*(int*)&some_float".  If
4941           optimization is enabled, it also runs in the back end, where it
4942           deals with multiple statement cases using flow-sensitive points-to
4943           information.  Only warns when the converted pointer is
4944           dereferenced.  Does not warn about incomplete types.
4945
4946       -Wstrict-overflow
4947       -Wstrict-overflow=n
4948           This option is only active when signed overflow is undefined.  It
4949           warns about cases where the compiler optimizes based on the
4950           assumption that signed overflow does not occur.  Note that it does
4951           not warn about all cases where the code might overflow: it only
4952           warns about cases where the compiler implements some optimization.
4953           Thus this warning depends on the optimization level.
4954
4955           An optimization that assumes that signed overflow does not occur is
4956           perfectly safe if the values of the variables involved are such
4957           that overflow never does, in fact, occur.  Therefore this warning
4958           can easily give a false positive: a warning about code that is not
4959           actually a problem.  To help focus on important issues, several
4960           warning levels are defined.  No warnings are issued for the use of
4961           undefined signed overflow when estimating how many iterations a
4962           loop requires, in particular when determining whether a loop will
4963           be executed at all.
4964
4965           -Wstrict-overflow=1
4966               Warn about cases that are both questionable and easy to avoid.
4967               For example the compiler simplifies "x + 1 > x" to 1.  This
4968               level of -Wstrict-overflow is enabled by -Wall; higher levels
4969               are not, and must be explicitly requested.
4970
4971           -Wstrict-overflow=2
4972               Also warn about other cases where a comparison is simplified to
4973               a constant.  For example: "abs (x) >= 0".  This can only be
4974               simplified when signed integer overflow is undefined, because
4975               "abs (INT_MIN)" overflows to "INT_MIN", which is less than
4976               zero.  -Wstrict-overflow (with no level) is the same as
4977               -Wstrict-overflow=2.
4978
4979           -Wstrict-overflow=3
4980               Also warn about other cases where a comparison is simplified.
4981               For example: "x + 1 > 1" is simplified to "x > 0".
4982
4983           -Wstrict-overflow=4
4984               Also warn about other simplifications not covered by the above
4985               cases.  For example: "(x * 10) / 5" is simplified to "x * 2".
4986
4987           -Wstrict-overflow=5
4988               Also warn about cases where the compiler reduces the magnitude
4989               of a constant involved in a comparison.  For example: "x + 2 >
4990               y" is simplified to "x + 1 >= y".  This is reported only at the
4991               highest warning level because this simplification applies to
4992               many comparisons, so this warning level gives a very large
4993               number of false positives.
4994
4995       -Wstring-compare
4996           Warn for calls to "strcmp" and "strncmp" whose result is determined
4997           to be either zero or non-zero in tests for such equality owing to
4998           the length of one argument being greater than the size of the array
4999           the other argument is stored in (or the bound in the case of
5000           "strncmp").  Such calls could be mistakes.  For example, the call
5001           to "strcmp" below is diagnosed because its result is necessarily
5002           non-zero irrespective of the contents of the array "a".
5003
5004                   extern char a[4];
5005                   void f (char *d)
5006                   {
5007                     strcpy (d, "string");
5008                     ...
5009                     if (0 == strcmp (a, d))   // cannot be true
5010                       puts ("a and d are the same");
5011                   }
5012
5013           -Wstring-compare is enabled by -Wextra.
5014
5015       -Wstringop-overflow
5016       -Wstringop-overflow=type
5017           Warn for calls to string manipulation functions such as "memcpy"
5018           and "strcpy" that are determined to overflow the destination
5019           buffer.  The optional argument is one greater than the type of
5020           Object Size Checking to perform to determine the size of the
5021           destination.  The argument is meaningful only for functions that
5022           operate on character arrays but not for raw memory functions like
5023           "memcpy" which always make use of Object Size type-0.  The option
5024           also warns for calls that specify a size in excess of the largest
5025           possible object or at most "SIZE_MAX / 2" bytes.  The option
5026           produces the best results with optimization enabled but can detect
5027           a small subset of simple buffer overflows even without optimization
5028           in calls to the GCC built-in functions like "__builtin_memcpy" that
5029           correspond to the standard functions.  In any case, the option
5030           warns about just a subset of buffer overflows detected by the
5031           corresponding overflow checking built-ins.  For example, the option
5032           issues a warning for the "strcpy" call below because it copies at
5033           least 5 characters (the string "blue" including the terminating
5034           NUL) into the buffer of size 4.
5035
5036                   enum Color { blue, purple, yellow };
5037                   const char* f (enum Color clr)
5038                   {
5039                     static char buf [4];
5040                     const char *str;
5041                     switch (clr)
5042                       {
5043                         case blue: str = "blue"; break;
5044                         case purple: str = "purple"; break;
5045                         case yellow: str = "yellow"; break;
5046                       }
5047
5048                     return strcpy (buf, str);   // warning here
5049                   }
5050
5051           Option -Wstringop-overflow=2 is enabled by default.
5052
5053           -Wstringop-overflow
5054           -Wstringop-overflow=1
5055               The -Wstringop-overflow=1 option uses type-zero Object Size
5056               Checking to determine the sizes of destination objects.  This
5057               is the default setting of the option.  At this setting the
5058               option does not warn for writes past the end of subobjects of
5059               larger objects accessed by pointers unless the size of the
5060               largest surrounding object is known.  When the destination may
5061               be one of several objects it is assumed to be the largest one
5062               of them.  On Linux systems, when optimization is enabled at
5063               this setting the option warns for the same code as when the
5064               "_FORTIFY_SOURCE" macro is defined to a non-zero value.
5065
5066           -Wstringop-overflow=2
5067               The -Wstringop-overflow=2 option uses type-one Object Size
5068               Checking to determine the sizes of destination objects.  At
5069               this setting the option warna about overflows when writing to
5070               members of the largest complete objects whose exact size is
5071               known.  However, it does not warn for excessive writes to the
5072               same members of unknown objects referenced by pointers since
5073               they may point to arrays containing unknown numbers of
5074               elements.
5075
5076           -Wstringop-overflow=3
5077               The -Wstringop-overflow=3 option uses type-two Object Size
5078               Checking to determine the sizes of destination objects.  At
5079               this setting the option warns about overflowing the smallest
5080               object or data member.  This is the most restrictive setting of
5081               the option that may result in warnings for safe code.
5082
5083           -Wstringop-overflow=4
5084               The -Wstringop-overflow=4 option uses type-three Object Size
5085               Checking to determine the sizes of destination objects.  At
5086               this setting the option warns about overflowing any data
5087               members, and when the destination is one of several objects it
5088               uses the size of the largest of them to decide whether to issue
5089               a warning.  Similarly to -Wstringop-overflow=3 this setting of
5090               the option may result in warnings for benign code.
5091
5092       -Wno-stringop-truncation
5093           Do not warn for calls to bounded string manipulation functions such
5094           as "strncat", "strncpy", and "stpncpy" that may either truncate the
5095           copied string or leave the destination unchanged.
5096
5097           In the following example, the call to "strncat" specifies a bound
5098           that is less than the length of the source string.  As a result,
5099           the copy of the source will be truncated and so the call is
5100           diagnosed.  To avoid the warning use "bufsize - strlen (buf) - 1)"
5101           as the bound.
5102
5103                   void append (char *buf, size_t bufsize)
5104                   {
5105                     strncat (buf, ".txt", 3);
5106                   }
5107
5108           As another example, the following call to "strncpy" results in
5109           copying to "d" just the characters preceding the terminating NUL,
5110           without appending the NUL to the end.  Assuming the result of
5111           "strncpy" is necessarily a NUL-terminated string is a common
5112           mistake, and so the call is diagnosed.  To avoid the warning when
5113           the result is not expected to be NUL-terminated, call "memcpy"
5114           instead.
5115
5116                   void copy (char *d, const char *s)
5117                   {
5118                     strncpy (d, s, strlen (s));
5119                   }
5120
5121           In the following example, the call to "strncpy" specifies the size
5122           of the destination buffer as the bound.  If the length of the
5123           source string is equal to or greater than this size the result of
5124           the copy will not be NUL-terminated.  Therefore, the call is also
5125           diagnosed.  To avoid the warning, specify "sizeof buf - 1" as the
5126           bound and set the last element of the buffer to "NUL".
5127
5128                   void copy (const char *s)
5129                   {
5130                     char buf[80];
5131                     strncpy (buf, s, sizeof buf);
5132                     ...
5133                   }
5134
5135           In situations where a character array is intended to store a
5136           sequence of bytes with no terminating "NUL" such an array may be
5137           annotated with attribute "nonstring" to avoid this warning.  Such
5138           arrays, however, are not suitable arguments to functions that
5139           expect "NUL"-terminated strings.  To help detect accidental misuses
5140           of such arrays GCC issues warnings unless it can prove that the use
5141           is safe.
5142
5143       -Wsuggest-attribute=[pure|const|noreturn|format|cold|malloc]
5144           Warn for cases where adding an attribute may be beneficial. The
5145           attributes currently supported are listed below.
5146
5147           -Wsuggest-attribute=pure
5148           -Wsuggest-attribute=const
5149           -Wsuggest-attribute=noreturn
5150           -Wmissing-noreturn
5151           -Wsuggest-attribute=malloc
5152               Warn about functions that might be candidates for attributes
5153               "pure", "const" or "noreturn" or "malloc". The compiler only
5154               warns for functions visible in other compilation units or (in
5155               the case of "pure" and "const") if it cannot prove that the
5156               function returns normally. A function returns normally if it
5157               doesn't contain an infinite loop or return abnormally by
5158               throwing, calling "abort" or trapping.  This analysis requires
5159               option -fipa-pure-const, which is enabled by default at -O and
5160               higher.  Higher optimization levels improve the accuracy of the
5161               analysis.
5162
5163           -Wsuggest-attribute=format
5164           -Wmissing-format-attribute
5165               Warn about function pointers that might be candidates for
5166               "format" attributes.  Note these are only possible candidates,
5167               not absolute ones.  GCC guesses that function pointers with
5168               "format" attributes that are used in assignment,
5169               initialization, parameter passing or return statements should
5170               have a corresponding "format" attribute in the resulting type.
5171               I.e. the left-hand side of the assignment or initialization,
5172               the type of the parameter variable, or the return type of the
5173               containing function respectively should also have a "format"
5174               attribute to avoid the warning.
5175
5176               GCC also warns about function definitions that might be
5177               candidates for "format" attributes.  Again, these are only
5178               possible candidates.  GCC guesses that "format" attributes
5179               might be appropriate for any function that calls a function
5180               like "vprintf" or "vscanf", but this might not always be the
5181               case, and some functions for which "format" attributes are
5182               appropriate may not be detected.
5183
5184           -Wsuggest-attribute=cold
5185               Warn about functions that might be candidates for "cold"
5186               attribute.  This is based on static detection and generally
5187               only warns about functions which always leads to a call to
5188               another "cold" function such as wrappers of C++ "throw" or
5189               fatal error reporting functions leading to "abort".
5190
5191       -Walloc-zero
5192           Warn about calls to allocation functions decorated with attribute
5193           "alloc_size" that specify zero bytes, including those to the built-
5194           in forms of the functions "aligned_alloc", "alloca", "calloc",
5195           "malloc", and "realloc".  Because the behavior of these functions
5196           when called with a zero size differs among implementations (and in
5197           the case of "realloc" has been deprecated) relying on it may result
5198           in subtle portability bugs and should be avoided.
5199
5200       -Walloc-size-larger-than=byte-size
5201           Warn about calls to functions decorated with attribute "alloc_size"
5202           that attempt to allocate objects larger than the specified number
5203           of bytes, or where the result of the size computation in an integer
5204           type with infinite precision would exceed the value of PTRDIFF_MAX
5205           on the target.  -Walloc-size-larger-than=PTRDIFF_MAX is enabled by
5206           default.  Warnings controlled by the option can be disabled either
5207           by specifying byte-size of SIZE_MAX or more or by
5208           -Wno-alloc-size-larger-than.
5209
5210       -Wno-alloc-size-larger-than
5211           Disable -Walloc-size-larger-than= warnings.  The option is
5212           equivalent to -Walloc-size-larger-than=SIZE_MAX or larger.
5213
5214       -Walloca
5215           This option warns on all uses of "alloca" in the source.
5216
5217       -Walloca-larger-than=byte-size
5218           This option warns on calls to "alloca" with an integer argument
5219           whose value is either zero, or that is not bounded by a controlling
5220           predicate that limits its value to at most byte-size.  It also
5221           warns for calls to "alloca" where the bound value is unknown.
5222           Arguments of non-integer types are considered unbounded even if
5223           they appear to be constrained to the expected range.
5224
5225           For example, a bounded case of "alloca" could be:
5226
5227                   void func (size_t n)
5228                   {
5229                     void *p;
5230                     if (n <= 1000)
5231                       p = alloca (n);
5232                     else
5233                       p = malloc (n);
5234                     f (p);
5235                   }
5236
5237           In the above example, passing "-Walloca-larger-than=1000" would not
5238           issue a warning because the call to "alloca" is known to be at most
5239           1000 bytes.  However, if "-Walloca-larger-than=500" were passed,
5240           the compiler would emit a warning.
5241
5242           Unbounded uses, on the other hand, are uses of "alloca" with no
5243           controlling predicate constraining its integer argument.  For
5244           example:
5245
5246                   void func ()
5247                   {
5248                     void *p = alloca (n);
5249                     f (p);
5250                   }
5251
5252           If "-Walloca-larger-than=500" were passed, the above would trigger
5253           a warning, but this time because of the lack of bounds checking.
5254
5255           Note, that even seemingly correct code involving signed integers
5256           could cause a warning:
5257
5258                   void func (signed int n)
5259                   {
5260                     if (n < 500)
5261                       {
5262                         p = alloca (n);
5263                         f (p);
5264                       }
5265                   }
5266
5267           In the above example, n could be negative, causing a larger than
5268           expected argument to be implicitly cast into the "alloca" call.
5269
5270           This option also warns when "alloca" is used in a loop.
5271
5272           -Walloca-larger-than=PTRDIFF_MAX is enabled by default but is
5273           usually only effective  when -ftree-vrp is active (default for -O2
5274           and above).
5275
5276           See also -Wvla-larger-than=byte-size.
5277
5278       -Wno-alloca-larger-than
5279           Disable -Walloca-larger-than= warnings.  The option is equivalent
5280           to -Walloca-larger-than=SIZE_MAX or larger.
5281
5282       -Warith-conversion
5283           Do warn about implicit conversions from arithmetic operations even
5284           when conversion of the operands to the same type cannot change
5285           their values.  This affects warnings from -Wconversion,
5286           -Wfloat-conversion, and -Wsign-conversion.
5287
5288                   void f (char c, int i)
5289                   {
5290                     c = c + i; // warns with B<-Wconversion>
5291                     c = c + 1; // only warns with B<-Warith-conversion>
5292                   }
5293
5294       -Warray-bounds
5295       -Warray-bounds=n
5296           This option is only active when -ftree-vrp is active (default for
5297           -O2 and above). It warns about subscripts to arrays that are always
5298           out of bounds. This warning is enabled by -Wall.
5299
5300           -Warray-bounds=1
5301               This is the warning level of -Warray-bounds and is enabled by
5302               -Wall; higher levels are not, and must be explicitly requested.
5303
5304           -Warray-bounds=2
5305               This warning level also warns about out of bounds access for
5306               arrays at the end of a struct and for arrays accessed through
5307               pointers. This warning level may give a larger number of false
5308               positives and is deactivated by default.
5309
5310       -Wattribute-alias=n
5311       -Wno-attribute-alias
5312           Warn about declarations using the "alias" and similar attributes
5313           whose target is incompatible with the type of the alias.
5314
5315           -Wattribute-alias=1
5316               The default warning level of the -Wattribute-alias option
5317               diagnoses incompatibilities between the type of the alias
5318               declaration and that of its target.  Such incompatibilities are
5319               typically indicative of bugs.
5320
5321           -Wattribute-alias=2
5322               At this level -Wattribute-alias also diagnoses cases where the
5323               attributes of the alias declaration are more restrictive than
5324               the attributes applied to its target.  These mismatches can
5325               potentially result in incorrect code generation.  In other
5326               cases they may be benign and could be resolved simply by adding
5327               the missing attribute to the target.  For comparison, see the
5328               -Wmissing-attributes option, which controls diagnostics when
5329               the alias declaration is less restrictive than the target,
5330               rather than more restrictive.
5331
5332               Attributes considered include "alloc_align", "alloc_size",
5333               "cold", "const", "hot", "leaf", "malloc", "nonnull",
5334               "noreturn", "nothrow", "pure", "returns_nonnull", and
5335               "returns_twice".
5336
5337           -Wattribute-alias is equivalent to -Wattribute-alias=1.  This is
5338           the default.  You can disable these warnings with either
5339           -Wno-attribute-alias or -Wattribute-alias=0.
5340
5341       -Wbool-compare
5342           Warn about boolean expression compared with an integer value
5343           different from "true"/"false".  For instance, the following
5344           comparison is always false:
5345
5346                   int n = 5;
5347                   ...
5348                   if ((n > 1) == 2) { ... }
5349
5350           This warning is enabled by -Wall.
5351
5352       -Wbool-operation
5353           Warn about suspicious operations on expressions of a boolean type.
5354           For instance, bitwise negation of a boolean is very likely a bug in
5355           the program.  For C, this warning also warns about incrementing or
5356           decrementing a boolean, which rarely makes sense.  (In C++,
5357           decrementing a boolean is always invalid.  Incrementing a boolean
5358           is invalid in C++17, and deprecated otherwise.)
5359
5360           This warning is enabled by -Wall.
5361
5362       -Wduplicated-branches
5363           Warn when an if-else has identical branches.  This warning detects
5364           cases like
5365
5366                   if (p != NULL)
5367                     return 0;
5368                   else
5369                     return 0;
5370
5371           It doesn't warn when both branches contain just a null statement.
5372           This warning also warn for conditional operators:
5373
5374                     int i = x ? *p : *p;
5375
5376       -Wduplicated-cond
5377           Warn about duplicated conditions in an if-else-if chain.  For
5378           instance, warn for the following code:
5379
5380                   if (p->q != NULL) { ... }
5381                   else if (p->q != NULL) { ... }
5382
5383       -Wframe-address
5384           Warn when the __builtin_frame_address or __builtin_return_address
5385           is called with an argument greater than 0.  Such calls may return
5386           indeterminate values or crash the program.  The warning is included
5387           in -Wall.
5388
5389       -Wno-discarded-qualifiers (C and Objective-C only)
5390           Do not warn if type qualifiers on pointers are being discarded.
5391           Typically, the compiler warns if a "const char *" variable is
5392           passed to a function that takes a "char *" parameter.  This option
5393           can be used to suppress such a warning.
5394
5395       -Wno-discarded-array-qualifiers (C and Objective-C only)
5396           Do not warn if type qualifiers on arrays which are pointer targets
5397           are being discarded.  Typically, the compiler warns if a "const int
5398           (*)[]" variable is passed to a function that takes a "int (*)[]"
5399           parameter.  This option can be used to suppress such a warning.
5400
5401       -Wno-incompatible-pointer-types (C and Objective-C only)
5402           Do not warn when there is a conversion between pointers that have
5403           incompatible types.  This warning is for cases not covered by
5404           -Wno-pointer-sign, which warns for pointer argument passing or
5405           assignment with different signedness.
5406
5407       -Wno-int-conversion (C and Objective-C only)
5408           Do not warn about incompatible integer to pointer and pointer to
5409           integer conversions.  This warning is about implicit conversions;
5410           for explicit conversions the warnings -Wno-int-to-pointer-cast and
5411           -Wno-pointer-to-int-cast may be used.
5412
5413       -Wzero-length-bounds
5414           Warn about accesses to elements of zero-length array members that
5415           might overlap other members of the same object.  Declaring interior
5416           zero-length arrays is discouraged because accesses to them are
5417           undefined.  See
5418
5419           For example, the first two stores in function "bad" are diagnosed
5420           because the array elements overlap the subsequent members "b" and
5421           "c".  The third store is diagnosed by -Warray-bounds because it is
5422           beyond the bounds of the enclosing object.
5423
5424                   struct X { int a[0]; int b, c; };
5425                   struct X x;
5426
5427                   void bad (void)
5428                   {
5429                     x.a[0] = 0;   // -Wzero-length-bounds
5430                     x.a[1] = 1;   // -Wzero-length-bounds
5431                     x.a[2] = 2;   // -Warray-bounds
5432                   }
5433
5434           Option -Wzero-length-bounds is enabled by -Warray-bounds.
5435
5436       -Wno-div-by-zero
5437           Do not warn about compile-time integer division by zero.  Floating-
5438           point division by zero is not warned about, as it can be a
5439           legitimate way of obtaining infinities and NaNs.
5440
5441       -Wsystem-headers
5442           Print warning messages for constructs found in system header files.
5443           Warnings from system headers are normally suppressed, on the
5444           assumption that they usually do not indicate real problems and
5445           would only make the compiler output harder to read.  Using this
5446           command-line option tells GCC to emit warnings from system headers
5447           as if they occurred in user code.  However, note that using -Wall
5448           in conjunction with this option does not warn about unknown pragmas
5449           in system headers---for that, -Wunknown-pragmas must also be used.
5450
5451       -Wtautological-compare
5452           Warn if a self-comparison always evaluates to true or false.  This
5453           warning detects various mistakes such as:
5454
5455                   int i = 1;
5456                   ...
5457                   if (i > i) { ... }
5458
5459           This warning also warns about bitwise comparisons that always
5460           evaluate to true or false, for instance:
5461
5462                   if ((a & 16) == 10) { ... }
5463
5464           will always be false.
5465
5466           This warning is enabled by -Wall.
5467
5468       -Wtrampolines
5469           Warn about trampolines generated for pointers to nested functions.
5470           A trampoline is a small piece of data or code that is created at
5471           run time on the stack when the address of a nested function is
5472           taken, and is used to call the nested function indirectly.  For
5473           some targets, it is made up of data only and thus requires no
5474           special treatment.  But, for most targets, it is made up of code
5475           and thus requires the stack to be made executable in order for the
5476           program to work properly.
5477
5478       -Wfloat-equal
5479           Warn if floating-point values are used in equality comparisons.
5480
5481           The idea behind this is that sometimes it is convenient (for the
5482           programmer) to consider floating-point values as approximations to
5483           infinitely precise real numbers.  If you are doing this, then you
5484           need to compute (by analyzing the code, or in some other way) the
5485           maximum or likely maximum error that the computation introduces,
5486           and allow for it when performing comparisons (and when producing
5487           output, but that's a different problem).  In particular, instead of
5488           testing for equality, you should check to see whether the two
5489           values have ranges that overlap; and this is done with the
5490           relational operators, so equality comparisons are probably
5491           mistaken.
5492
5493       -Wtraditional (C and Objective-C only)
5494           Warn about certain constructs that behave differently in
5495           traditional and ISO C.  Also warn about ISO C constructs that have
5496           no traditional C equivalent, and/or problematic constructs that
5497           should be avoided.
5498
5499           *   Macro parameters that appear within string literals in the
5500               macro body.  In traditional C macro replacement takes place
5501               within string literals, but in ISO C it does not.
5502
5503           *   In traditional C, some preprocessor directives did not exist.
5504               Traditional preprocessors only considered a line to be a
5505               directive if the # appeared in column 1 on the line.  Therefore
5506               -Wtraditional warns about directives that traditional C
5507               understands but ignores because the # does not appear as the
5508               first character on the line.  It also suggests you hide
5509               directives like "#pragma" not understood by traditional C by
5510               indenting them.  Some traditional implementations do not
5511               recognize "#elif", so this option suggests avoiding it
5512               altogether.
5513
5514           *   A function-like macro that appears without arguments.
5515
5516           *   The unary plus operator.
5517
5518           *   The U integer constant suffix, or the F or L floating-point
5519               constant suffixes.  (Traditional C does support the L suffix on
5520               integer constants.)  Note, these suffixes appear in macros
5521               defined in the system headers of most modern systems, e.g. the
5522               _MIN/_MAX macros in "<limits.h>".  Use of these macros in user
5523               code might normally lead to spurious warnings, however GCC's
5524               integrated preprocessor has enough context to avoid warning in
5525               these cases.
5526
5527           *   A function declared external in one block and then used after
5528               the end of the block.
5529
5530           *   A "switch" statement has an operand of type "long".
5531
5532           *   A non-"static" function declaration follows a "static" one.
5533               This construct is not accepted by some traditional C compilers.
5534
5535           *   The ISO type of an integer constant has a different width or
5536               signedness from its traditional type.  This warning is only
5537               issued if the base of the constant is ten.  I.e. hexadecimal or
5538               octal values, which typically represent bit patterns, are not
5539               warned about.
5540
5541           *   Usage of ISO string concatenation is detected.
5542
5543           *   Initialization of automatic aggregates.
5544
5545           *   Identifier conflicts with labels.  Traditional C lacks a
5546               separate namespace for labels.
5547
5548           *   Initialization of unions.  If the initializer is zero, the
5549               warning is omitted.  This is done under the assumption that the
5550               zero initializer in user code appears conditioned on e.g.
5551               "__STDC__" to avoid missing initializer warnings and relies on
5552               default initialization to zero in the traditional C case.
5553
5554           *   Conversions by prototypes between fixed/floating-point values
5555               and vice versa.  The absence of these prototypes when compiling
5556               with traditional C causes serious problems.  This is a subset
5557               of the possible conversion warnings; for the full set use
5558               -Wtraditional-conversion.
5559
5560           *   Use of ISO C style function definitions.  This warning
5561               intentionally is not issued for prototype declarations or
5562               variadic functions because these ISO C features appear in your
5563               code when using libiberty's traditional C compatibility macros,
5564               "PARAMS" and "VPARAMS".  This warning is also bypassed for
5565               nested functions because that feature is already a GCC
5566               extension and thus not relevant to traditional C compatibility.
5567
5568       -Wtraditional-conversion (C and Objective-C only)
5569           Warn if a prototype causes a type conversion that is different from
5570           what would happen to the same argument in the absence of a
5571           prototype.  This includes conversions of fixed point to floating
5572           and vice versa, and conversions changing the width or signedness of
5573           a fixed-point argument except when the same as the default
5574           promotion.
5575
5576       -Wdeclaration-after-statement (C and Objective-C only)
5577           Warn when a declaration is found after a statement in a block.
5578           This construct, known from C++, was introduced with ISO C99 and is
5579           by default allowed in GCC.  It is not supported by ISO C90.
5580
5581       -Wshadow
5582           Warn whenever a local variable or type declaration shadows another
5583           variable, parameter, type, class member (in C++), or instance
5584           variable (in Objective-C) or whenever a built-in function is
5585           shadowed.  Note that in C++, the compiler warns if a local variable
5586           shadows an explicit typedef, but not if it shadows a
5587           struct/class/enum.  If this warning is enabled, it includes also
5588           all instances of local shadowing.  This means that
5589           -Wno-shadow=local and -Wno-shadow=compatible-local are ignored when
5590           -Wshadow is used.  Same as -Wshadow=global.
5591
5592       -Wno-shadow-ivar (Objective-C only)
5593           Do not warn whenever a local variable shadows an instance variable
5594           in an Objective-C method.
5595
5596       -Wshadow=global
5597           Warn for any shadowing.  Same as -Wshadow.
5598
5599       -Wshadow=local
5600           Warn when a local variable shadows another local variable or
5601           parameter.
5602
5603       -Wshadow=compatible-local
5604           Warn when a local variable shadows another local variable or
5605           parameter whose type is compatible with that of the shadowing
5606           variable.  In C++, type compatibility here means the type of the
5607           shadowing variable can be converted to that of the shadowed
5608           variable.  The creation of this flag (in addition to
5609           -Wshadow=local) is based on the idea that when a local variable
5610           shadows another one of incompatible type, it is most likely
5611           intentional, not a bug or typo, as shown in the following example:
5612
5613                   for (SomeIterator i = SomeObj.begin(); i != SomeObj.end(); ++i)
5614                   {
5615                     for (int i = 0; i < N; ++i)
5616                     {
5617                       ...
5618                     }
5619                     ...
5620                   }
5621
5622           Since the two variable "i" in the example above have incompatible
5623           types, enabling only -Wshadow=compatible-local does not emit a
5624           warning.  Because their types are incompatible, if a programmer
5625           accidentally uses one in place of the other, type checking is
5626           expected to catch that and emit an error or warning.  Use of this
5627           flag instead of -Wshadow=local can possibly reduce the number of
5628           warnings triggered by intentional shadowing.  Note that this also
5629           means that shadowing "const char *i" by "char *i" does not emit a
5630           warning.
5631
5632           This warning is also enabled by -Wshadow=local.
5633
5634       -Wlarger-than=byte-size
5635           Warn whenever an object is defined whose size exceeds byte-size.
5636           -Wlarger-than=PTRDIFF_MAX is enabled by default.  Warnings
5637           controlled by the option can be disabled either by specifying byte-
5638           size of SIZE_MAX or more or by -Wno-larger-than.
5639
5640       -Wno-larger-than
5641           Disable -Wlarger-than= warnings.  The option is equivalent to
5642           -Wlarger-than=SIZE_MAX or larger.
5643
5644       -Wframe-larger-than=byte-size
5645           Warn if the size of a function frame exceeds byte-size.  The
5646           computation done to determine the stack frame size is approximate
5647           and not conservative.  The actual requirements may be somewhat
5648           greater than byte-size even if you do not get a warning.  In
5649           addition, any space allocated via "alloca", variable-length arrays,
5650           or related constructs is not included by the compiler when
5651           determining whether or not to issue a warning.
5652           -Wframe-larger-than=PTRDIFF_MAX is enabled by default.  Warnings
5653           controlled by the option can be disabled either by specifying byte-
5654           size of SIZE_MAX or more or by -Wno-frame-larger-than.
5655
5656       -Wno-frame-larger-than
5657           Disable -Wframe-larger-than= warnings.  The option is equivalent to
5658           -Wframe-larger-than=SIZE_MAX or larger.
5659
5660       -Wno-free-nonheap-object
5661           Do not warn when attempting to free an object that was not
5662           allocated on the heap.
5663
5664       -Wstack-usage=byte-size
5665           Warn if the stack usage of a function might exceed byte-size.  The
5666           computation done to determine the stack usage is conservative.  Any
5667           space allocated via "alloca", variable-length arrays, or related
5668           constructs is included by the compiler when determining whether or
5669           not to issue a warning.
5670
5671           The message is in keeping with the output of -fstack-usage.
5672
5673           *   If the stack usage is fully static but exceeds the specified
5674               amount, it's:
5675
5676                         warning: stack usage is 1120 bytes
5677
5678           *   If the stack usage is (partly) dynamic but bounded, it's:
5679
5680                         warning: stack usage might be 1648 bytes
5681
5682           *   If the stack usage is (partly) dynamic and not bounded, it's:
5683
5684                         warning: stack usage might be unbounded
5685
5686           -Wstack-usage=PTRDIFF_MAX is enabled by default.  Warnings
5687           controlled by the option can be disabled either by specifying byte-
5688           size of SIZE_MAX or more or by -Wno-stack-usage.
5689
5690       -Wno-stack-usage
5691           Disable -Wstack-usage= warnings.  The option is equivalent to
5692           -Wstack-usage=SIZE_MAX or larger.
5693
5694       -Wunsafe-loop-optimizations
5695           Warn if the loop cannot be optimized because the compiler cannot
5696           assume anything on the bounds of the loop indices.  With
5697           -funsafe-loop-optimizations warn if the compiler makes such
5698           assumptions.
5699
5700       -Wno-pedantic-ms-format (MinGW targets only)
5701           When used in combination with -Wformat and -pedantic without GNU
5702           extensions, this option disables the warnings about non-ISO
5703           "printf" / "scanf" format width specifiers "I32", "I64", and "I"
5704           used on Windows targets, which depend on the MS runtime.
5705
5706       -Wpointer-arith
5707           Warn about anything that depends on the "size of" a function type
5708           or of "void".  GNU C assigns these types a size of 1, for
5709           convenience in calculations with "void *" pointers and pointers to
5710           functions.  In C++, warn also when an arithmetic operation involves
5711           "NULL".  This warning is also enabled by -Wpedantic.
5712
5713       -Wno-pointer-compare
5714           Do not warn if a pointer is compared with a zero character
5715           constant.  This usually means that the pointer was meant to be
5716           dereferenced.  For example:
5717
5718                   const char *p = foo ();
5719                   if (p == '\0')
5720                     return 42;
5721
5722           Note that the code above is invalid in C++11.
5723
5724           This warning is enabled by default.
5725
5726       -Wtype-limits
5727           Warn if a comparison is always true or always false due to the
5728           limited range of the data type, but do not warn for constant
5729           expressions.  For example, warn if an unsigned variable is compared
5730           against zero with "<" or ">=".  This warning is also enabled by
5731           -Wextra.
5732
5733       -Wabsolute-value (C and Objective-C only)
5734           Warn for calls to standard functions that compute the absolute
5735           value of an argument when a more appropriate standard function is
5736           available.  For example, calling "abs(3.14)" triggers the warning
5737           because the appropriate function to call to compute the absolute
5738           value of a double argument is "fabs".  The option also triggers
5739           warnings when the argument in a call to such a function has an
5740           unsigned type.  This warning can be suppressed with an explicit
5741           type cast and it is also enabled by -Wextra.
5742
5743       -Wcomment
5744       -Wcomments
5745           Warn whenever a comment-start sequence /* appears in a /* comment,
5746           or whenever a backslash-newline appears in a // comment.  This
5747           warning is enabled by -Wall.
5748
5749       -Wtrigraphs
5750           Warn if any trigraphs are encountered that might change the meaning
5751           of the program.  Trigraphs within comments are not warned about,
5752           except those that would form escaped newlines.
5753
5754           This option is implied by -Wall.  If -Wall is not given, this
5755           option is still enabled unless trigraphs are enabled.  To get
5756           trigraph conversion without warnings, but get the other -Wall
5757           warnings, use -trigraphs -Wall -Wno-trigraphs.
5758
5759       -Wundef
5760           Warn if an undefined identifier is evaluated in an "#if" directive.
5761           Such identifiers are replaced with zero.
5762
5763       -Wexpansion-to-defined
5764           Warn whenever defined is encountered in the expansion of a macro
5765           (including the case where the macro is expanded by an #if
5766           directive).  Such usage is not portable.  This warning is also
5767           enabled by -Wpedantic and -Wextra.
5768
5769       -Wunused-macros
5770           Warn about macros defined in the main file that are unused.  A
5771           macro is used if it is expanded or tested for existence at least
5772           once.  The preprocessor also warns if the macro has not been used
5773           at the time it is redefined or undefined.
5774
5775           Built-in macros, macros defined on the command line, and macros
5776           defined in include files are not warned about.
5777
5778           Note: If a macro is actually used, but only used in skipped
5779           conditional blocks, then the preprocessor reports it as unused.  To
5780           avoid the warning in such a case, you might improve the scope of
5781           the macro's definition by, for example, moving it into the first
5782           skipped block.  Alternatively, you could provide a dummy use with
5783           something like:
5784
5785                   #if defined the_macro_causing_the_warning
5786                   #endif
5787
5788       -Wno-endif-labels
5789           Do not warn whenever an "#else" or an "#endif" are followed by
5790           text.  This sometimes happens in older programs with code of the
5791           form
5792
5793                   #if FOO
5794                   ...
5795                   #else FOO
5796                   ...
5797                   #endif FOO
5798
5799           The second and third "FOO" should be in comments.  This warning is
5800           on by default.
5801
5802       -Wbad-function-cast (C and Objective-C only)
5803           Warn when a function call is cast to a non-matching type.  For
5804           example, warn if a call to a function returning an integer type is
5805           cast to a pointer type.
5806
5807       -Wc90-c99-compat (C and Objective-C only)
5808           Warn about features not present in ISO C90, but present in ISO C99.
5809           For instance, warn about use of variable length arrays, "long long"
5810           type, "bool" type, compound literals, designated initializers, and
5811           so on.  This option is independent of the standards mode.  Warnings
5812           are disabled in the expression that follows "__extension__".
5813
5814       -Wc99-c11-compat (C and Objective-C only)
5815           Warn about features not present in ISO C99, but present in ISO C11.
5816           For instance, warn about use of anonymous structures and unions,
5817           "_Atomic" type qualifier, "_Thread_local" storage-class specifier,
5818           "_Alignas" specifier, "Alignof" operator, "_Generic" keyword, and
5819           so on.  This option is independent of the standards mode.  Warnings
5820           are disabled in the expression that follows "__extension__".
5821
5822       -Wc11-c2x-compat (C and Objective-C only)
5823           Warn about features not present in ISO C11, but present in ISO C2X.
5824           For instance, warn about omitting the string in "_Static_assert",
5825           use of [[]] syntax for attributes, use of decimal floating-point
5826           types, and so on.  This option is independent of the standards
5827           mode.  Warnings are disabled in the expression that follows
5828           "__extension__".
5829
5830       -Wc++-compat (C and Objective-C only)
5831           Warn about ISO C constructs that are outside of the common subset
5832           of ISO C and ISO C++, e.g. request for implicit conversion from
5833           "void *" to a pointer to non-"void" type.
5834
5835       -Wc++11-compat (C++ and Objective-C++ only)
5836           Warn about C++ constructs whose meaning differs between ISO C++
5837           1998 and ISO C++ 2011, e.g., identifiers in ISO C++ 1998 that are
5838           keywords in ISO C++ 2011.  This warning turns on -Wnarrowing and is
5839           enabled by -Wall.
5840
5841       -Wc++14-compat (C++ and Objective-C++ only)
5842           Warn about C++ constructs whose meaning differs between ISO C++
5843           2011 and ISO C++ 2014.  This warning is enabled by -Wall.
5844
5845       -Wc++17-compat (C++ and Objective-C++ only)
5846           Warn about C++ constructs whose meaning differs between ISO C++
5847           2014 and ISO C++ 2017.  This warning is enabled by -Wall.
5848
5849       -Wc++20-compat (C++ and Objective-C++ only)
5850           Warn about C++ constructs whose meaning differs between ISO C++
5851           2017 and ISO C++ 2020.  This warning is enabled by -Wall.
5852
5853       -Wcast-qual
5854           Warn whenever a pointer is cast so as to remove a type qualifier
5855           from the target type.  For example, warn if a "const char *" is
5856           cast to an ordinary "char *".
5857
5858           Also warn when making a cast that introduces a type qualifier in an
5859           unsafe way.  For example, casting "char **" to "const char **" is
5860           unsafe, as in this example:
5861
5862                     /* p is char ** value.  */
5863                     const char **q = (const char **) p;
5864                     /* Assignment of readonly string to const char * is OK.  */
5865                     *q = "string";
5866                     /* Now char** pointer points to read-only memory.  */
5867                     **p = 'b';
5868
5869       -Wcast-align
5870           Warn whenever a pointer is cast such that the required alignment of
5871           the target is increased.  For example, warn if a "char *" is cast
5872           to an "int *" on machines where integers can only be accessed at
5873           two- or four-byte boundaries.
5874
5875       -Wcast-align=strict
5876           Warn whenever a pointer is cast such that the required alignment of
5877           the target is increased.  For example, warn if a "char *" is cast
5878           to an "int *" regardless of the target machine.
5879
5880       -Wcast-function-type
5881           Warn when a function pointer is cast to an incompatible function
5882           pointer.  In a cast involving function types with a variable
5883           argument list only the types of initial arguments that are provided
5884           are considered.  Any parameter of pointer-type matches any other
5885           pointer-type.  Any benign differences in integral types are
5886           ignored, like "int" vs. "long" on ILP32 targets.  Likewise type
5887           qualifiers are ignored.  The function type "void (*) (void)" is
5888           special and matches everything, which can be used to suppress this
5889           warning.  In a cast involving pointer to member types this warning
5890           warns whenever the type cast is changing the pointer to member
5891           type.  This warning is enabled by -Wextra.
5892
5893       -Wwrite-strings
5894           When compiling C, give string constants the type "const
5895           char[length]" so that copying the address of one into a non-"const"
5896           "char *" pointer produces a warning.  These warnings help you find
5897           at compile time code that can try to write into a string constant,
5898           but only if you have been very careful about using "const" in
5899           declarations and prototypes.  Otherwise, it is just a nuisance.
5900           This is why we did not make -Wall request these warnings.
5901
5902           When compiling C++, warn about the deprecated conversion from
5903           string literals to "char *".  This warning is enabled by default
5904           for C++ programs.
5905
5906       -Wclobbered
5907           Warn for variables that might be changed by "longjmp" or "vfork".
5908           This warning is also enabled by -Wextra.
5909
5910       -Wconversion
5911           Warn for implicit conversions that may alter a value. This includes
5912           conversions between real and integer, like "abs (x)" when "x" is
5913           "double"; conversions between signed and unsigned, like "unsigned
5914           ui = -1"; and conversions to smaller types, like "sqrtf (M_PI)". Do
5915           not warn for explicit casts like "abs ((int) x)" and "ui =
5916           (unsigned) -1", or if the value is not changed by the conversion
5917           like in "abs (2.0)".  Warnings about conversions between signed and
5918           unsigned integers can be disabled by using -Wno-sign-conversion.
5919
5920           For C++, also warn for confusing overload resolution for user-
5921           defined conversions; and conversions that never use a type
5922           conversion operator: conversions to "void", the same type, a base
5923           class or a reference to them. Warnings about conversions between
5924           signed and unsigned integers are disabled by default in C++ unless
5925           -Wsign-conversion is explicitly enabled.
5926
5927           Warnings about conversion from arithmetic on a small type back to
5928           that type are only given with -Warith-conversion.
5929
5930       -Wdangling-else
5931           Warn about constructions where there may be confusion to which "if"
5932           statement an "else" branch belongs.  Here is an example of such a
5933           case:
5934
5935                   {
5936                     if (a)
5937                       if (b)
5938                         foo ();
5939                     else
5940                       bar ();
5941                   }
5942
5943           In C/C++, every "else" branch belongs to the innermost possible
5944           "if" statement, which in this example is "if (b)".  This is often
5945           not what the programmer expected, as illustrated in the above
5946           example by indentation the programmer chose.  When there is the
5947           potential for this confusion, GCC issues a warning when this flag
5948           is specified.  To eliminate the warning, add explicit braces around
5949           the innermost "if" statement so there is no way the "else" can
5950           belong to the enclosing "if".  The resulting code looks like this:
5951
5952                   {
5953                     if (a)
5954                       {
5955                         if (b)
5956                           foo ();
5957                         else
5958                           bar ();
5959                       }
5960                   }
5961
5962           This warning is enabled by -Wparentheses.
5963
5964       -Wdate-time
5965           Warn when macros "__TIME__", "__DATE__" or "__TIMESTAMP__" are
5966           encountered as they might prevent bit-wise-identical reproducible
5967           compilations.
5968
5969       -Wempty-body
5970           Warn if an empty body occurs in an "if", "else" or "do while"
5971           statement.  This warning is also enabled by -Wextra.
5972
5973       -Wno-endif-labels
5974           Do not warn about stray tokens after "#else" and "#endif".
5975
5976       -Wenum-compare
5977           Warn about a comparison between values of different enumerated
5978           types.  In C++ enumerated type mismatches in conditional
5979           expressions are also diagnosed and the warning is enabled by
5980           default.  In C this warning is enabled by -Wall.
5981
5982       -Wenum-conversion (C, Objective-C only)
5983           Warn when a value of enumerated type is implicitly converted to a
5984           different enumerated type.  This warning is enabled by -Wextra.
5985
5986       -Wjump-misses-init (C, Objective-C only)
5987           Warn if a "goto" statement or a "switch" statement jumps forward
5988           across the initialization of a variable, or jumps backward to a
5989           label after the variable has been initialized.  This only warns
5990           about variables that are initialized when they are declared.  This
5991           warning is only supported for C and Objective-C; in C++ this sort
5992           of branch is an error in any case.
5993
5994           -Wjump-misses-init is included in -Wc++-compat.  It can be disabled
5995           with the -Wno-jump-misses-init option.
5996
5997       -Wsign-compare
5998           Warn when a comparison between signed and unsigned values could
5999           produce an incorrect result when the signed value is converted to
6000           unsigned.  In C++, this warning is also enabled by -Wall.  In C, it
6001           is also enabled by -Wextra.
6002
6003       -Wsign-conversion
6004           Warn for implicit conversions that may change the sign of an
6005           integer value, like assigning a signed integer expression to an
6006           unsigned integer variable. An explicit cast silences the warning.
6007           In C, this option is enabled also by -Wconversion.
6008
6009       -Wfloat-conversion
6010           Warn for implicit conversions that reduce the precision of a real
6011           value.  This includes conversions from real to integer, and from
6012           higher precision real to lower precision real values.  This option
6013           is also enabled by -Wconversion.
6014
6015       -Wno-scalar-storage-order
6016           Do not warn on suspicious constructs involving reverse scalar
6017           storage order.
6018
6019       -Wsizeof-pointer-div
6020           Warn for suspicious divisions of two sizeof expressions that divide
6021           the pointer size by the element size, which is the usual way to
6022           compute the array size but won't work out correctly with pointers.
6023           This warning warns e.g. about "sizeof (ptr) / sizeof (ptr[0])" if
6024           "ptr" is not an array, but a pointer.  This warning is enabled by
6025           -Wall.
6026
6027       -Wsizeof-pointer-memaccess
6028           Warn for suspicious length parameters to certain string and memory
6029           built-in functions if the argument uses "sizeof".  This warning
6030           triggers for example for "memset (ptr, 0, sizeof (ptr));" if "ptr"
6031           is not an array, but a pointer, and suggests a possible fix, or
6032           about "memcpy (&foo, ptr, sizeof (&foo));".
6033           -Wsizeof-pointer-memaccess also warns about calls to bounded string
6034           copy functions like "strncat" or "strncpy" that specify as the
6035           bound a "sizeof" expression of the source array.  For example, in
6036           the following function the call to "strncat" specifies the size of
6037           the source string as the bound.  That is almost certainly a mistake
6038           and so the call is diagnosed.
6039
6040                   void make_file (const char *name)
6041                   {
6042                     char path[PATH_MAX];
6043                     strncpy (path, name, sizeof path - 1);
6044                     strncat (path, ".text", sizeof ".text");
6045                     ...
6046                   }
6047
6048           The -Wsizeof-pointer-memaccess option is enabled by -Wall.
6049
6050       -Wno-sizeof-array-argument
6051           Do not warn when the "sizeof" operator is applied to a parameter
6052           that is declared as an array in a function definition.  This
6053           warning is enabled by default for C and C++ programs.
6054
6055       -Wmemset-elt-size
6056           Warn for suspicious calls to the "memset" built-in function, if the
6057           first argument references an array, and the third argument is a
6058           number equal to the number of elements, but not equal to the size
6059           of the array in memory.  This indicates that the user has omitted a
6060           multiplication by the element size.  This warning is enabled by
6061           -Wall.
6062
6063       -Wmemset-transposed-args
6064           Warn for suspicious calls to the "memset" built-in function where
6065           the second argument is not zero and the third argument is zero.
6066           For example, the call "memset (buf, sizeof buf, 0)" is diagnosed
6067           because "memset (buf, 0, sizeof buf)" was meant instead.  The
6068           diagnostic is only emitted if the third argument is a literal zero.
6069           Otherwise, if it is an expression that is folded to zero, or a cast
6070           of zero to some type, it is far less likely that the arguments have
6071           been mistakenly transposed and no warning is emitted.  This warning
6072           is enabled by -Wall.
6073
6074       -Waddress
6075           Warn about suspicious uses of memory addresses. These include using
6076           the address of a function in a conditional expression, such as
6077           "void func(void); if (func)", and comparisons against the memory
6078           address of a string literal, such as "if (x == "abc")".  Such uses
6079           typically indicate a programmer error: the address of a function
6080           always evaluates to true, so their use in a conditional usually
6081           indicate that the programmer forgot the parentheses in a function
6082           call; and comparisons against string literals result in unspecified
6083           behavior and are not portable in C, so they usually indicate that
6084           the programmer intended to use "strcmp".  This warning is enabled
6085           by -Wall.
6086
6087       -Wno-address-of-packed-member
6088           Do not warn when the address of packed member of struct or union is
6089           taken, which usually results in an unaligned pointer value.  This
6090           is enabled by default.
6091
6092       -Wlogical-op
6093           Warn about suspicious uses of logical operators in expressions.
6094           This includes using logical operators in contexts where a bit-wise
6095           operator is likely to be expected.  Also warns when the operands of
6096           a logical operator are the same:
6097
6098                   extern int a;
6099                   if (a < 0 && a < 0) { ... }
6100
6101       -Wlogical-not-parentheses
6102           Warn about logical not used on the left hand side operand of a
6103           comparison.  This option does not warn if the right operand is
6104           considered to be a boolean expression.  Its purpose is to detect
6105           suspicious code like the following:
6106
6107                   int a;
6108                   ...
6109                   if (!a > 1) { ... }
6110
6111           It is possible to suppress the warning by wrapping the LHS into
6112           parentheses:
6113
6114                   if ((!a) > 1) { ... }
6115
6116           This warning is enabled by -Wall.
6117
6118       -Waggregate-return
6119           Warn if any functions that return structures or unions are defined
6120           or called.  (In languages where you can return an array, this also
6121           elicits a warning.)
6122
6123       -Wno-aggressive-loop-optimizations
6124           Warn if in a loop with constant number of iterations the compiler
6125           detects undefined behavior in some statement during one or more of
6126           the iterations.
6127
6128       -Wno-attributes
6129           Do not warn if an unexpected "__attribute__" is used, such as
6130           unrecognized attributes, function attributes applied to variables,
6131           etc.  This does not stop errors for incorrect use of supported
6132           attributes.
6133
6134       -Wno-builtin-declaration-mismatch
6135           Warn if a built-in function is declared with an incompatible
6136           signature or as a non-function, or when a built-in function
6137           declared with a type that does not include a prototype is called
6138           with arguments whose promoted types do not match those expected by
6139           the function.  When -Wextra is specified, also warn when a built-in
6140           function that takes arguments is declared without a prototype.  The
6141           -Wbuiltin-declaration-mismatch warning is enabled by default.  To
6142           avoid the warning include the appropriate header to bring the
6143           prototypes of built-in functions into scope.
6144
6145           For example, the call to "memset" below is diagnosed by the warning
6146           because the function expects a value of type "size_t" as its
6147           argument but the type of 32 is "int".  With -Wextra, the
6148           declaration of the function is diagnosed as well.
6149
6150                   extern void* memset ();
6151                   void f (void *d)
6152                   {
6153                     memset (d, '\0', 32);
6154                   }
6155
6156       -Wno-builtin-macro-redefined
6157           Do not warn if certain built-in macros are redefined.  This
6158           suppresses warnings for redefinition of "__TIMESTAMP__",
6159           "__TIME__", "__DATE__", "__FILE__", and "__BASE_FILE__".
6160
6161       -Wstrict-prototypes (C and Objective-C only)
6162           Warn if a function is declared or defined without specifying the
6163           argument types.  (An old-style function definition is permitted
6164           without a warning if preceded by a declaration that specifies the
6165           argument types.)
6166
6167       -Wold-style-declaration (C and Objective-C only)
6168           Warn for obsolescent usages, according to the C Standard, in a
6169           declaration. For example, warn if storage-class specifiers like
6170           "static" are not the first things in a declaration.  This warning
6171           is also enabled by -Wextra.
6172
6173       -Wold-style-definition (C and Objective-C only)
6174           Warn if an old-style function definition is used.  A warning is
6175           given even if there is a previous prototype.  A definition using ()
6176           is not considered an old-style definition in C2X mode, because it
6177           is equivalent to (void) in that case, but is considered an old-
6178           style definition for older standards.
6179
6180       -Wmissing-parameter-type (C and Objective-C only)
6181           A function parameter is declared without a type specifier in
6182           K&R-style functions:
6183
6184                   void foo(bar) { }
6185
6186           This warning is also enabled by -Wextra.
6187
6188       -Wmissing-prototypes (C and Objective-C only)
6189           Warn if a global function is defined without a previous prototype
6190           declaration.  This warning is issued even if the definition itself
6191           provides a prototype.  Use this option to detect global functions
6192           that do not have a matching prototype declaration in a header file.
6193           This option is not valid for C++ because all function declarations
6194           provide prototypes and a non-matching declaration declares an
6195           overload rather than conflict with an earlier declaration.  Use
6196           -Wmissing-declarations to detect missing declarations in C++.
6197
6198       -Wmissing-declarations
6199           Warn if a global function is defined without a previous
6200           declaration.  Do so even if the definition itself provides a
6201           prototype.  Use this option to detect global functions that are not
6202           declared in header files.  In C, no warnings are issued for
6203           functions with previous non-prototype declarations; use
6204           -Wmissing-prototypes to detect missing prototypes.  In C++, no
6205           warnings are issued for function templates, or for inline
6206           functions, or for functions in anonymous namespaces.
6207
6208       -Wmissing-field-initializers
6209           Warn if a structure's initializer has some fields missing.  For
6210           example, the following code causes such a warning, because "x.h" is
6211           implicitly zero:
6212
6213                   struct s { int f, g, h; };
6214                   struct s x = { 3, 4 };
6215
6216           This option does not warn about designated initializers, so the
6217           following modification does not trigger a warning:
6218
6219                   struct s { int f, g, h; };
6220                   struct s x = { .f = 3, .g = 4 };
6221
6222           In C this option does not warn about the universal zero initializer
6223           { 0 }:
6224
6225                   struct s { int f, g, h; };
6226                   struct s x = { 0 };
6227
6228           Likewise, in C++ this option does not warn about the empty { }
6229           initializer, for example:
6230
6231                   struct s { int f, g, h; };
6232                   s x = { };
6233
6234           This warning is included in -Wextra.  To get other -Wextra warnings
6235           without this one, use -Wextra -Wno-missing-field-initializers.
6236
6237       -Wno-multichar
6238           Do not warn if a multicharacter constant ('FOOF') is used.  Usually
6239           they indicate a typo in the user's code, as they have
6240           implementation-defined values, and should not be used in portable
6241           code.
6242
6243       -Wnormalized=[none|id|nfc|nfkc]
6244           In ISO C and ISO C++, two identifiers are different if they are
6245           different sequences of characters.  However, sometimes when
6246           characters outside the basic ASCII character set are used, you can
6247           have two different character sequences that look the same.  To
6248           avoid confusion, the ISO 10646 standard sets out some normalization
6249           rules which when applied ensure that two sequences that look the
6250           same are turned into the same sequence.  GCC can warn you if you
6251           are using identifiers that have not been normalized; this option
6252           controls that warning.
6253
6254           There are four levels of warning supported by GCC.  The default is
6255           -Wnormalized=nfc, which warns about any identifier that is not in
6256           the ISO 10646 "C" normalized form, NFC.  NFC is the recommended
6257           form for most uses.  It is equivalent to -Wnormalized.
6258
6259           Unfortunately, there are some characters allowed in identifiers by
6260           ISO C and ISO C++ that, when turned into NFC, are not allowed in
6261           identifiers.  That is, there's no way to use these symbols in
6262           portable ISO C or C++ and have all your identifiers in NFC.
6263           -Wnormalized=id suppresses the warning for these characters.  It is
6264           hoped that future versions of the standards involved will correct
6265           this, which is why this option is not the default.
6266
6267           You can switch the warning off for all characters by writing
6268           -Wnormalized=none or -Wno-normalized.  You should only do this if
6269           you are using some other normalization scheme (like "D"), because
6270           otherwise you can easily create bugs that are literally impossible
6271           to see.
6272
6273           Some characters in ISO 10646 have distinct meanings but look
6274           identical in some fonts or display methodologies, especially once
6275           formatting has been applied.  For instance "\u207F", "SUPERSCRIPT
6276           LATIN SMALL LETTER N", displays just like a regular "n" that has
6277           been placed in a superscript.  ISO 10646 defines the NFKC
6278           normalization scheme to convert all these into a standard form as
6279           well, and GCC warns if your code is not in NFKC if you use
6280           -Wnormalized=nfkc.  This warning is comparable to warning about
6281           every identifier that contains the letter O because it might be
6282           confused with the digit 0, and so is not the default, but may be
6283           useful as a local coding convention if the programming environment
6284           cannot be fixed to display these characters distinctly.
6285
6286       -Wno-attribute-warning
6287           Do not warn about usage of functions declared with "warning"
6288           attribute.  By default, this warning is enabled.
6289           -Wno-attribute-warning can be used to disable the warning or
6290           -Wno-error=attribute-warning can be used to disable the error when
6291           compiled with -Werror flag.
6292
6293       -Wno-deprecated
6294           Do not warn about usage of deprecated features.
6295
6296       -Wno-deprecated-declarations
6297           Do not warn about uses of functions, variables, and types marked as
6298           deprecated by using the "deprecated" attribute.
6299
6300       -Wno-overflow
6301           Do not warn about compile-time overflow in constant expressions.
6302
6303       -Wno-odr
6304           Warn about One Definition Rule violations during link-time
6305           optimization.  Enabled by default.
6306
6307       -Wopenmp-simd
6308           Warn if the vectorizer cost model overrides the OpenMP simd
6309           directive set by user.  The -fsimd-cost-model=unlimited option can
6310           be used to relax the cost model.
6311
6312       -Woverride-init (C and Objective-C only)
6313           Warn if an initialized field without side effects is overridden
6314           when using designated initializers.
6315
6316           This warning is included in -Wextra.  To get other -Wextra warnings
6317           without this one, use -Wextra -Wno-override-init.
6318
6319       -Wno-override-init-side-effects (C and Objective-C only)
6320           Do not warn if an initialized field with side effects is overridden
6321           when using designated initializers.  This warning is enabled by
6322           default.
6323
6324       -Wpacked
6325           Warn if a structure is given the packed attribute, but the packed
6326           attribute has no effect on the layout or size of the structure.
6327           Such structures may be mis-aligned for little benefit.  For
6328           instance, in this code, the variable "f.x" in "struct bar" is
6329           misaligned even though "struct bar" does not itself have the packed
6330           attribute:
6331
6332                   struct foo {
6333                     int x;
6334                     char a, b, c, d;
6335                   } __attribute__((packed));
6336                   struct bar {
6337                     char z;
6338                     struct foo f;
6339                   };
6340
6341       -Wnopacked-bitfield-compat
6342           The 4.1, 4.2 and 4.3 series of GCC ignore the "packed" attribute on
6343           bit-fields of type "char".  This was fixed in GCC 4.4 but the
6344           change can lead to differences in the structure layout.  GCC
6345           informs you when the offset of such a field has changed in GCC 4.4.
6346           For example there is no longer a 4-bit padding between field "a"
6347           and "b" in this structure:
6348
6349                   struct foo
6350                   {
6351                     char a:4;
6352                     char b:8;
6353                   } __attribute__ ((packed));
6354
6355           This warning is enabled by default.  Use
6356           -Wno-packed-bitfield-compat to disable this warning.
6357
6358       -Wpacked-not-aligned (C, C++, Objective-C and Objective-C++ only)
6359           Warn if a structure field with explicitly specified alignment in a
6360           packed struct or union is misaligned.  For example, a warning will
6361           be issued on "struct S", like, "warning: alignment 1 of 'struct S'
6362           is less than 8", in this code:
6363
6364                   struct __attribute__ ((aligned (8))) S8 { char a[8]; };
6365                   struct __attribute__ ((packed)) S {
6366                     struct S8 s8;
6367                   };
6368
6369           This warning is enabled by -Wall.
6370
6371       -Wpadded
6372           Warn if padding is included in a structure, either to align an
6373           element of the structure or to align the whole structure.
6374           Sometimes when this happens it is possible to rearrange the fields
6375           of the structure to reduce the padding and so make the structure
6376           smaller.
6377
6378       -Wredundant-decls
6379           Warn if anything is declared more than once in the same scope, even
6380           in cases where multiple declaration is valid and changes nothing.
6381
6382       -Wrestrict
6383           Warn when an object referenced by a "restrict"-qualified parameter
6384           (or, in C++, a "__restrict"-qualified parameter) is aliased by
6385           another argument, or when copies between such objects overlap.  For
6386           example, the call to the "strcpy" function below attempts to
6387           truncate the string by replacing its initial characters with the
6388           last four.  However, because the call writes the terminating NUL
6389           into "a[4]", the copies overlap and the call is diagnosed.
6390
6391                   void foo (void)
6392                   {
6393                     char a[] = "abcd1234";
6394                     strcpy (a, a + 4);
6395                     ...
6396                   }
6397
6398           The -Wrestrict option detects some instances of simple overlap even
6399           without optimization but works best at -O2 and above.  It is
6400           included in -Wall.
6401
6402       -Wnested-externs (C and Objective-C only)
6403           Warn if an "extern" declaration is encountered within a function.
6404
6405       -Winline
6406           Warn if a function that is declared as inline cannot be inlined.
6407           Even with this option, the compiler does not warn about failures to
6408           inline functions declared in system headers.
6409
6410           The compiler uses a variety of heuristics to determine whether or
6411           not to inline a function.  For example, the compiler takes into
6412           account the size of the function being inlined and the amount of
6413           inlining that has already been done in the current function.
6414           Therefore, seemingly insignificant changes in the source program
6415           can cause the warnings produced by -Winline to appear or disappear.
6416
6417       -Wint-in-bool-context
6418           Warn for suspicious use of integer values where boolean values are
6419           expected, such as conditional expressions (?:) using non-boolean
6420           integer constants in boolean context, like "if (a <= b ? 2 : 3)".
6421           Or left shifting of signed integers in boolean context, like "for
6422           (a = 0; 1 << a; a++);".  Likewise for all kinds of multiplications
6423           regardless of the data type.  This warning is enabled by -Wall.
6424
6425       -Wno-int-to-pointer-cast
6426           Suppress warnings from casts to pointer type of an integer of a
6427           different size. In C++, casting to a pointer type of smaller size
6428           is an error. Wint-to-pointer-cast is enabled by default.
6429
6430       -Wno-pointer-to-int-cast (C and Objective-C only)
6431           Suppress warnings from casts from a pointer to an integer type of a
6432           different size.
6433
6434       -Winvalid-pch
6435           Warn if a precompiled header is found in the search path but cannot
6436           be used.
6437
6438       -Wlong-long
6439           Warn if "long long" type is used.  This is enabled by either
6440           -Wpedantic or -Wtraditional in ISO C90 and C++98 modes.  To inhibit
6441           the warning messages, use -Wno-long-long.
6442
6443       -Wvariadic-macros
6444           Warn if variadic macros are used in ISO C90 mode, or if the GNU
6445           alternate syntax is used in ISO C99 mode.  This is enabled by
6446           either -Wpedantic or -Wtraditional.  To inhibit the warning
6447           messages, use -Wno-variadic-macros.
6448
6449       -Wno-varargs
6450           Do not warn upon questionable usage of the macros used to handle
6451           variable arguments like "va_start".  These warnings are enabled by
6452           default.
6453
6454       -Wvector-operation-performance
6455           Warn if vector operation is not implemented via SIMD capabilities
6456           of the architecture.  Mainly useful for the performance tuning.
6457           Vector operation can be implemented "piecewise", which means that
6458           the scalar operation is performed on every vector element; "in
6459           parallel", which means that the vector operation is implemented
6460           using scalars of wider type, which normally is more performance
6461           efficient; and "as a single scalar", which means that vector fits
6462           into a scalar type.
6463
6464       -Wvla
6465           Warn if a variable-length array is used in the code.  -Wno-vla
6466           prevents the -Wpedantic warning of the variable-length array.
6467
6468       -Wvla-larger-than=byte-size
6469           If this option is used, the compiler warns for declarations of
6470           variable-length arrays whose size is either unbounded, or bounded
6471           by an argument that allows the array size to exceed byte-size
6472           bytes.  This is similar to how -Walloca-larger-than=byte-size
6473           works, but with variable-length arrays.
6474
6475           Note that GCC may optimize small variable-length arrays of a known
6476           value into plain arrays, so this warning may not get triggered for
6477           such arrays.
6478
6479           -Wvla-larger-than=PTRDIFF_MAX is enabled by default but is
6480           typically only effective when -ftree-vrp is active (default for -O2
6481           and above).
6482
6483           See also -Walloca-larger-than=byte-size.
6484
6485       -Wno-vla-larger-than
6486           Disable -Wvla-larger-than= warnings.  The option is equivalent to
6487           -Wvla-larger-than=SIZE_MAX or larger.
6488
6489       -Wvolatile-register-var
6490           Warn if a register variable is declared volatile.  The volatile
6491           modifier does not inhibit all optimizations that may eliminate
6492           reads and/or writes to register variables.  This warning is enabled
6493           by -Wall.
6494
6495       -Wdisabled-optimization
6496           Warn if a requested optimization pass is disabled.  This warning
6497           does not generally indicate that there is anything wrong with your
6498           code; it merely indicates that GCC's optimizers are unable to
6499           handle the code effectively.  Often, the problem is that your code
6500           is too big or too complex; GCC refuses to optimize programs when
6501           the optimization itself is likely to take inordinate amounts of
6502           time.
6503
6504       -Wpointer-sign (C and Objective-C only)
6505           Warn for pointer argument passing or assignment with different
6506           signedness.  This option is only supported for C and Objective-C.
6507           It is implied by -Wall and by -Wpedantic, which can be disabled
6508           with -Wno-pointer-sign.
6509
6510       -Wstack-protector
6511           This option is only active when -fstack-protector is active.  It
6512           warns about functions that are not protected against stack
6513           smashing.
6514
6515       -Woverlength-strings
6516           Warn about string constants that are longer than the "minimum
6517           maximum" length specified in the C standard.  Modern compilers
6518           generally allow string constants that are much longer than the
6519           standard's minimum limit, but very portable programs should avoid
6520           using longer strings.
6521
6522           The limit applies after string constant concatenation, and does not
6523           count the trailing NUL.  In C90, the limit was 509 characters; in
6524           C99, it was raised to 4095.  C++98 does not specify a normative
6525           minimum maximum, so we do not diagnose overlength strings in C++.
6526
6527           This option is implied by -Wpedantic, and can be disabled with
6528           -Wno-overlength-strings.
6529
6530       -Wunsuffixed-float-constants (C and Objective-C only)
6531           Issue a warning for any floating constant that does not have a
6532           suffix.  When used together with -Wsystem-headers it warns about
6533           such constants in system header files.  This can be useful when
6534           preparing code to use with the "FLOAT_CONST_DECIMAL64" pragma from
6535           the decimal floating-point extension to C99.
6536
6537       -Wno-lto-type-mismatch
6538           During the link-time optimization, do not warn about type
6539           mismatches in global declarations from different compilation units.
6540           Requires -flto to be enabled.  Enabled by default.
6541
6542       -Wno-designated-init (C and Objective-C only)
6543           Suppress warnings when a positional initializer is used to
6544           initialize a structure that has been marked with the
6545           "designated_init" attribute.
6546
6547       -Wno-hsa
6548           Do not warn when HSAIL cannot be emitted for the compiled function
6549           or OpenMP construct.  These warnings are enabled by default.
6550
6551   Options That Control Static Analysis
6552       -fanalyzer
6553           This option enables an static analysis of program flow which looks
6554           for "interesting" interprocedural paths through the code, and
6555           issues warnings for problems found on them.
6556
6557           This analysis is much more expensive than other GCC warnings.
6558
6559           Enabling this option effectively enables the following warnings:
6560
6561           -Wanalyzer-double-fclose -Wanalyzer-double-free
6562           -Wanalyzer-exposure-through-output-file -Wanalyzer-file-leak
6563           -Wanalyzer-free-of-non-heap -Wanalyzer-malloc-leak
6564           -Wanalyzer-possible-null-argument
6565           -Wanalyzer-possible-null-dereference -Wanalyzer-null-argument
6566           -Wanalyzer-null-dereference -Wanalyzer-stale-setjmp-buffer
6567           -Wanalyzer-tainted-array-index
6568           -Wanalyzer-unsafe-call-within-signal-handler
6569           -Wanalyzer-use-after-free
6570           -Wanalyzer-use-of-pointer-in-stale-stack-frame
6571
6572           This option is only available if GCC was configured with analyzer
6573           support enabled.
6574
6575       -Wanalyzer-too-complex
6576           If -fanalyzer is enabled, the analyzer uses various heuristics to
6577           attempt to explore the control flow and data flow in the program,
6578           but these can be defeated by sufficiently complicated code.
6579
6580           By default, the analysis silently stops if the code is too
6581           complicated for the analyzer to fully explore and it reaches an
6582           internal limit.  The -Wanalyzer-too-complex option warns if this
6583           occurs.
6584
6585       -Wno-analyzer-double-fclose
6586           This warning requires -fanalyzer, which enables it; use
6587           -Wno-analyzer-double-fclose to disable it.
6588
6589           This diagnostic warns for paths through the code in which a "FILE
6590           *" can have "fclose" called on it more than once.
6591
6592       -Wno-analyzer-double-free
6593           This warning requires -fanalyzer, which enables it; use
6594           -Wno-analyzer-double-free to disable it.
6595
6596           This diagnostic warns for paths through the code in which a pointer
6597           can have "free" called on it more than once.
6598
6599       -Wno-analyzer-exposure-through-output-file
6600           This warning requires -fanalyzer, which enables it; use
6601           -Wno-analyzer-exposure-through-output-file to disable it.
6602
6603           This diagnostic warns for paths through the code in which a
6604           security-sensitive value is written to an output file (such as
6605           writing a password to a log file).
6606
6607       -Wno-analyzer-file-leak
6608           This warning requires -fanalyzer, which enables it; use
6609           -Wno-analyzer-file-leak to disable it.
6610
6611           This diagnostic warns for paths through the code in which a
6612           "<stdio.h>" "FILE *" stream object is leaked.
6613
6614       -Wno-analyzer-free-of-non-heap
6615           This warning requires -fanalyzer, which enables it; use
6616           -Wno-analyzer-free-of-non-heap to disable it.
6617
6618           This diagnostic warns for paths through the code in which "free" is
6619           called on a non-heap pointer (e.g. an on-stack buffer, or a
6620           global).
6621
6622       -Wno-analyzer-malloc-leak
6623           This warning requires -fanalyzer, which enables it; use
6624           -Wno-analyzer-malloc-leak to disable it.
6625
6626           This diagnostic warns for paths through the code in which a pointer
6627           allocated via "malloc" is leaked.
6628
6629       -Wno-analyzer-possible-null-argument
6630           This warning requires -fanalyzer, which enables it; use
6631           -Wno-analyzer-possible-null-argument to disable it.
6632
6633           This diagnostic warns for paths through the code in which a
6634           possibly-NULL value is passed to a function argument marked with
6635           "__attribute__((nonnull))" as requiring a non-NULL value.
6636
6637       -Wno-analyzer-possible-null-dereference
6638           This warning requires -fanalyzer, which enables it; use
6639           -Wno-analyzer-possible-null-dereference to disable it.
6640
6641           This diagnostic warns for paths through the code in which a
6642           possibly-NULL value is dereferenced.
6643
6644       -Wno-analyzer-null-argument
6645           This warning requires -fanalyzer, which enables it; use
6646           -Wno-analyzer-null-argument to disable it.
6647
6648           This diagnostic warns for paths through the code in which a value
6649           known to be NULL is passed to a function argument marked with
6650           "__attribute__((nonnull))" as requiring a non-NULL value.
6651
6652       -Wno-analyzer-null-dereference
6653           This warning requires -fanalyzer, which enables it; use
6654           -Wno-analyzer-null-dereference to disable it.
6655
6656           This diagnostic warns for paths through the code in which a value
6657           known to be NULL is dereferenced.
6658
6659       -Wno-analyzer-stale-setjmp-buffer
6660           This warning requires -fanalyzer, which enables it; use
6661           -Wno-analyzer-stale-setjmp-buffer to disable it.
6662
6663           This diagnostic warns for paths through the code in which "longjmp"
6664           is called to rewind to a "jmp_buf" relating to a "setjmp" call in a
6665           function that has returned.
6666
6667           When "setjmp" is called on a "jmp_buf" to record a rewind location,
6668           it records the stack frame.  The stack frame becomes invalid when
6669           the function containing the "setjmp" call returns.  Attempting to
6670           rewind to it via "longjmp" would reference a stack frame that no
6671           longer exists, and likely lead to a crash (or worse).
6672
6673       -Wno-analyzer-tainted-array-index
6674           This warning requires both -fanalyzer and -fanalyzer-checker=taint
6675           to enable it; use -Wno-analyzer-tainted-array-index to disable it.
6676
6677           This diagnostic warns for paths through the code in which a value
6678           that could be under an attacker's control is used as the index of
6679           an array access without being sanitized.
6680
6681       -Wno-analyzer-unsafe-call-within-signal-handler
6682           This warning requires -fanalyzer, which enables it; use
6683           -Wno-analyzer-unsafe-call-within-signal-handler to disable it.
6684
6685           This diagnostic warns for paths through the code in which a
6686           function known to be async-signal-unsafe (such as "fprintf") is
6687           called from a signal handler.
6688
6689       -Wno-analyzer-use-after-free
6690           This warning requires -fanalyzer, which enables it; use
6691           -Wno-analyzer-use-after-free to disable it.
6692
6693           This diagnostic warns for paths through the code in which a pointer
6694           is used after "free" is called on it.
6695
6696       -Wno-analyzer-use-of-pointer-in-stale-stack-frame
6697           This warning requires -fanalyzer, which enables it; use
6698           -Wno-analyzer-use-of-pointer-in-stale-stack-frame to disable it.
6699
6700           This diagnostic warns for paths through the code in which a pointer
6701           is dereferenced that points to a variable in a stale stack frame.
6702
6703       Pertinent parameters for controlling the exploration are: --param
6704       analyzer-bb-explosion-factor=value, --param
6705       analyzer-max-enodes-per-program-point=value, --param
6706       analyzer-max-recursion-depth=value, and --param
6707       analyzer-min-snodes-for-call-summary=value.
6708
6709       The following options control the analyzer.
6710
6711       -fanalyzer-call-summaries
6712           Simplify interprocedural analysis by computing the effect of
6713           certain calls, rather than exploring all paths through the function
6714           from callsite to each possible return.
6715
6716           If enabled, call summaries are only used for functions with more
6717           than one call site, and that are sufficiently complicated (as per
6718           --param analyzer-min-snodes-for-call-summary=value).
6719
6720       -fanalyzer-checker=name
6721           Restrict the analyzer to run just the named checker, and enable it.
6722
6723           Some checkers are disabled by default (even with -fanalyzer), such
6724           as the "taint" checker that implements
6725           -Wanalyzer-tainted-array-index, and this option is required to
6726           enable them.
6727
6728       -fanalyzer-fine-grained
6729           This option is intended for analyzer developers.
6730
6731           Internally the analyzer builds an "exploded graph" that combines
6732           control flow graphs with data flow information.
6733
6734           By default, an edge in this graph can contain the effects of a run
6735           of multiple statements within a basic block.  With
6736           -fanalyzer-fine-grained, each statement gets its own edge.
6737
6738       -fanalyzer-show-duplicate-count
6739           This option is intended for analyzer developers: if multiple
6740           diagnostics have been detected as being duplicates of each other,
6741           it emits a note when reporting the best diagnostic, giving the
6742           number of additional diagnostics that were suppressed by the
6743           deduplication logic.
6744
6745       -fno-analyzer-state-merge
6746           This option is intended for analyzer developers.
6747
6748           By default the analyzer attempts to simplify analysis by merging
6749           sufficiently similar states at each program point as it builds its
6750           "exploded graph".  With -fno-analyzer-state-merge this merging can
6751           be suppressed, for debugging state-handling issues.
6752
6753       -fno-analyzer-state-purge
6754           This option is intended for analyzer developers.
6755
6756           By default the analyzer attempts to simplify analysis by purging
6757           aspects of state at a program point that appear to no longer be
6758           relevant e.g. the values of locals that aren't accessed later in
6759           the function and which aren't relevant to leak analysis.
6760
6761           With -fno-analyzer-state-purge this purging of state can be
6762           suppressed, for debugging state-handling issues.
6763
6764       -fanalyzer-transitivity
6765           This option enables transitivity of constraints within the
6766           analyzer.
6767
6768       -fanalyzer-verbose-edges
6769           This option is intended for analyzer developers.  It enables more
6770           verbose, lower-level detail in the descriptions of control flow
6771           within diagnostic paths.
6772
6773       -fanalyzer-verbose-state-changes
6774           This option is intended for analyzer developers.  It enables more
6775           verbose, lower-level detail in the descriptions of events relating
6776           to state machines within diagnostic paths.
6777
6778       -fanalyzer-verbosity=level
6779           This option controls the complexity of the control flow paths that
6780           are emitted for analyzer diagnostics.
6781
6782           The level can be one of:
6783
6784           0   At this level, interprocedural call and return events are
6785               displayed, along with the most pertinent state-change events
6786               relating to a diagnostic.  For example, for a double-"free"
6787               diagnostic, both calls to "free" will be shown.
6788
6789           1   As per the previous level, but also show events for the entry
6790               to each function.
6791
6792           2   As per the previous level, but also show events relating to
6793               control flow that are significant to triggering the issue (e.g.
6794               "true path taken" at a conditional).
6795
6796               This level is the default.
6797
6798           3   As per the previous level, but show all control flow events,
6799               not just significant ones.
6800
6801           4   This level is intended for analyzer developers; it adds various
6802               other events intended for debugging the analyzer.
6803
6804       -fdump-analyzer
6805           Dump internal details about what the analyzer is doing to
6806           file.analyzer.txt.  This option is overridden by
6807           -fdump-analyzer-stderr.
6808
6809       -fdump-analyzer-stderr
6810           Dump internal details about what the analyzer is doing to stderr.
6811           This option overrides -fdump-analyzer.
6812
6813       -fdump-analyzer-callgraph
6814           Dump a representation of the call graph suitable for viewing with
6815           GraphViz to file.callgraph.dot.
6816
6817       -fdump-analyzer-exploded-graph
6818           Dump a representation of the "exploded graph" suitable for viewing
6819           with GraphViz to file.eg.dot.  Nodes are color-coded based on
6820           state-machine states to emphasize state changes.
6821
6822       -fdump-analyzer-exploded-nodes
6823           Emit diagnostics showing where nodes in the "exploded graph" are in
6824           relation to the program source.
6825
6826       -fdump-analyzer-exploded-nodes-2
6827           Dump a textual representation of the "exploded graph" to
6828           file.eg.txt.
6829
6830       -fdump-analyzer-exploded-nodes-3
6831           Dump a textual representation of the "exploded graph" to one dump
6832           file per node, to file.eg-id.txt.  This is typically a large number
6833           of dump files.
6834
6835       -fdump-analyzer-state-purge
6836           As per -fdump-analyzer-supergraph, dump a representation of the
6837           "supergraph" suitable for viewing with GraphViz, but annotate the
6838           graph with information on what state will be purged at each node.
6839           The graph is written to file.state-purge.dot.
6840
6841       -fdump-analyzer-supergraph
6842           Dump representations of the "supergraph" suitable for viewing with
6843           GraphViz to file.supergraph.dot and to file.supergraph-eg.dot.
6844           These show all of the control flow graphs in the program, with
6845           interprocedural edges for calls and returns.  The second dump
6846           contains annotations showing nodes in the "exploded graph" and
6847           diagnostics associated with them.
6848
6849   Options for Debugging Your Program
6850       To tell GCC to emit extra information for use by a debugger, in almost
6851       all cases you need only to add -g to your other options.
6852
6853       GCC allows you to use -g with -O.  The shortcuts taken by optimized
6854       code may occasionally be surprising: some variables you declared may
6855       not exist at all; flow of control may briefly move where you did not
6856       expect it; some statements may not be executed because they compute
6857       constant results or their values are already at hand; some statements
6858       may execute in different places because they have been moved out of
6859       loops.  Nevertheless it is possible to debug optimized output.  This
6860       makes it reasonable to use the optimizer for programs that might have
6861       bugs.
6862
6863       If you are not using some other optimization option, consider using -Og
6864       with -g.  With no -O option at all, some compiler passes that collect
6865       information useful for debugging do not run at all, so that -Og may
6866       result in a better debugging experience.
6867
6868       -g  Produce debugging information in the operating system's native
6869           format (stabs, COFF, XCOFF, or DWARF).  GDB can work with this
6870           debugging information.
6871
6872           On most systems that use stabs format, -g enables use of extra
6873           debugging information that only GDB can use; this extra information
6874           makes debugging work better in GDB but probably makes other
6875           debuggers crash or refuse to read the program.  If you want to
6876           control for certain whether to generate the extra information, use
6877           -gstabs+, -gstabs, -gxcoff+, -gxcoff, or -gvms (see below).
6878
6879       -ggdb
6880           Produce debugging information for use by GDB.  This means to use
6881           the most expressive format available (DWARF, stabs, or the native
6882           format if neither of those are supported), including GDB extensions
6883           if at all possible.
6884
6885       -gdwarf
6886       -gdwarf-version
6887           Produce debugging information in DWARF format (if that is
6888           supported).  The value of version may be either 2, 3, 4 or 5; the
6889           default version for most targets is 4.  DWARF Version 5 is only
6890           experimental.
6891
6892           Note that with DWARF Version 2, some ports require and always use
6893           some non-conflicting DWARF 3 extensions in the unwind tables.
6894
6895           Version 4 may require GDB 7.0 and -fvar-tracking-assignments for
6896           maximum benefit.
6897
6898           GCC no longer supports DWARF Version 1, which is substantially
6899           different than Version 2 and later.  For historical reasons, some
6900           other DWARF-related options such as -fno-dwarf2-cfi-asm) retain a
6901           reference to DWARF Version 2 in their names, but apply to all
6902           currently-supported versions of DWARF.
6903
6904       -gstabs
6905           Produce debugging information in stabs format (if that is
6906           supported), without GDB extensions.  This is the format used by DBX
6907           on most BSD systems.  On MIPS, Alpha and System V Release 4 systems
6908           this option produces stabs debugging output that is not understood
6909           by DBX.  On System V Release 4 systems this option requires the GNU
6910           assembler.
6911
6912       -gstabs+
6913           Produce debugging information in stabs format (if that is
6914           supported), using GNU extensions understood only by the GNU
6915           debugger (GDB).  The use of these extensions is likely to make
6916           other debuggers crash or refuse to read the program.
6917
6918       -gxcoff
6919           Produce debugging information in XCOFF format (if that is
6920           supported).  This is the format used by the DBX debugger on IBM
6921           RS/6000 systems.
6922
6923       -gxcoff+
6924           Produce debugging information in XCOFF format (if that is
6925           supported), using GNU extensions understood only by the GNU
6926           debugger (GDB).  The use of these extensions is likely to make
6927           other debuggers crash or refuse to read the program, and may cause
6928           assemblers other than the GNU assembler (GAS) to fail with an
6929           error.
6930
6931       -gvms
6932           Produce debugging information in Alpha/VMS debug format (if that is
6933           supported).  This is the format used by DEBUG on Alpha/VMS systems.
6934
6935       -glevel
6936       -ggdblevel
6937       -gstabslevel
6938       -gxcofflevel
6939       -gvmslevel
6940           Request debugging information and also use level to specify how
6941           much information.  The default level is 2.
6942
6943           Level 0 produces no debug information at all.  Thus, -g0 negates
6944           -g.
6945
6946           Level 1 produces minimal information, enough for making backtraces
6947           in parts of the program that you don't plan to debug.  This
6948           includes descriptions of functions and external variables, and line
6949           number tables, but no information about local variables.
6950
6951           Level 3 includes extra information, such as all the macro
6952           definitions present in the program.  Some debuggers support macro
6953           expansion when you use -g3.
6954
6955           If you use multiple -g options, with or without level numbers, the
6956           last such option is the one that is effective.
6957
6958           -gdwarf does not accept a concatenated debug level, to avoid
6959           confusion with -gdwarf-level.  Instead use an additional -glevel
6960           option to change the debug level for DWARF.
6961
6962       -fno-eliminate-unused-debug-symbols
6963           By default, no debug information is produced for symbols that are
6964           not actually used. Use this option if you want debug information
6965           for all symbols.
6966
6967       -femit-class-debug-always
6968           Instead of emitting debugging information for a C++ class in only
6969           one object file, emit it in all object files using the class.  This
6970           option should be used only with debuggers that are unable to handle
6971           the way GCC normally emits debugging information for classes
6972           because using this option increases the size of debugging
6973           information by as much as a factor of two.
6974
6975       -fno-merge-debug-strings
6976           Direct the linker to not merge together strings in the debugging
6977           information that are identical in different object files.  Merging
6978           is not supported by all assemblers or linkers.  Merging decreases
6979           the size of the debug information in the output file at the cost of
6980           increasing link processing time.  Merging is enabled by default.
6981
6982       -fdebug-prefix-map=old=new
6983           When compiling files residing in directory old, record debugging
6984           information describing them as if the files resided in directory
6985           new instead.  This can be used to replace a build-time path with an
6986           install-time path in the debug info.  It can also be used to change
6987           an absolute path to a relative path by using . for new.  This can
6988           give more reproducible builds, which are location independent, but
6989           may require an extra command to tell GDB where to find the source
6990           files. See also -ffile-prefix-map.
6991
6992       -fvar-tracking
6993           Run variable tracking pass.  It computes where variables are stored
6994           at each position in code.  Better debugging information is then
6995           generated (if the debugging information format supports this
6996           information).
6997
6998           It is enabled by default when compiling with optimization (-Os, -O,
6999           -O2, ...), debugging information (-g) and the debug info format
7000           supports it.
7001
7002       -fvar-tracking-assignments
7003           Annotate assignments to user variables early in the compilation and
7004           attempt to carry the annotations over throughout the compilation
7005           all the way to the end, in an attempt to improve debug information
7006           while optimizing.  Use of -gdwarf-4 is recommended along with it.
7007
7008           It can be enabled even if var-tracking is disabled, in which case
7009           annotations are created and maintained, but discarded at the end.
7010           By default, this flag is enabled together with -fvar-tracking,
7011           except when selective scheduling is enabled.
7012
7013       -gsplit-dwarf
7014           Separate as much DWARF debugging information as possible into a
7015           separate output file with the extension .dwo.  This option allows
7016           the build system to avoid linking files with debug information.  To
7017           be useful, this option requires a debugger capable of reading .dwo
7018           files.
7019
7020       -gdescribe-dies
7021           Add description attributes to some DWARF DIEs that have no name
7022           attribute, such as artificial variables, external references and
7023           call site parameter DIEs.
7024
7025       -gpubnames
7026           Generate DWARF ".debug_pubnames" and ".debug_pubtypes" sections.
7027
7028       -ggnu-pubnames
7029           Generate ".debug_pubnames" and ".debug_pubtypes" sections in a
7030           format suitable for conversion into a GDB index.  This option is
7031           only useful with a linker that can produce GDB index version 7.
7032
7033       -fdebug-types-section
7034           When using DWARF Version 4 or higher, type DIEs can be put into
7035           their own ".debug_types" section instead of making them part of the
7036           ".debug_info" section.  It is more efficient to put them in a
7037           separate comdat section since the linker can then remove
7038           duplicates.  But not all DWARF consumers support ".debug_types"
7039           sections yet and on some objects ".debug_types" produces larger
7040           instead of smaller debugging information.
7041
7042       -grecord-gcc-switches
7043       -gno-record-gcc-switches
7044           This switch causes the command-line options used to invoke the
7045           compiler that may affect code generation to be appended to the
7046           DW_AT_producer attribute in DWARF debugging information.  The
7047           options are concatenated with spaces separating them from each
7048           other and from the compiler version.  It is enabled by default.
7049           See also -frecord-gcc-switches for another way of storing compiler
7050           options into the object file.
7051
7052       -gstrict-dwarf
7053           Disallow using extensions of later DWARF standard version than
7054           selected with -gdwarf-version.  On most targets using non-
7055           conflicting DWARF extensions from later standard versions is
7056           allowed.
7057
7058       -gno-strict-dwarf
7059           Allow using extensions of later DWARF standard version than
7060           selected with -gdwarf-version.
7061
7062       -gas-loc-support
7063           Inform the compiler that the assembler supports ".loc" directives.
7064           It may then use them for the assembler to generate DWARF2+ line
7065           number tables.
7066
7067           This is generally desirable, because assembler-generated line-
7068           number tables are a lot more compact than those the compiler can
7069           generate itself.
7070
7071           This option will be enabled by default if, at GCC configure time,
7072           the assembler was found to support such directives.
7073
7074       -gno-as-loc-support
7075           Force GCC to generate DWARF2+ line number tables internally, if
7076           DWARF2+ line number tables are to be generated.
7077
7078       -gas-locview-support
7079           Inform the compiler that the assembler supports "view" assignment
7080           and reset assertion checking in ".loc" directives.
7081
7082           This option will be enabled by default if, at GCC configure time,
7083           the assembler was found to support them.
7084
7085       -gno-as-locview-support
7086           Force GCC to assign view numbers internally, if
7087           -gvariable-location-views are explicitly requested.
7088
7089       -gcolumn-info
7090       -gno-column-info
7091           Emit location column information into DWARF debugging information,
7092           rather than just file and line.  This option is enabled by default.
7093
7094       -gstatement-frontiers
7095       -gno-statement-frontiers
7096           This option causes GCC to create markers in the internal
7097           representation at the beginning of statements, and to keep them
7098           roughly in place throughout compilation, using them to guide the
7099           output of "is_stmt" markers in the line number table.  This is
7100           enabled by default when compiling with optimization (-Os, -O, -O2,
7101           ...), and outputting DWARF 2 debug information at the normal level.
7102
7103       -gvariable-location-views
7104       -gvariable-location-views=incompat5
7105       -gno-variable-location-views
7106           Augment variable location lists with progressive view numbers
7107           implied from the line number table.  This enables debug information
7108           consumers to inspect state at certain points of the program, even
7109           if no instructions associated with the corresponding source
7110           locations are present at that point.  If the assembler lacks
7111           support for view numbers in line number tables, this will cause the
7112           compiler to emit the line number table, which generally makes them
7113           somewhat less compact.  The augmented line number tables and
7114           location lists are fully backward-compatible, so they can be
7115           consumed by debug information consumers that are not aware of these
7116           augmentations, but they won't derive any benefit from them either.
7117
7118           This is enabled by default when outputting DWARF 2 debug
7119           information at the normal level, as long as there is assembler
7120           support, -fvar-tracking-assignments is enabled and -gstrict-dwarf
7121           is not.  When assembler support is not available, this may still be
7122           enabled, but it will force GCC to output internal line number
7123           tables, and if -ginternal-reset-location-views is not enabled, that
7124           will most certainly lead to silently mismatching location views.
7125
7126           There is a proposed representation for view numbers that is not
7127           backward compatible with the location list format introduced in
7128           DWARF 5, that can be enabled with
7129           -gvariable-location-views=incompat5.  This option may be removed in
7130           the future, is only provided as a reference implementation of the
7131           proposed representation.  Debug information consumers are not
7132           expected to support this extended format, and they would be
7133           rendered unable to decode location lists using it.
7134
7135       -ginternal-reset-location-views
7136       -gno-internal-reset-location-views
7137           Attempt to determine location views that can be omitted from
7138           location view lists.  This requires the compiler to have very
7139           accurate insn length estimates, which isn't always the case, and it
7140           may cause incorrect view lists to be generated silently when using
7141           an assembler that does not support location view lists.  The GNU
7142           assembler will flag any such error as a "view number mismatch".
7143           This is only enabled on ports that define a reliable estimation
7144           function.
7145
7146       -ginline-points
7147       -gno-inline-points
7148           Generate extended debug information for inlined functions.
7149           Location view tracking markers are inserted at inlined entry
7150           points, so that address and view numbers can be computed and output
7151           in debug information.  This can be enabled independently of
7152           location views, in which case the view numbers won't be output, but
7153           it can only be enabled along with statement frontiers, and it is
7154           only enabled by default if location views are enabled.
7155
7156       -gz[=type]
7157           Produce compressed debug sections in DWARF format, if that is
7158           supported.  If type is not given, the default type depends on the
7159           capabilities of the assembler and linker used.  type may be one of
7160           none (don't compress debug sections), zlib (use zlib compression in
7161           ELF gABI format), or zlib-gnu (use zlib compression in traditional
7162           GNU format).  If the linker doesn't support writing compressed
7163           debug sections, the option is rejected.  Otherwise, if the
7164           assembler does not support them, -gz is silently ignored when
7165           producing object files.
7166
7167       -femit-struct-debug-baseonly
7168           Emit debug information for struct-like types only when the base
7169           name of the compilation source file matches the base name of file
7170           in which the struct is defined.
7171
7172           This option substantially reduces the size of debugging
7173           information, but at significant potential loss in type information
7174           to the debugger.  See -femit-struct-debug-reduced for a less
7175           aggressive option.  See -femit-struct-debug-detailed for more
7176           detailed control.
7177
7178           This option works only with DWARF debug output.
7179
7180       -femit-struct-debug-reduced
7181           Emit debug information for struct-like types only when the base
7182           name of the compilation source file matches the base name of file
7183           in which the type is defined, unless the struct is a template or
7184           defined in a system header.
7185
7186           This option significantly reduces the size of debugging
7187           information, with some potential loss in type information to the
7188           debugger.  See -femit-struct-debug-baseonly for a more aggressive
7189           option.  See -femit-struct-debug-detailed for more detailed
7190           control.
7191
7192           This option works only with DWARF debug output.
7193
7194       -femit-struct-debug-detailed[=spec-list]
7195           Specify the struct-like types for which the compiler generates
7196           debug information.  The intent is to reduce duplicate struct debug
7197           information between different object files within the same program.
7198
7199           This option is a detailed version of -femit-struct-debug-reduced
7200           and -femit-struct-debug-baseonly, which serves for most needs.
7201
7202           A specification has the
7203           syntax[dir:|ind:][ord:|gen:](any|sys|base|none)
7204
7205           The optional first word limits the specification to structs that
7206           are used directly (dir:) or used indirectly (ind:).  A struct type
7207           is used directly when it is the type of a variable, member.
7208           Indirect uses arise through pointers to structs.  That is, when use
7209           of an incomplete struct is valid, the use is indirect.  An example
7210           is struct one direct; struct two * indirect;.
7211
7212           The optional second word limits the specification to ordinary
7213           structs (ord:) or generic structs (gen:).  Generic structs are a
7214           bit complicated to explain.  For C++, these are non-explicit
7215           specializations of template classes, or non-template classes within
7216           the above.  Other programming languages have generics, but
7217           -femit-struct-debug-detailed does not yet implement them.
7218
7219           The third word specifies the source files for those structs for
7220           which the compiler should emit debug information.  The values none
7221           and any have the normal meaning.  The value base means that the
7222           base of name of the file in which the type declaration appears must
7223           match the base of the name of the main compilation file.  In
7224           practice, this means that when compiling foo.c, debug information
7225           is generated for types declared in that file and foo.h, but not
7226           other header files.  The value sys means those types satisfying
7227           base or declared in system or compiler headers.
7228
7229           You may need to experiment to determine the best settings for your
7230           application.
7231
7232           The default is -femit-struct-debug-detailed=all.
7233
7234           This option works only with DWARF debug output.
7235
7236       -fno-dwarf2-cfi-asm
7237           Emit DWARF unwind info as compiler generated ".eh_frame" section
7238           instead of using GAS ".cfi_*" directives.
7239
7240       -fno-eliminate-unused-debug-types
7241           Normally, when producing DWARF output, GCC avoids producing debug
7242           symbol output for types that are nowhere used in the source file
7243           being compiled.  Sometimes it is useful to have GCC emit debugging
7244           information for all types declared in a compilation unit,
7245           regardless of whether or not they are actually used in that
7246           compilation unit, for example if, in the debugger, you want to cast
7247           a value to a type that is not actually used in your program (but is
7248           declared).  More often, however, this results in a significant
7249           amount of wasted space.
7250
7251   Options That Control Optimization
7252       These options control various sorts of optimizations.
7253
7254       Without any optimization option, the compiler's goal is to reduce the
7255       cost of compilation and to make debugging produce the expected results.
7256       Statements are independent: if you stop the program with a breakpoint
7257       between statements, you can then assign a new value to any variable or
7258       change the program counter to any other statement in the function and
7259       get exactly the results you expect from the source code.
7260
7261       Turning on optimization flags makes the compiler attempt to improve the
7262       performance and/or code size at the expense of compilation time and
7263       possibly the ability to debug the program.
7264
7265       The compiler performs optimization based on the knowledge it has of the
7266       program.  Compiling multiple files at once to a single output file mode
7267       allows the compiler to use information gained from all of the files
7268       when compiling each of them.
7269
7270       Not all optimizations are controlled directly by a flag.  Only
7271       optimizations that have a flag are listed in this section.
7272
7273       Most optimizations are completely disabled at -O0 or if an -O level is
7274       not set on the command line, even if individual optimization flags are
7275       specified.  Similarly, -Og suppresses many optimization passes.
7276
7277       Depending on the target and how GCC was configured, a slightly
7278       different set of optimizations may be enabled at each -O level than
7279       those listed here.  You can invoke GCC with -Q --help=optimizers to
7280       find out the exact set of optimizations that are enabled at each level.
7281
7282       -O
7283       -O1 Optimize.  Optimizing compilation takes somewhat more time, and a
7284           lot more memory for a large function.
7285
7286           With -O, the compiler tries to reduce code size and execution time,
7287           without performing any optimizations that take a great deal of
7288           compilation time.
7289
7290           -O turns on the following optimization flags:
7291
7292           -fauto-inc-dec -fbranch-count-reg -fcombine-stack-adjustments
7293           -fcompare-elim -fcprop-registers -fdce -fdefer-pop -fdelayed-branch
7294           -fdse -fforward-propagate -fguess-branch-probability
7295           -fif-conversion -fif-conversion2 -finline-functions-called-once
7296           -fipa-profile -fipa-pure-const -fipa-reference
7297           -fipa-reference-addressable -fmerge-constants
7298           -fmove-loop-invariants -fomit-frame-pointer -freorder-blocks
7299           -fshrink-wrap -fshrink-wrap-separate -fsplit-wide-types
7300           -fssa-backprop -fssa-phiopt -ftree-bit-ccp -ftree-ccp -ftree-ch
7301           -ftree-coalesce-vars -ftree-copy-prop -ftree-dce
7302           -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre
7303           -ftree-phiprop -ftree-pta -ftree-scev-cprop -ftree-sink -ftree-slsr
7304           -ftree-sra -ftree-ter -funit-at-a-time
7305
7306       -O2 Optimize even more.  GCC performs nearly all supported
7307           optimizations that do not involve a space-speed tradeoff.  As
7308           compared to -O, this option increases both compilation time and the
7309           performance of the generated code.
7310
7311           -O2 turns on all optimization flags specified by -O.  It also turns
7312           on the following optimization flags:
7313
7314           -falign-functions  -falign-jumps -falign-labels  -falign-loops
7315           -fcaller-saves -fcode-hoisting -fcrossjumping -fcse-follow-jumps
7316           -fcse-skip-blocks -fdelete-null-pointer-checks -fdevirtualize
7317           -fdevirtualize-speculatively -fexpensive-optimizations
7318           -ffinite-loops -fgcse  -fgcse-lm -fhoist-adjacent-loads
7319           -finline-functions -finline-small-functions -findirect-inlining
7320           -fipa-bit-cp  -fipa-cp  -fipa-icf -fipa-ra  -fipa-sra  -fipa-vrp
7321           -fisolate-erroneous-paths-dereference -flra-remat
7322           -foptimize-sibling-calls -foptimize-strlen -fpartial-inlining
7323           -fpeephole2 -freorder-blocks-algorithm=stc
7324           -freorder-blocks-and-partition  -freorder-functions
7325           -frerun-cse-after-loop -fschedule-insns  -fschedule-insns2
7326           -fsched-interblock  -fsched-spec -fstore-merging -fstrict-aliasing
7327           -fthread-jumps -ftree-builtin-call-dce -ftree-pre
7328           -ftree-switch-conversion  -ftree-tail-merge -ftree-vrp
7329
7330           Please note the warning under -fgcse about invoking -O2 on programs
7331           that use computed gotos.
7332
7333       -O3 Optimize yet more.  -O3 turns on all optimizations specified by -O2
7334           and also turns on the following optimization flags:
7335
7336           -fgcse-after-reload -fipa-cp-clone -floop-interchange
7337           -floop-unroll-and-jam -fpeel-loops -fpredictive-commoning
7338           -fsplit-loops -fsplit-paths -ftree-loop-distribution
7339           -ftree-loop-vectorize -ftree-partial-pre -ftree-slp-vectorize
7340           -funswitch-loops -fvect-cost-model -fvect-cost-model=dynamic
7341           -fversion-loops-for-strides
7342
7343       -O0 Reduce compilation time and make debugging produce the expected
7344           results.  This is the default.
7345
7346       -Os Optimize for size.  -Os enables all -O2 optimizations except those
7347           that often increase code size:
7348
7349           -falign-functions  -falign-jumps -falign-labels  -falign-loops
7350           -fprefetch-loop-arrays  -freorder-blocks-algorithm=stc
7351
7352           It also enables -finline-functions, causes the compiler to tune for
7353           code size rather than execution speed, and performs further
7354           optimizations designed to reduce code size.
7355
7356       -Ofast
7357           Disregard strict standards compliance.  -Ofast enables all -O3
7358           optimizations.  It also enables optimizations that are not valid
7359           for all standard-compliant programs.  It turns on -ffast-math,
7360           -fallow-store-data-races and the Fortran-specific -fstack-arrays,
7361           unless -fmax-stack-var-size is specified, and -fno-protect-parens.
7362
7363       -Og Optimize debugging experience.  -Og should be the optimization
7364           level of choice for the standard edit-compile-debug cycle, offering
7365           a reasonable level of optimization while maintaining fast
7366           compilation and a good debugging experience.  It is a better choice
7367           than -O0 for producing debuggable code because some compiler passes
7368           that collect debug information are disabled at -O0.
7369
7370           Like -O0, -Og completely disables a number of optimization passes
7371           so that individual options controlling them have no effect.
7372           Otherwise -Og enables all -O1 optimization flags except for those
7373           that may interfere with debugging:
7374
7375           -fbranch-count-reg  -fdelayed-branch -fdse  -fif-conversion
7376           -fif-conversion2 -finline-functions-called-once
7377           -fmove-loop-invariants  -fssa-phiopt -ftree-bit-ccp  -ftree-dse
7378           -ftree-pta  -ftree-sra
7379
7380       If you use multiple -O options, with or without level numbers, the last
7381       such option is the one that is effective.
7382
7383       Options of the form -fflag specify machine-independent flags.  Most
7384       flags have both positive and negative forms; the negative form of -ffoo
7385       is -fno-foo.  In the table below, only one of the forms is listed---the
7386       one you typically use.  You can figure out the other form by either
7387       removing no- or adding it.
7388
7389       The following options control specific optimizations.  They are either
7390       activated by -O options or are related to ones that are.  You can use
7391       the following flags in the rare cases when "fine-tuning" of
7392       optimizations to be performed is desired.
7393
7394       -fno-defer-pop
7395           For machines that must pop arguments after a function call, always
7396           pop the arguments as soon as each function returns.  At levels -O1
7397           and higher, -fdefer-pop is the default; this allows the compiler to
7398           let arguments accumulate on the stack for several function calls
7399           and pop them all at once.
7400
7401       -fforward-propagate
7402           Perform a forward propagation pass on RTL.  The pass tries to
7403           combine two instructions and checks if the result can be
7404           simplified.  If loop unrolling is active, two passes are performed
7405           and the second is scheduled after loop unrolling.
7406
7407           This option is enabled by default at optimization levels -O, -O2,
7408           -O3, -Os.
7409
7410       -ffp-contract=style
7411           -ffp-contract=off disables floating-point expression contraction.
7412           -ffp-contract=fast enables floating-point expression contraction
7413           such as forming of fused multiply-add operations if the target has
7414           native support for them.  -ffp-contract=on enables floating-point
7415           expression contraction if allowed by the language standard.  This
7416           is currently not implemented and treated equal to
7417           -ffp-contract=off.
7418
7419           The default is -ffp-contract=fast.
7420
7421       -fomit-frame-pointer
7422           Omit the frame pointer in functions that don't need one.  This
7423           avoids the instructions to save, set up and restore the frame
7424           pointer; on many targets it also makes an extra register available.
7425
7426           On some targets this flag has no effect because the standard
7427           calling sequence always uses a frame pointer, so it cannot be
7428           omitted.
7429
7430           Note that -fno-omit-frame-pointer doesn't guarantee the frame
7431           pointer is used in all functions.  Several targets always omit the
7432           frame pointer in leaf functions.
7433
7434           Enabled by default at -O and higher.
7435
7436       -foptimize-sibling-calls
7437           Optimize sibling and tail recursive calls.
7438
7439           Enabled at levels -O2, -O3, -Os.
7440
7441       -foptimize-strlen
7442           Optimize various standard C string functions (e.g. "strlen",
7443           "strchr" or "strcpy") and their "_FORTIFY_SOURCE" counterparts into
7444           faster alternatives.
7445
7446           Enabled at levels -O2, -O3.
7447
7448       -fno-inline
7449           Do not expand any functions inline apart from those marked with the
7450           "always_inline" attribute.  This is the default when not
7451           optimizing.
7452
7453           Single functions can be exempted from inlining by marking them with
7454           the "noinline" attribute.
7455
7456       -finline-small-functions
7457           Integrate functions into their callers when their body is smaller
7458           than expected function call code (so overall size of program gets
7459           smaller).  The compiler heuristically decides which functions are
7460           simple enough to be worth integrating in this way.  This inlining
7461           applies to all functions, even those not declared inline.
7462
7463           Enabled at levels -O2, -O3, -Os.
7464
7465       -findirect-inlining
7466           Inline also indirect calls that are discovered to be known at
7467           compile time thanks to previous inlining.  This option has any
7468           effect only when inlining itself is turned on by the
7469           -finline-functions or -finline-small-functions options.
7470
7471           Enabled at levels -O2, -O3, -Os.
7472
7473       -finline-functions
7474           Consider all functions for inlining, even if they are not declared
7475           inline.  The compiler heuristically decides which functions are
7476           worth integrating in this way.
7477
7478           If all calls to a given function are integrated, and the function
7479           is declared "static", then the function is normally not output as
7480           assembler code in its own right.
7481
7482           Enabled at levels -O2, -O3, -Os.  Also enabled by -fprofile-use and
7483           -fauto-profile.
7484
7485       -finline-functions-called-once
7486           Consider all "static" functions called once for inlining into their
7487           caller even if they are not marked "inline".  If a call to a given
7488           function is integrated, then the function is not output as
7489           assembler code in its own right.
7490
7491           Enabled at levels -O1, -O2, -O3 and -Os, but not -Og.
7492
7493       -fearly-inlining
7494           Inline functions marked by "always_inline" and functions whose body
7495           seems smaller than the function call overhead early before doing
7496           -fprofile-generate instrumentation and real inlining pass.  Doing
7497           so makes profiling significantly cheaper and usually inlining
7498           faster on programs having large chains of nested wrapper functions.
7499
7500           Enabled by default.
7501
7502       -fipa-sra
7503           Perform interprocedural scalar replacement of aggregates, removal
7504           of unused parameters and replacement of parameters passed by
7505           reference by parameters passed by value.
7506
7507           Enabled at levels -O2, -O3 and -Os.
7508
7509       -finline-limit=n
7510           By default, GCC limits the size of functions that can be inlined.
7511           This flag allows coarse control of this limit.  n is the size of
7512           functions that can be inlined in number of pseudo instructions.
7513
7514           Inlining is actually controlled by a number of parameters, which
7515           may be specified individually by using --param name=value.  The
7516           -finline-limit=n option sets some of these parameters as follows:
7517
7518           max-inline-insns-single
7519               is set to n/2.
7520
7521           max-inline-insns-auto
7522               is set to n/2.
7523
7524           See below for a documentation of the individual parameters
7525           controlling inlining and for the defaults of these parameters.
7526
7527           Note: there may be no value to -finline-limit that results in
7528           default behavior.
7529
7530           Note: pseudo instruction represents, in this particular context, an
7531           abstract measurement of function's size.  In no way does it
7532           represent a count of assembly instructions and as such its exact
7533           meaning might change from one release to an another.
7534
7535       -fno-keep-inline-dllexport
7536           This is a more fine-grained version of -fkeep-inline-functions,
7537           which applies only to functions that are declared using the
7538           "dllexport" attribute or declspec.
7539
7540       -fkeep-inline-functions
7541           In C, emit "static" functions that are declared "inline" into the
7542           object file, even if the function has been inlined into all of its
7543           callers.  This switch does not affect functions using the "extern
7544           inline" extension in GNU C90.  In C++, emit any and all inline
7545           functions into the object file.
7546
7547       -fkeep-static-functions
7548           Emit "static" functions into the object file, even if the function
7549           is never used.
7550
7551       -fkeep-static-consts
7552           Emit variables declared "static const" when optimization isn't
7553           turned on, even if the variables aren't referenced.
7554
7555           GCC enables this option by default.  If you want to force the
7556           compiler to check if a variable is referenced, regardless of
7557           whether or not optimization is turned on, use the
7558           -fno-keep-static-consts option.
7559
7560       -fmerge-constants
7561           Attempt to merge identical constants (string constants and
7562           floating-point constants) across compilation units.
7563
7564           This option is the default for optimized compilation if the
7565           assembler and linker support it.  Use -fno-merge-constants to
7566           inhibit this behavior.
7567
7568           Enabled at levels -O, -O2, -O3, -Os.
7569
7570       -fmerge-all-constants
7571           Attempt to merge identical constants and identical variables.
7572
7573           This option implies -fmerge-constants.  In addition to
7574           -fmerge-constants this considers e.g. even constant initialized
7575           arrays or initialized constant variables with integral or floating-
7576           point types.  Languages like C or C++ require each variable,
7577           including multiple instances of the same variable in recursive
7578           calls, to have distinct locations, so using this option results in
7579           non-conforming behavior.
7580
7581       -fmodulo-sched
7582           Perform swing modulo scheduling immediately before the first
7583           scheduling pass.  This pass looks at innermost loops and reorders
7584           their instructions by overlapping different iterations.
7585
7586       -fmodulo-sched-allow-regmoves
7587           Perform more aggressive SMS-based modulo scheduling with register
7588           moves allowed.  By setting this flag certain anti-dependences edges
7589           are deleted, which triggers the generation of reg-moves based on
7590           the life-range analysis.  This option is effective only with
7591           -fmodulo-sched enabled.
7592
7593       -fno-branch-count-reg
7594           Disable the optimization pass that scans for opportunities to use
7595           "decrement and branch" instructions on a count register instead of
7596           instruction sequences that decrement a register, compare it against
7597           zero, and then branch based upon the result.  This option is only
7598           meaningful on architectures that support such instructions, which
7599           include x86, PowerPC, IA-64 and S/390.  Note that the
7600           -fno-branch-count-reg option doesn't remove the decrement and
7601           branch instructions from the generated instruction stream
7602           introduced by other optimization passes.
7603
7604           The default is -fbranch-count-reg at -O1 and higher, except for
7605           -Og.
7606
7607       -fno-function-cse
7608           Do not put function addresses in registers; make each instruction
7609           that calls a constant function contain the function's address
7610           explicitly.
7611
7612           This option results in less efficient code, but some strange hacks
7613           that alter the assembler output may be confused by the
7614           optimizations performed when this option is not used.
7615
7616           The default is -ffunction-cse
7617
7618       -fno-zero-initialized-in-bss
7619           If the target supports a BSS section, GCC by default puts variables
7620           that are initialized to zero into BSS.  This can save space in the
7621           resulting code.
7622
7623           This option turns off this behavior because some programs
7624           explicitly rely on variables going to the data section---e.g., so
7625           that the resulting executable can find the beginning of that
7626           section and/or make assumptions based on that.
7627
7628           The default is -fzero-initialized-in-bss.
7629
7630       -fthread-jumps
7631           Perform optimizations that check to see if a jump branches to a
7632           location where another comparison subsumed by the first is found.
7633           If so, the first branch is redirected to either the destination of
7634           the second branch or a point immediately following it, depending on
7635           whether the condition is known to be true or false.
7636
7637           Enabled at levels -O2, -O3, -Os.
7638
7639       -fsplit-wide-types
7640           When using a type that occupies multiple registers, such as "long
7641           long" on a 32-bit system, split the registers apart and allocate
7642           them independently.  This normally generates better code for those
7643           types, but may make debugging more difficult.
7644
7645           Enabled at levels -O, -O2, -O3, -Os.
7646
7647       -fsplit-wide-types-early
7648           Fully split wide types early, instead of very late.  This option
7649           has no effect unless -fsplit-wide-types is turned on.
7650
7651           This is the default on some targets.
7652
7653       -fcse-follow-jumps
7654           In common subexpression elimination (CSE), scan through jump
7655           instructions when the target of the jump is not reached by any
7656           other path.  For example, when CSE encounters an "if" statement
7657           with an "else" clause, CSE follows the jump when the condition
7658           tested is false.
7659
7660           Enabled at levels -O2, -O3, -Os.
7661
7662       -fcse-skip-blocks
7663           This is similar to -fcse-follow-jumps, but causes CSE to follow
7664           jumps that conditionally skip over blocks.  When CSE encounters a
7665           simple "if" statement with no else clause, -fcse-skip-blocks causes
7666           CSE to follow the jump around the body of the "if".
7667
7668           Enabled at levels -O2, -O3, -Os.
7669
7670       -frerun-cse-after-loop
7671           Re-run common subexpression elimination after loop optimizations
7672           are performed.
7673
7674           Enabled at levels -O2, -O3, -Os.
7675
7676       -fgcse
7677           Perform a global common subexpression elimination pass.  This pass
7678           also performs global constant and copy propagation.
7679
7680           Note: When compiling a program using computed gotos, a GCC
7681           extension, you may get better run-time performance if you disable
7682           the global common subexpression elimination pass by adding
7683           -fno-gcse to the command line.
7684
7685           Enabled at levels -O2, -O3, -Os.
7686
7687       -fgcse-lm
7688           When -fgcse-lm is enabled, global common subexpression elimination
7689           attempts to move loads that are only killed by stores into
7690           themselves.  This allows a loop containing a load/store sequence to
7691           be changed to a load outside the loop, and a copy/store within the
7692           loop.
7693
7694           Enabled by default when -fgcse is enabled.
7695
7696       -fgcse-sm
7697           When -fgcse-sm is enabled, a store motion pass is run after global
7698           common subexpression elimination.  This pass attempts to move
7699           stores out of loops.  When used in conjunction with -fgcse-lm,
7700           loops containing a load/store sequence can be changed to a load
7701           before the loop and a store after the loop.
7702
7703           Not enabled at any optimization level.
7704
7705       -fgcse-las
7706           When -fgcse-las is enabled, the global common subexpression
7707           elimination pass eliminates redundant loads that come after stores
7708           to the same memory location (both partial and full redundancies).
7709
7710           Not enabled at any optimization level.
7711
7712       -fgcse-after-reload
7713           When -fgcse-after-reload is enabled, a redundant load elimination
7714           pass is performed after reload.  The purpose of this pass is to
7715           clean up redundant spilling.
7716
7717           Enabled by -fprofile-use and -fauto-profile.
7718
7719       -faggressive-loop-optimizations
7720           This option tells the loop optimizer to use language constraints to
7721           derive bounds for the number of iterations of a loop.  This assumes
7722           that loop code does not invoke undefined behavior by for example
7723           causing signed integer overflows or out-of-bound array accesses.
7724           The bounds for the number of iterations of a loop are used to guide
7725           loop unrolling and peeling and loop exit test optimizations.  This
7726           option is enabled by default.
7727
7728       -funconstrained-commons
7729           This option tells the compiler that variables declared in common
7730           blocks (e.g. Fortran) may later be overridden with longer trailing
7731           arrays. This prevents certain optimizations that depend on knowing
7732           the array bounds.
7733
7734       -fcrossjumping
7735           Perform cross-jumping transformation.  This transformation unifies
7736           equivalent code and saves code size.  The resulting code may or may
7737           not perform better than without cross-jumping.
7738
7739           Enabled at levels -O2, -O3, -Os.
7740
7741       -fauto-inc-dec
7742           Combine increments or decrements of addresses with memory accesses.
7743           This pass is always skipped on architectures that do not have
7744           instructions to support this.  Enabled by default at -O and higher
7745           on architectures that support this.
7746
7747       -fdce
7748           Perform dead code elimination (DCE) on RTL.  Enabled by default at
7749           -O and higher.
7750
7751       -fdse
7752           Perform dead store elimination (DSE) on RTL.  Enabled by default at
7753           -O and higher.
7754
7755       -fif-conversion
7756           Attempt to transform conditional jumps into branch-less
7757           equivalents.  This includes use of conditional moves, min, max, set
7758           flags and abs instructions, and some tricks doable by standard
7759           arithmetics.  The use of conditional execution on chips where it is
7760           available is controlled by -fif-conversion2.
7761
7762           Enabled at levels -O, -O2, -O3, -Os, but not with -Og.
7763
7764       -fif-conversion2
7765           Use conditional execution (where available) to transform
7766           conditional jumps into branch-less equivalents.
7767
7768           Enabled at levels -O, -O2, -O3, -Os, but not with -Og.
7769
7770       -fdeclone-ctor-dtor
7771           The C++ ABI requires multiple entry points for constructors and
7772           destructors: one for a base subobject, one for a complete object,
7773           and one for a virtual destructor that calls operator delete
7774           afterwards.  For a hierarchy with virtual bases, the base and
7775           complete variants are clones, which means two copies of the
7776           function.  With this option, the base and complete variants are
7777           changed to be thunks that call a common implementation.
7778
7779           Enabled by -Os.
7780
7781       -fdelete-null-pointer-checks
7782           Assume that programs cannot safely dereference null pointers, and
7783           that no code or data element resides at address zero.  This option
7784           enables simple constant folding optimizations at all optimization
7785           levels.  In addition, other optimization passes in GCC use this
7786           flag to control global dataflow analyses that eliminate useless
7787           checks for null pointers; these assume that a memory access to
7788           address zero always results in a trap, so that if a pointer is
7789           checked after it has already been dereferenced, it cannot be null.
7790
7791           Note however that in some environments this assumption is not true.
7792           Use -fno-delete-null-pointer-checks to disable this optimization
7793           for programs that depend on that behavior.
7794
7795           This option is enabled by default on most targets.  On Nios II ELF,
7796           it defaults to off.  On AVR, CR16, and MSP430, this option is
7797           completely disabled.
7798
7799           Passes that use the dataflow information are enabled independently
7800           at different optimization levels.
7801
7802       -fdevirtualize
7803           Attempt to convert calls to virtual functions to direct calls.
7804           This is done both within a procedure and interprocedurally as part
7805           of indirect inlining (-findirect-inlining) and interprocedural
7806           constant propagation (-fipa-cp).  Enabled at levels -O2, -O3, -Os.
7807
7808       -fdevirtualize-speculatively
7809           Attempt to convert calls to virtual functions to speculative direct
7810           calls.  Based on the analysis of the type inheritance graph,
7811           determine for a given call the set of likely targets. If the set is
7812           small, preferably of size 1, change the call into a conditional
7813           deciding between direct and indirect calls.  The speculative calls
7814           enable more optimizations, such as inlining.  When they seem
7815           useless after further optimization, they are converted back into
7816           original form.
7817
7818       -fdevirtualize-at-ltrans
7819           Stream extra information needed for aggressive devirtualization
7820           when running the link-time optimizer in local transformation mode.
7821           This option enables more devirtualization but significantly
7822           increases the size of streamed data. For this reason it is disabled
7823           by default.
7824
7825       -fexpensive-optimizations
7826           Perform a number of minor optimizations that are relatively
7827           expensive.
7828
7829           Enabled at levels -O2, -O3, -Os.
7830
7831       -free
7832           Attempt to remove redundant extension instructions.  This is
7833           especially helpful for the x86-64 architecture, which implicitly
7834           zero-extends in 64-bit registers after writing to their lower
7835           32-bit half.
7836
7837           Enabled for Alpha, AArch64 and x86 at levels -O2, -O3, -Os.
7838
7839       -fno-lifetime-dse
7840           In C++ the value of an object is only affected by changes within
7841           its lifetime: when the constructor begins, the object has an
7842           indeterminate value, and any changes during the lifetime of the
7843           object are dead when the object is destroyed.  Normally dead store
7844           elimination will take advantage of this; if your code relies on the
7845           value of the object storage persisting beyond the lifetime of the
7846           object, you can use this flag to disable this optimization.  To
7847           preserve stores before the constructor starts (e.g. because your
7848           operator new clears the object storage) but still treat the object
7849           as dead after the destructor, you can use -flifetime-dse=1.  The
7850           default behavior can be explicitly selected with -flifetime-dse=2.
7851           -flifetime-dse=0 is equivalent to -fno-lifetime-dse.
7852
7853       -flive-range-shrinkage
7854           Attempt to decrease register pressure through register live range
7855           shrinkage.  This is helpful for fast processors with small or
7856           moderate size register sets.
7857
7858       -fira-algorithm=algorithm
7859           Use the specified coloring algorithm for the integrated register
7860           allocator.  The algorithm argument can be priority, which specifies
7861           Chow's priority coloring, or CB, which specifies Chaitin-Briggs
7862           coloring.  Chaitin-Briggs coloring is not implemented for all
7863           architectures, but for those targets that do support it, it is the
7864           default because it generates better code.
7865
7866       -fira-region=region
7867           Use specified regions for the integrated register allocator.  The
7868           region argument should be one of the following:
7869
7870           all Use all loops as register allocation regions.  This can give
7871               the best results for machines with a small and/or irregular
7872               register set.
7873
7874           mixed
7875               Use all loops except for loops with small register pressure as
7876               the regions.  This value usually gives the best results in most
7877               cases and for most architectures, and is enabled by default
7878               when compiling with optimization for speed (-O, -O2, ...).
7879
7880           one Use all functions as a single region.  This typically results
7881               in the smallest code size, and is enabled by default for -Os or
7882               -O0.
7883
7884       -fira-hoist-pressure
7885           Use IRA to evaluate register pressure in the code hoisting pass for
7886           decisions to hoist expressions.  This option usually results in
7887           smaller code, but it can slow the compiler down.
7888
7889           This option is enabled at level -Os for all targets.
7890
7891       -fira-loop-pressure
7892           Use IRA to evaluate register pressure in loops for decisions to
7893           move loop invariants.  This option usually results in generation of
7894           faster and smaller code on machines with large register files (>=
7895           32 registers), but it can slow the compiler down.
7896
7897           This option is enabled at level -O3 for some targets.
7898
7899       -fno-ira-share-save-slots
7900           Disable sharing of stack slots used for saving call-used hard
7901           registers living through a call.  Each hard register gets a
7902           separate stack slot, and as a result function stack frames are
7903           larger.
7904
7905       -fno-ira-share-spill-slots
7906           Disable sharing of stack slots allocated for pseudo-registers.
7907           Each pseudo-register that does not get a hard register gets a
7908           separate stack slot, and as a result function stack frames are
7909           larger.
7910
7911       -flra-remat
7912           Enable CFG-sensitive rematerialization in LRA.  Instead of loading
7913           values of spilled pseudos, LRA tries to rematerialize (recalculate)
7914           values if it is profitable.
7915
7916           Enabled at levels -O2, -O3, -Os.
7917
7918       -fdelayed-branch
7919           If supported for the target machine, attempt to reorder
7920           instructions to exploit instruction slots available after delayed
7921           branch instructions.
7922
7923           Enabled at levels -O, -O2, -O3, -Os, but not at -Og.
7924
7925       -fschedule-insns
7926           If supported for the target machine, attempt to reorder
7927           instructions to eliminate execution stalls due to required data
7928           being unavailable.  This helps machines that have slow floating
7929           point or memory load instructions by allowing other instructions to
7930           be issued until the result of the load or floating-point
7931           instruction is required.
7932
7933           Enabled at levels -O2, -O3.
7934
7935       -fschedule-insns2
7936           Similar to -fschedule-insns, but requests an additional pass of
7937           instruction scheduling after register allocation has been done.
7938           This is especially useful on machines with a relatively small
7939           number of registers and where memory load instructions take more
7940           than one cycle.
7941
7942           Enabled at levels -O2, -O3, -Os.
7943
7944       -fno-sched-interblock
7945           Disable instruction scheduling across basic blocks, which is
7946           normally enabled when scheduling before register allocation, i.e.
7947           with -fschedule-insns or at -O2 or higher.
7948
7949       -fno-sched-spec
7950           Disable speculative motion of non-load instructions, which is
7951           normally enabled when scheduling before register allocation, i.e.
7952           with -fschedule-insns or at -O2 or higher.
7953
7954       -fsched-pressure
7955           Enable register pressure sensitive insn scheduling before register
7956           allocation.  This only makes sense when scheduling before register
7957           allocation is enabled, i.e. with -fschedule-insns or at -O2 or
7958           higher.  Usage of this option can improve the generated code and
7959           decrease its size by preventing register pressure increase above
7960           the number of available hard registers and subsequent spills in
7961           register allocation.
7962
7963       -fsched-spec-load
7964           Allow speculative motion of some load instructions.  This only
7965           makes sense when scheduling before register allocation, i.e. with
7966           -fschedule-insns or at -O2 or higher.
7967
7968       -fsched-spec-load-dangerous
7969           Allow speculative motion of more load instructions.  This only
7970           makes sense when scheduling before register allocation, i.e. with
7971           -fschedule-insns or at -O2 or higher.
7972
7973       -fsched-stalled-insns
7974       -fsched-stalled-insns=n
7975           Define how many insns (if any) can be moved prematurely from the
7976           queue of stalled insns into the ready list during the second
7977           scheduling pass.  -fno-sched-stalled-insns means that no insns are
7978           moved prematurely, -fsched-stalled-insns=0 means there is no limit
7979           on how many queued insns can be moved prematurely.
7980           -fsched-stalled-insns without a value is equivalent to
7981           -fsched-stalled-insns=1.
7982
7983       -fsched-stalled-insns-dep
7984       -fsched-stalled-insns-dep=n
7985           Define how many insn groups (cycles) are examined for a dependency
7986           on a stalled insn that is a candidate for premature removal from
7987           the queue of stalled insns.  This has an effect only during the
7988           second scheduling pass, and only if -fsched-stalled-insns is used.
7989           -fno-sched-stalled-insns-dep is equivalent to
7990           -fsched-stalled-insns-dep=0.  -fsched-stalled-insns-dep without a
7991           value is equivalent to -fsched-stalled-insns-dep=1.
7992
7993       -fsched2-use-superblocks
7994           When scheduling after register allocation, use superblock
7995           scheduling.  This allows motion across basic block boundaries,
7996           resulting in faster schedules.  This option is experimental, as not
7997           all machine descriptions used by GCC model the CPU closely enough
7998           to avoid unreliable results from the algorithm.
7999
8000           This only makes sense when scheduling after register allocation,
8001           i.e. with -fschedule-insns2 or at -O2 or higher.
8002
8003       -fsched-group-heuristic
8004           Enable the group heuristic in the scheduler.  This heuristic favors
8005           the instruction that belongs to a schedule group.  This is enabled
8006           by default when scheduling is enabled, i.e. with -fschedule-insns
8007           or -fschedule-insns2 or at -O2 or higher.
8008
8009       -fsched-critical-path-heuristic
8010           Enable the critical-path heuristic in the scheduler.  This
8011           heuristic favors instructions on the critical path.  This is
8012           enabled by default when scheduling is enabled, i.e. with
8013           -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
8014
8015       -fsched-spec-insn-heuristic
8016           Enable the speculative instruction heuristic in the scheduler.
8017           This heuristic favors speculative instructions with greater
8018           dependency weakness.  This is enabled by default when scheduling is
8019           enabled, i.e.  with -fschedule-insns or -fschedule-insns2 or at -O2
8020           or higher.
8021
8022       -fsched-rank-heuristic
8023           Enable the rank heuristic in the scheduler.  This heuristic favors
8024           the instruction belonging to a basic block with greater size or
8025           frequency.  This is enabled by default when scheduling is enabled,
8026           i.e.  with -fschedule-insns or -fschedule-insns2 or at -O2 or
8027           higher.
8028
8029       -fsched-last-insn-heuristic
8030           Enable the last-instruction heuristic in the scheduler.  This
8031           heuristic favors the instruction that is less dependent on the last
8032           instruction scheduled.  This is enabled by default when scheduling
8033           is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at
8034           -O2 or higher.
8035
8036       -fsched-dep-count-heuristic
8037           Enable the dependent-count heuristic in the scheduler.  This
8038           heuristic favors the instruction that has more instructions
8039           depending on it.  This is enabled by default when scheduling is
8040           enabled, i.e.  with -fschedule-insns or -fschedule-insns2 or at -O2
8041           or higher.
8042
8043       -freschedule-modulo-scheduled-loops
8044           Modulo scheduling is performed before traditional scheduling.  If a
8045           loop is modulo scheduled, later scheduling passes may change its
8046           schedule.  Use this option to control that behavior.
8047
8048       -fselective-scheduling
8049           Schedule instructions using selective scheduling algorithm.
8050           Selective scheduling runs instead of the first scheduler pass.
8051
8052       -fselective-scheduling2
8053           Schedule instructions using selective scheduling algorithm.
8054           Selective scheduling runs instead of the second scheduler pass.
8055
8056       -fsel-sched-pipelining
8057           Enable software pipelining of innermost loops during selective
8058           scheduling.  This option has no effect unless one of
8059           -fselective-scheduling or -fselective-scheduling2 is turned on.
8060
8061       -fsel-sched-pipelining-outer-loops
8062           When pipelining loops during selective scheduling, also pipeline
8063           outer loops.  This option has no effect unless
8064           -fsel-sched-pipelining is turned on.
8065
8066       -fsemantic-interposition
8067           Some object formats, like ELF, allow interposing of symbols by the
8068           dynamic linker.  This means that for symbols exported from the DSO,
8069           the compiler cannot perform interprocedural propagation, inlining
8070           and other optimizations in anticipation that the function or
8071           variable in question may change. While this feature is useful, for
8072           example, to rewrite memory allocation functions by a debugging
8073           implementation, it is expensive in the terms of code quality.  With
8074           -fno-semantic-interposition the compiler assumes that if
8075           interposition happens for functions the overwriting function will
8076           have precisely the same semantics (and side effects).  Similarly if
8077           interposition happens for variables, the constructor of the
8078           variable will be the same. The flag has no effect for functions
8079           explicitly declared inline (where it is never allowed for
8080           interposition to change semantics) and for symbols explicitly
8081           declared weak.
8082
8083       -fshrink-wrap
8084           Emit function prologues only before parts of the function that need
8085           it, rather than at the top of the function.  This flag is enabled
8086           by default at -O and higher.
8087
8088       -fshrink-wrap-separate
8089           Shrink-wrap separate parts of the prologue and epilogue separately,
8090           so that those parts are only executed when needed.  This option is
8091           on by default, but has no effect unless -fshrink-wrap is also
8092           turned on and the target supports this.
8093
8094       -fcaller-saves
8095           Enable allocation of values to registers that are clobbered by
8096           function calls, by emitting extra instructions to save and restore
8097           the registers around such calls.  Such allocation is done only when
8098           it seems to result in better code.
8099
8100           This option is always enabled by default on certain machines,
8101           usually those which have no call-preserved registers to use
8102           instead.
8103
8104           Enabled at levels -O2, -O3, -Os.
8105
8106       -fcombine-stack-adjustments
8107           Tracks stack adjustments (pushes and pops) and stack memory
8108           references and then tries to find ways to combine them.
8109
8110           Enabled by default at -O1 and higher.
8111
8112       -fipa-ra
8113           Use caller save registers for allocation if those registers are not
8114           used by any called function.  In that case it is not necessary to
8115           save and restore them around calls.  This is only possible if
8116           called functions are part of same compilation unit as current
8117           function and they are compiled before it.
8118
8119           Enabled at levels -O2, -O3, -Os, however the option is disabled if
8120           generated code will be instrumented for profiling (-p, or -pg) or
8121           if callee's register usage cannot be known exactly (this happens on
8122           targets that do not expose prologues and epilogues in RTL).
8123
8124       -fconserve-stack
8125           Attempt to minimize stack usage.  The compiler attempts to use less
8126           stack space, even if that makes the program slower.  This option
8127           implies setting the large-stack-frame parameter to 100 and the
8128           large-stack-frame-growth parameter to 400.
8129
8130       -ftree-reassoc
8131           Perform reassociation on trees.  This flag is enabled by default at
8132           -O and higher.
8133
8134       -fcode-hoisting
8135           Perform code hoisting.  Code hoisting tries to move the evaluation
8136           of expressions executed on all paths to the function exit as early
8137           as possible.  This is especially useful as a code size
8138           optimization, but it often helps for code speed as well.  This flag
8139           is enabled by default at -O2 and higher.
8140
8141       -ftree-pre
8142           Perform partial redundancy elimination (PRE) on trees.  This flag
8143           is enabled by default at -O2 and -O3.
8144
8145       -ftree-partial-pre
8146           Make partial redundancy elimination (PRE) more aggressive.  This
8147           flag is enabled by default at -O3.
8148
8149       -ftree-forwprop
8150           Perform forward propagation on trees.  This flag is enabled by
8151           default at -O and higher.
8152
8153       -ftree-fre
8154           Perform full redundancy elimination (FRE) on trees.  The difference
8155           between FRE and PRE is that FRE only considers expressions that are
8156           computed on all paths leading to the redundant computation.  This
8157           analysis is faster than PRE, though it exposes fewer redundancies.
8158           This flag is enabled by default at -O and higher.
8159
8160       -ftree-phiprop
8161           Perform hoisting of loads from conditional pointers on trees.  This
8162           pass is enabled by default at -O and higher.
8163
8164       -fhoist-adjacent-loads
8165           Speculatively hoist loads from both branches of an if-then-else if
8166           the loads are from adjacent locations in the same structure and the
8167           target architecture has a conditional move instruction.  This flag
8168           is enabled by default at -O2 and higher.
8169
8170       -ftree-copy-prop
8171           Perform copy propagation on trees.  This pass eliminates
8172           unnecessary copy operations.  This flag is enabled by default at -O
8173           and higher.
8174
8175       -fipa-pure-const
8176           Discover which functions are pure or constant.  Enabled by default
8177           at -O and higher.
8178
8179       -fipa-reference
8180           Discover which static variables do not escape the compilation unit.
8181           Enabled by default at -O and higher.
8182
8183       -fipa-reference-addressable
8184           Discover read-only, write-only and non-addressable static
8185           variables.  Enabled by default at -O and higher.
8186
8187       -fipa-stack-alignment
8188           Reduce stack alignment on call sites if possible.  Enabled by
8189           default.
8190
8191       -fipa-pta
8192           Perform interprocedural pointer analysis and interprocedural
8193           modification and reference analysis.  This option can cause
8194           excessive memory and compile-time usage on large compilation units.
8195           It is not enabled by default at any optimization level.
8196
8197       -fipa-profile
8198           Perform interprocedural profile propagation.  The functions called
8199           only from cold functions are marked as cold. Also functions
8200           executed once (such as "cold", "noreturn", static constructors or
8201           destructors) are identified. Cold functions and loop less parts of
8202           functions executed once are then optimized for size.  Enabled by
8203           default at -O and higher.
8204
8205       -fipa-cp
8206           Perform interprocedural constant propagation.  This optimization
8207           analyzes the program to determine when values passed to functions
8208           are constants and then optimizes accordingly.  This optimization
8209           can substantially increase performance if the application has
8210           constants passed to functions.  This flag is enabled by default at
8211           -O2, -Os and -O3.  It is also enabled by -fprofile-use and
8212           -fauto-profile.
8213
8214       -fipa-cp-clone
8215           Perform function cloning to make interprocedural constant
8216           propagation stronger.  When enabled, interprocedural constant
8217           propagation performs function cloning when externally visible
8218           function can be called with constant arguments.  Because this
8219           optimization can create multiple copies of functions, it may
8220           significantly increase code size (see --param
8221           ipa-cp-unit-growth=value).  This flag is enabled by default at -O3.
8222           It is also enabled by -fprofile-use and -fauto-profile.
8223
8224       -fipa-bit-cp
8225           When enabled, perform interprocedural bitwise constant propagation.
8226           This flag is enabled by default at -O2 and by -fprofile-use and
8227           -fauto-profile.  It requires that -fipa-cp is enabled.
8228
8229       -fipa-vrp
8230           When enabled, perform interprocedural propagation of value ranges.
8231           This flag is enabled by default at -O2. It requires that -fipa-cp
8232           is enabled.
8233
8234       -fipa-icf
8235           Perform Identical Code Folding for functions and read-only
8236           variables.  The optimization reduces code size and may disturb
8237           unwind stacks by replacing a function by equivalent one with a
8238           different name. The optimization works more effectively with link-
8239           time optimization enabled.
8240
8241           Although the behavior is similar to the Gold Linker's ICF
8242           optimization, GCC ICF works on different levels and thus the
8243           optimizations are not same - there are equivalences that are found
8244           only by GCC and equivalences found only by Gold.
8245
8246           This flag is enabled by default at -O2 and -Os.
8247
8248       -flive-patching=level
8249           Control GCC's optimizations to produce output suitable for live-
8250           patching.
8251
8252           If the compiler's optimization uses a function's body or
8253           information extracted from its body to optimize/change another
8254           function, the latter is called an impacted function of the former.
8255           If a function is patched, its impacted functions should be patched
8256           too.
8257
8258           The impacted functions are determined by the compiler's
8259           interprocedural optimizations.  For example, a caller is impacted
8260           when inlining a function into its caller, cloning a function and
8261           changing its caller to call this new clone, or extracting a
8262           function's pureness/constness information to optimize its direct or
8263           indirect callers, etc.
8264
8265           Usually, the more IPA optimizations enabled, the larger the number
8266           of impacted functions for each function.  In order to control the
8267           number of impacted functions and more easily compute the list of
8268           impacted function, IPA optimizations can be partially enabled at
8269           two different levels.
8270
8271           The level argument should be one of the following:
8272
8273           inline-clone
8274               Only enable inlining and cloning optimizations, which includes
8275               inlining, cloning, interprocedural scalar replacement of
8276               aggregates and partial inlining.  As a result, when patching a
8277               function, all its callers and its clones' callers are impacted,
8278               therefore need to be patched as well.
8279
8280               -flive-patching=inline-clone disables the following
8281               optimization flags: -fwhole-program  -fipa-pta  -fipa-reference
8282               -fipa-ra -fipa-icf  -fipa-icf-functions  -fipa-icf-variables
8283               -fipa-bit-cp  -fipa-vrp  -fipa-pure-const
8284               -fipa-reference-addressable -fipa-stack-alignment
8285
8286           inline-only-static
8287               Only enable inlining of static functions.  As a result, when
8288               patching a static function, all its callers are impacted and so
8289               need to be patched as well.
8290
8291               In addition to all the flags that -flive-patching=inline-clone
8292               disables, -flive-patching=inline-only-static disables the
8293               following additional optimization flags: -fipa-cp-clone
8294               -fipa-sra  -fpartial-inlining  -fipa-cp
8295
8296           When -flive-patching is specified without any value, the default
8297           value is inline-clone.
8298
8299           This flag is disabled by default.
8300
8301           Note that -flive-patching is not supported with link-time
8302           optimization (-flto).
8303
8304       -fisolate-erroneous-paths-dereference
8305           Detect paths that trigger erroneous or undefined behavior due to
8306           dereferencing a null pointer.  Isolate those paths from the main
8307           control flow and turn the statement with erroneous or undefined
8308           behavior into a trap.  This flag is enabled by default at -O2 and
8309           higher and depends on -fdelete-null-pointer-checks also being
8310           enabled.
8311
8312       -fisolate-erroneous-paths-attribute
8313           Detect paths that trigger erroneous or undefined behavior due to a
8314           null value being used in a way forbidden by a "returns_nonnull" or
8315           "nonnull" attribute.  Isolate those paths from the main control
8316           flow and turn the statement with erroneous or undefined behavior
8317           into a trap.  This is not currently enabled, but may be enabled by
8318           -O2 in the future.
8319
8320       -ftree-sink
8321           Perform forward store motion on trees.  This flag is enabled by
8322           default at -O and higher.
8323
8324       -ftree-bit-ccp
8325           Perform sparse conditional bit constant propagation on trees and
8326           propagate pointer alignment information.  This pass only operates
8327           on local scalar variables and is enabled by default at -O1 and
8328           higher, except for -Og.  It requires that -ftree-ccp is enabled.
8329
8330       -ftree-ccp
8331           Perform sparse conditional constant propagation (CCP) on trees.
8332           This pass only operates on local scalar variables and is enabled by
8333           default at -O and higher.
8334
8335       -fssa-backprop
8336           Propagate information about uses of a value up the definition chain
8337           in order to simplify the definitions.  For example, this pass
8338           strips sign operations if the sign of a value never matters.  The
8339           flag is enabled by default at -O and higher.
8340
8341       -fssa-phiopt
8342           Perform pattern matching on SSA PHI nodes to optimize conditional
8343           code.  This pass is enabled by default at -O1 and higher, except
8344           for -Og.
8345
8346       -ftree-switch-conversion
8347           Perform conversion of simple initializations in a switch to
8348           initializations from a scalar array.  This flag is enabled by
8349           default at -O2 and higher.
8350
8351       -ftree-tail-merge
8352           Look for identical code sequences.  When found, replace one with a
8353           jump to the other.  This optimization is known as tail merging or
8354           cross jumping.  This flag is enabled by default at -O2 and higher.
8355           The compilation time in this pass can be limited using max-tail-
8356           merge-comparisons parameter and max-tail-merge-iterations
8357           parameter.
8358
8359       -ftree-dce
8360           Perform dead code elimination (DCE) on trees.  This flag is enabled
8361           by default at -O and higher.
8362
8363       -ftree-builtin-call-dce
8364           Perform conditional dead code elimination (DCE) for calls to built-
8365           in functions that may set "errno" but are otherwise free of side
8366           effects.  This flag is enabled by default at -O2 and higher if -Os
8367           is not also specified.
8368
8369       -ffinite-loops
8370           Assume that a loop with an exit will eventually take the exit and
8371           not loop indefinitely.  This allows the compiler to remove loops
8372           that otherwise have no side-effects, not considering eventual
8373           endless looping as such.
8374
8375           This option is enabled by default at -O2 for C++ with -std=c++11 or
8376           higher.
8377
8378       -ftree-dominator-opts
8379           Perform a variety of simple scalar cleanups (constant/copy
8380           propagation, redundancy elimination, range propagation and
8381           expression simplification) based on a dominator tree traversal.
8382           This also performs jump threading (to reduce jumps to jumps). This
8383           flag is enabled by default at -O and higher.
8384
8385       -ftree-dse
8386           Perform dead store elimination (DSE) on trees.  A dead store is a
8387           store into a memory location that is later overwritten by another
8388           store without any intervening loads.  In this case the earlier
8389           store can be deleted.  This flag is enabled by default at -O and
8390           higher.
8391
8392       -ftree-ch
8393           Perform loop header copying on trees.  This is beneficial since it
8394           increases effectiveness of code motion optimizations.  It also
8395           saves one jump.  This flag is enabled by default at -O and higher.
8396           It is not enabled for -Os, since it usually increases code size.
8397
8398       -ftree-loop-optimize
8399           Perform loop optimizations on trees.  This flag is enabled by
8400           default at -O and higher.
8401
8402       -ftree-loop-linear
8403       -floop-strip-mine
8404       -floop-block
8405           Perform loop nest optimizations.  Same as -floop-nest-optimize.  To
8406           use this code transformation, GCC has to be configured with
8407           --with-isl to enable the Graphite loop transformation
8408           infrastructure.
8409
8410       -fgraphite-identity
8411           Enable the identity transformation for graphite.  For every SCoP we
8412           generate the polyhedral representation and transform it back to
8413           gimple.  Using -fgraphite-identity we can check the costs or
8414           benefits of the GIMPLE -> GRAPHITE -> GIMPLE transformation.  Some
8415           minimal optimizations are also performed by the code generator isl,
8416           like index splitting and dead code elimination in loops.
8417
8418       -floop-nest-optimize
8419           Enable the isl based loop nest optimizer.  This is a generic loop
8420           nest optimizer based on the Pluto optimization algorithms.  It
8421           calculates a loop structure optimized for data-locality and
8422           parallelism.  This option is experimental.
8423
8424       -floop-parallelize-all
8425           Use the Graphite data dependence analysis to identify loops that
8426           can be parallelized.  Parallelize all the loops that can be
8427           analyzed to not contain loop carried dependences without checking
8428           that it is profitable to parallelize the loops.
8429
8430       -ftree-coalesce-vars
8431           While transforming the program out of the SSA representation,
8432           attempt to reduce copying by coalescing versions of different user-
8433           defined variables, instead of just compiler temporaries.  This may
8434           severely limit the ability to debug an optimized program compiled
8435           with -fno-var-tracking-assignments.  In the negated form, this flag
8436           prevents SSA coalescing of user variables.  This option is enabled
8437           by default if optimization is enabled, and it does very little
8438           otherwise.
8439
8440       -ftree-loop-if-convert
8441           Attempt to transform conditional jumps in the innermost loops to
8442           branch-less equivalents.  The intent is to remove control-flow from
8443           the innermost loops in order to improve the ability of the
8444           vectorization pass to handle these loops.  This is enabled by
8445           default if vectorization is enabled.
8446
8447       -ftree-loop-distribution
8448           Perform loop distribution.  This flag can improve cache performance
8449           on big loop bodies and allow further loop optimizations, like
8450           parallelization or vectorization, to take place.  For example, the
8451           loop
8452
8453                   DO I = 1, N
8454                     A(I) = B(I) + C
8455                     D(I) = E(I) * F
8456                   ENDDO
8457
8458           is transformed to
8459
8460                   DO I = 1, N
8461                      A(I) = B(I) + C
8462                   ENDDO
8463                   DO I = 1, N
8464                      D(I) = E(I) * F
8465                   ENDDO
8466
8467           This flag is enabled by default at -O3.  It is also enabled by
8468           -fprofile-use and -fauto-profile.
8469
8470       -ftree-loop-distribute-patterns
8471           Perform loop distribution of patterns that can be code generated
8472           with calls to a library.  This flag is enabled by default at -O2
8473           and higher, and by -fprofile-use and -fauto-profile.
8474
8475           This pass distributes the initialization loops and generates a call
8476           to memset zero.  For example, the loop
8477
8478                   DO I = 1, N
8479                     A(I) = 0
8480                     B(I) = A(I) + I
8481                   ENDDO
8482
8483           is transformed to
8484
8485                   DO I = 1, N
8486                      A(I) = 0
8487                   ENDDO
8488                   DO I = 1, N
8489                      B(I) = A(I) + I
8490                   ENDDO
8491
8492           and the initialization loop is transformed into a call to memset
8493           zero.  This flag is enabled by default at -O3.  It is also enabled
8494           by -fprofile-use and -fauto-profile.
8495
8496       -floop-interchange
8497           Perform loop interchange outside of graphite.  This flag can
8498           improve cache performance on loop nest and allow further loop
8499           optimizations, like vectorization, to take place.  For example, the
8500           loop
8501
8502                   for (int i = 0; i < N; i++)
8503                     for (int j = 0; j < N; j++)
8504                       for (int k = 0; k < N; k++)
8505                         c[i][j] = c[i][j] + a[i][k]*b[k][j];
8506
8507           is transformed to
8508
8509                   for (int i = 0; i < N; i++)
8510                     for (int k = 0; k < N; k++)
8511                       for (int j = 0; j < N; j++)
8512                         c[i][j] = c[i][j] + a[i][k]*b[k][j];
8513
8514           This flag is enabled by default at -O3.  It is also enabled by
8515           -fprofile-use and -fauto-profile.
8516
8517       -floop-unroll-and-jam
8518           Apply unroll and jam transformations on feasible loops.  In a loop
8519           nest this unrolls the outer loop by some factor and fuses the
8520           resulting multiple inner loops.  This flag is enabled by default at
8521           -O3.  It is also enabled by -fprofile-use and -fauto-profile.
8522
8523       -ftree-loop-im
8524           Perform loop invariant motion on trees.  This pass moves only
8525           invariants that are hard to handle at RTL level (function calls,
8526           operations that expand to nontrivial sequences of insns).  With
8527           -funswitch-loops it also moves operands of conditions that are
8528           invariant out of the loop, so that we can use just trivial
8529           invariantness analysis in loop unswitching.  The pass also includes
8530           store motion.
8531
8532       -ftree-loop-ivcanon
8533           Create a canonical counter for number of iterations in loops for
8534           which determining number of iterations requires complicated
8535           analysis.  Later optimizations then may determine the number
8536           easily.  Useful especially in connection with unrolling.
8537
8538       -ftree-scev-cprop
8539           Perform final value replacement.  If a variable is modified in a
8540           loop in such a way that its value when exiting the loop can be
8541           determined using only its initial value and the number of loop
8542           iterations, replace uses of the final value by such a computation,
8543           provided it is sufficiently cheap.  This reduces data dependencies
8544           and may allow further simplifications.  Enabled by default at -O
8545           and higher.
8546
8547       -fivopts
8548           Perform induction variable optimizations (strength reduction,
8549           induction variable merging and induction variable elimination) on
8550           trees.
8551
8552       -ftree-parallelize-loops=n
8553           Parallelize loops, i.e., split their iteration space to run in n
8554           threads.  This is only possible for loops whose iterations are
8555           independent and can be arbitrarily reordered.  The optimization is
8556           only profitable on multiprocessor machines, for loops that are CPU-
8557           intensive, rather than constrained e.g. by memory bandwidth.  This
8558           option implies -pthread, and thus is only supported on targets that
8559           have support for -pthread.
8560
8561       -ftree-pta
8562           Perform function-local points-to analysis on trees.  This flag is
8563           enabled by default at -O1 and higher, except for -Og.
8564
8565       -ftree-sra
8566           Perform scalar replacement of aggregates.  This pass replaces
8567           structure references with scalars to prevent committing structures
8568           to memory too early.  This flag is enabled by default at -O1 and
8569           higher, except for -Og.
8570
8571       -fstore-merging
8572           Perform merging of narrow stores to consecutive memory addresses.
8573           This pass merges contiguous stores of immediate values narrower
8574           than a word into fewer wider stores to reduce the number of
8575           instructions.  This is enabled by default at -O2 and higher as well
8576           as -Os.
8577
8578       -ftree-ter
8579           Perform temporary expression replacement during the SSA->normal
8580           phase.  Single use/single def temporaries are replaced at their use
8581           location with their defining expression.  This results in non-
8582           GIMPLE code, but gives the expanders much more complex trees to
8583           work on resulting in better RTL generation.  This is enabled by
8584           default at -O and higher.
8585
8586       -ftree-slsr
8587           Perform straight-line strength reduction on trees.  This recognizes
8588           related expressions involving multiplications and replaces them by
8589           less expensive calculations when possible.  This is enabled by
8590           default at -O and higher.
8591
8592       -ftree-vectorize
8593           Perform vectorization on trees. This flag enables
8594           -ftree-loop-vectorize and -ftree-slp-vectorize if not explicitly
8595           specified.
8596
8597       -ftree-loop-vectorize
8598           Perform loop vectorization on trees. This flag is enabled by
8599           default at -O3 and by -ftree-vectorize, -fprofile-use, and
8600           -fauto-profile.
8601
8602       -ftree-slp-vectorize
8603           Perform basic block vectorization on trees. This flag is enabled by
8604           default at -O3 and by -ftree-vectorize, -fprofile-use, and
8605           -fauto-profile.
8606
8607       -fvect-cost-model=model
8608           Alter the cost model used for vectorization.  The model argument
8609           should be one of unlimited, dynamic or cheap.  With the unlimited
8610           model the vectorized code-path is assumed to be profitable while
8611           with the dynamic model a runtime check guards the vectorized code-
8612           path to enable it only for iteration counts that will likely
8613           execute faster than when executing the original scalar loop.  The
8614           cheap model disables vectorization of loops where doing so would be
8615           cost prohibitive for example due to required runtime checks for
8616           data dependence or alignment but otherwise is equal to the dynamic
8617           model.  The default cost model depends on other optimization flags
8618           and is either dynamic or cheap.
8619
8620       -fsimd-cost-model=model
8621           Alter the cost model used for vectorization of loops marked with
8622           the OpenMP simd directive.  The model argument should be one of
8623           unlimited, dynamic, cheap.  All values of model have the same
8624           meaning as described in -fvect-cost-model and by default a cost
8625           model defined with -fvect-cost-model is used.
8626
8627       -ftree-vrp
8628           Perform Value Range Propagation on trees.  This is similar to the
8629           constant propagation pass, but instead of values, ranges of values
8630           are propagated.  This allows the optimizers to remove unnecessary
8631           range checks like array bound checks and null pointer checks.  This
8632           is enabled by default at -O2 and higher.  Null pointer check
8633           elimination is only done if -fdelete-null-pointer-checks is
8634           enabled.
8635
8636       -fsplit-paths
8637           Split paths leading to loop backedges.  This can improve dead code
8638           elimination and common subexpression elimination.  This is enabled
8639           by default at -O3 and above.
8640
8641       -fsplit-ivs-in-unroller
8642           Enables expression of values of induction variables in later
8643           iterations of the unrolled loop using the value in the first
8644           iteration.  This breaks long dependency chains, thus improving
8645           efficiency of the scheduling passes.
8646
8647           A combination of -fweb and CSE is often sufficient to obtain the
8648           same effect.  However, that is not reliable in cases where the loop
8649           body is more complicated than a single basic block.  It also does
8650           not work at all on some architectures due to restrictions in the
8651           CSE pass.
8652
8653           This optimization is enabled by default.
8654
8655       -fvariable-expansion-in-unroller
8656           With this option, the compiler creates multiple copies of some
8657           local variables when unrolling a loop, which can result in superior
8658           code.
8659
8660           This optimization is enabled by default for PowerPC targets, but
8661           disabled by default otherwise.
8662
8663       -fpartial-inlining
8664           Inline parts of functions.  This option has any effect only when
8665           inlining itself is turned on by the -finline-functions or
8666           -finline-small-functions options.
8667
8668           Enabled at levels -O2, -O3, -Os.
8669
8670       -fpredictive-commoning
8671           Perform predictive commoning optimization, i.e., reusing
8672           computations (especially memory loads and stores) performed in
8673           previous iterations of loops.
8674
8675           This option is enabled at level -O3.  It is also enabled by
8676           -fprofile-use and -fauto-profile.
8677
8678       -fprefetch-loop-arrays
8679           If supported by the target machine, generate instructions to
8680           prefetch memory to improve the performance of loops that access
8681           large arrays.
8682
8683           This option may generate better or worse code; results are highly
8684           dependent on the structure of loops within the source code.
8685
8686           Disabled at level -Os.
8687
8688       -fno-printf-return-value
8689           Do not substitute constants for known return value of formatted
8690           output functions such as "sprintf", "snprintf", "vsprintf", and
8691           "vsnprintf" (but not "printf" of "fprintf").  This transformation
8692           allows GCC to optimize or even eliminate branches based on the
8693           known return value of these functions called with arguments that
8694           are either constant, or whose values are known to be in a range
8695           that makes determining the exact return value possible.  For
8696           example, when -fprintf-return-value is in effect, both the branch
8697           and the body of the "if" statement (but not the call to "snprint")
8698           can be optimized away when "i" is a 32-bit or smaller integer
8699           because the return value is guaranteed to be at most 8.
8700
8701                   char buf[9];
8702                   if (snprintf (buf, "%08x", i) >= sizeof buf)
8703                     ...
8704
8705           The -fprintf-return-value option relies on other optimizations and
8706           yields best results with -O2 and above.  It works in tandem with
8707           the -Wformat-overflow and -Wformat-truncation options.  The
8708           -fprintf-return-value option is enabled by default.
8709
8710       -fno-peephole
8711       -fno-peephole2
8712           Disable any machine-specific peephole optimizations.  The
8713           difference between -fno-peephole and -fno-peephole2 is in how they
8714           are implemented in the compiler; some targets use one, some use the
8715           other, a few use both.
8716
8717           -fpeephole is enabled by default.  -fpeephole2 enabled at levels
8718           -O2, -O3, -Os.
8719
8720       -fno-guess-branch-probability
8721           Do not guess branch probabilities using heuristics.
8722
8723           GCC uses heuristics to guess branch probabilities if they are not
8724           provided by profiling feedback (-fprofile-arcs).  These heuristics
8725           are based on the control flow graph.  If some branch probabilities
8726           are specified by "__builtin_expect", then the heuristics are used
8727           to guess branch probabilities for the rest of the control flow
8728           graph, taking the "__builtin_expect" info into account.  The
8729           interactions between the heuristics and "__builtin_expect" can be
8730           complex, and in some cases, it may be useful to disable the
8731           heuristics so that the effects of "__builtin_expect" are easier to
8732           understand.
8733
8734           It is also possible to specify expected probability of the
8735           expression with "__builtin_expect_with_probability" built-in
8736           function.
8737
8738           The default is -fguess-branch-probability at levels -O, -O2, -O3,
8739           -Os.
8740
8741       -freorder-blocks
8742           Reorder basic blocks in the compiled function in order to reduce
8743           number of taken branches and improve code locality.
8744
8745           Enabled at levels -O, -O2, -O3, -Os.
8746
8747       -freorder-blocks-algorithm=algorithm
8748           Use the specified algorithm for basic block reordering.  The
8749           algorithm argument can be simple, which does not increase code size
8750           (except sometimes due to secondary effects like alignment), or stc,
8751           the "software trace cache" algorithm, which tries to put all often
8752           executed code together, minimizing the number of branches executed
8753           by making extra copies of code.
8754
8755           The default is simple at levels -O, -Os, and stc at levels -O2,
8756           -O3.
8757
8758       -freorder-blocks-and-partition
8759           In addition to reordering basic blocks in the compiled function, in
8760           order to reduce number of taken branches, partitions hot and cold
8761           basic blocks into separate sections of the assembly and .o files,
8762           to improve paging and cache locality performance.
8763
8764           This optimization is automatically turned off in the presence of
8765           exception handling or unwind tables (on targets using
8766           setjump/longjump or target specific scheme), for linkonce sections,
8767           for functions with a user-defined section attribute and on any
8768           architecture that does not support named sections.  When
8769           -fsplit-stack is used this option is not enabled by default (to
8770           avoid linker errors), but may be enabled explicitly (if using a
8771           working linker).
8772
8773           Enabled for x86 at levels -O2, -O3, -Os.
8774
8775       -freorder-functions
8776           Reorder functions in the object file in order to improve code
8777           locality.  This is implemented by using special subsections
8778           ".text.hot" for most frequently executed functions and
8779           ".text.unlikely" for unlikely executed functions.  Reordering is
8780           done by the linker so object file format must support named
8781           sections and linker must place them in a reasonable way.
8782
8783           This option isn't effective unless you either provide profile
8784           feedback (see -fprofile-arcs for details) or manually annotate
8785           functions with "hot" or "cold" attributes.
8786
8787           Enabled at levels -O2, -O3, -Os.
8788
8789       -fstrict-aliasing
8790           Allow the compiler to assume the strictest aliasing rules
8791           applicable to the language being compiled.  For C (and C++), this
8792           activates optimizations based on the type of expressions.  In
8793           particular, an object of one type is assumed never to reside at the
8794           same address as an object of a different type, unless the types are
8795           almost the same.  For example, an "unsigned int" can alias an
8796           "int", but not a "void*" or a "double".  A character type may alias
8797           any other type.
8798
8799           Pay special attention to code like this:
8800
8801                   union a_union {
8802                     int i;
8803                     double d;
8804                   };
8805
8806                   int f() {
8807                     union a_union t;
8808                     t.d = 3.0;
8809                     return t.i;
8810                   }
8811
8812           The practice of reading from a different union member than the one
8813           most recently written to (called "type-punning") is common.  Even
8814           with -fstrict-aliasing, type-punning is allowed, provided the
8815           memory is accessed through the union type.  So, the code above
8816           works as expected.    However, this code might not:
8817
8818                   int f() {
8819                     union a_union t;
8820                     int* ip;
8821                     t.d = 3.0;
8822                     ip = &t.i;
8823                     return *ip;
8824                   }
8825
8826           Similarly, access by taking the address, casting the resulting
8827           pointer and dereferencing the result has undefined behavior, even
8828           if the cast uses a union type, e.g.:
8829
8830                   int f() {
8831                     double d = 3.0;
8832                     return ((union a_union *) &d)->i;
8833                   }
8834
8835           The -fstrict-aliasing option is enabled at levels -O2, -O3, -Os.
8836
8837       -falign-functions
8838       -falign-functions=n
8839       -falign-functions=n:m
8840       -falign-functions=n:m:n2
8841       -falign-functions=n:m:n2:m2
8842           Align the start of functions to the next power-of-two greater than
8843           or equal to n, skipping up to m-1 bytes.  This ensures that at
8844           least the first m bytes of the function can be fetched by the CPU
8845           without crossing an n-byte alignment boundary.
8846
8847           If m is not specified, it defaults to n.
8848
8849           Examples: -falign-functions=32 aligns functions to the next 32-byte
8850           boundary, -falign-functions=24 aligns to the next 32-byte boundary
8851           only if this can be done by skipping 23 bytes or less,
8852           -falign-functions=32:7 aligns to the next 32-byte boundary only if
8853           this can be done by skipping 6 bytes or less.
8854
8855           The second pair of n2:m2 values allows you to specify a secondary
8856           alignment: -falign-functions=64:7:32:3 aligns to the next 64-byte
8857           boundary if this can be done by skipping 6 bytes or less, otherwise
8858           aligns to the next 32-byte boundary if this can be done by skipping
8859           2 bytes or less.  If m2 is not specified, it defaults to n2.
8860
8861           Some assemblers only support this flag when n is a power of two; in
8862           that case, it is rounded up.
8863
8864           -fno-align-functions and -falign-functions=1 are equivalent and
8865           mean that functions are not aligned.
8866
8867           If n is not specified or is zero, use a machine-dependent default.
8868           The maximum allowed n option value is 65536.
8869
8870           Enabled at levels -O2, -O3.
8871
8872       -flimit-function-alignment
8873           If this option is enabled, the compiler tries to avoid
8874           unnecessarily overaligning functions. It attempts to instruct the
8875           assembler to align by the amount specified by -falign-functions,
8876           but not to skip more bytes than the size of the function.
8877
8878       -falign-labels
8879       -falign-labels=n
8880       -falign-labels=n:m
8881       -falign-labels=n:m:n2
8882       -falign-labels=n:m:n2:m2
8883           Align all branch targets to a power-of-two boundary.
8884
8885           Parameters of this option are analogous to the -falign-functions
8886           option.  -fno-align-labels and -falign-labels=1 are equivalent and
8887           mean that labels are not aligned.
8888
8889           If -falign-loops or -falign-jumps are applicable and are greater
8890           than this value, then their values are used instead.
8891
8892           If n is not specified or is zero, use a machine-dependent default
8893           which is very likely to be 1, meaning no alignment.  The maximum
8894           allowed n option value is 65536.
8895
8896           Enabled at levels -O2, -O3.
8897
8898       -falign-loops
8899       -falign-loops=n
8900       -falign-loops=n:m
8901       -falign-loops=n:m:n2
8902       -falign-loops=n:m:n2:m2
8903           Align loops to a power-of-two boundary.  If the loops are executed
8904           many times, this makes up for any execution of the dummy padding
8905           instructions.
8906
8907           If -falign-labels is greater than this value, then its value is
8908           used instead.
8909
8910           Parameters of this option are analogous to the -falign-functions
8911           option.  -fno-align-loops and -falign-loops=1 are equivalent and
8912           mean that loops are not aligned.  The maximum allowed n option
8913           value is 65536.
8914
8915           If n is not specified or is zero, use a machine-dependent default.
8916
8917           Enabled at levels -O2, -O3.
8918
8919       -falign-jumps
8920       -falign-jumps=n
8921       -falign-jumps=n:m
8922       -falign-jumps=n:m:n2
8923       -falign-jumps=n:m:n2:m2
8924           Align branch targets to a power-of-two boundary, for branch targets
8925           where the targets can only be reached by jumping.  In this case, no
8926           dummy operations need be executed.
8927
8928           If -falign-labels is greater than this value, then its value is
8929           used instead.
8930
8931           Parameters of this option are analogous to the -falign-functions
8932           option.  -fno-align-jumps and -falign-jumps=1 are equivalent and
8933           mean that loops are not aligned.
8934
8935           If n is not specified or is zero, use a machine-dependent default.
8936           The maximum allowed n option value is 65536.
8937
8938           Enabled at levels -O2, -O3.
8939
8940       -fno-allocation-dce
8941           Do not remove unused C++ allocations in dead code elimination.
8942
8943       -fallow-store-data-races
8944           Allow the compiler to introduce new data races on stores.
8945
8946           Enabled at level -Ofast.
8947
8948       -funit-at-a-time
8949           This option is left for compatibility reasons. -funit-at-a-time has
8950           no effect, while -fno-unit-at-a-time implies -fno-toplevel-reorder
8951           and -fno-section-anchors.
8952
8953           Enabled by default.
8954
8955       -fno-toplevel-reorder
8956           Do not reorder top-level functions, variables, and "asm"
8957           statements.  Output them in the same order that they appear in the
8958           input file.  When this option is used, unreferenced static
8959           variables are not removed.  This option is intended to support
8960           existing code that relies on a particular ordering.  For new code,
8961           it is better to use attributes when possible.
8962
8963           -ftoplevel-reorder is the default at -O1 and higher, and also at
8964           -O0 if -fsection-anchors is explicitly requested.  Additionally
8965           -fno-toplevel-reorder implies -fno-section-anchors.
8966
8967       -fweb
8968           Constructs webs as commonly used for register allocation purposes
8969           and assign each web individual pseudo register.  This allows the
8970           register allocation pass to operate on pseudos directly, but also
8971           strengthens several other optimization passes, such as CSE, loop
8972           optimizer and trivial dead code remover.  It can, however, make
8973           debugging impossible, since variables no longer stay in a "home
8974           register".
8975
8976           Enabled by default with -funroll-loops.
8977
8978       -fwhole-program
8979           Assume that the current compilation unit represents the whole
8980           program being compiled.  All public functions and variables with
8981           the exception of "main" and those merged by attribute
8982           "externally_visible" become static functions and in effect are
8983           optimized more aggressively by interprocedural optimizers.
8984
8985           This option should not be used in combination with -flto.  Instead
8986           relying on a linker plugin should provide safer and more precise
8987           information.
8988
8989       -flto[=n]
8990           This option runs the standard link-time optimizer.  When invoked
8991           with source code, it generates GIMPLE (one of GCC's internal
8992           representations) and writes it to special ELF sections in the
8993           object file.  When the object files are linked together, all the
8994           function bodies are read from these ELF sections and instantiated
8995           as if they had been part of the same translation unit.
8996
8997           To use the link-time optimizer, -flto and optimization options
8998           should be specified at compile time and during the final link.  It
8999           is recommended that you compile all the files participating in the
9000           same link with the same options and also specify those options at
9001           link time.  For example:
9002
9003                   gcc -c -O2 -flto foo.c
9004                   gcc -c -O2 -flto bar.c
9005                   gcc -o myprog -flto -O2 foo.o bar.o
9006
9007           The first two invocations to GCC save a bytecode representation of
9008           GIMPLE into special ELF sections inside foo.o and bar.o.  The final
9009           invocation reads the GIMPLE bytecode from foo.o and bar.o, merges
9010           the two files into a single internal image, and compiles the result
9011           as usual.  Since both foo.o and bar.o are merged into a single
9012           image, this causes all the interprocedural analyses and
9013           optimizations in GCC to work across the two files as if they were a
9014           single one.  This means, for example, that the inliner is able to
9015           inline functions in bar.o into functions in foo.o and vice-versa.
9016
9017           Another (simpler) way to enable link-time optimization is:
9018
9019                   gcc -o myprog -flto -O2 foo.c bar.c
9020
9021           The above generates bytecode for foo.c and bar.c, merges them
9022           together into a single GIMPLE representation and optimizes them as
9023           usual to produce myprog.
9024
9025           The important thing to keep in mind is that to enable link-time
9026           optimizations you need to use the GCC driver to perform the link
9027           step.  GCC automatically performs link-time optimization if any of
9028           the objects involved were compiled with the -flto command-line
9029           option.  You can always override the automatic decision to do link-
9030           time optimization by passing -fno-lto to the link command.
9031
9032           To make whole program optimization effective, it is necessary to
9033           make certain whole program assumptions.  The compiler needs to know
9034           what functions and variables can be accessed by libraries and
9035           runtime outside of the link-time optimized unit.  When supported by
9036           the linker, the linker plugin (see -fuse-linker-plugin) passes
9037           information to the compiler about used and externally visible
9038           symbols.  When the linker plugin is not available, -fwhole-program
9039           should be used to allow the compiler to make these assumptions,
9040           which leads to more aggressive optimization decisions.
9041
9042           When a file is compiled with -flto without -fuse-linker-plugin, the
9043           generated object file is larger than a regular object file because
9044           it contains GIMPLE bytecodes and the usual final code (see
9045           -ffat-lto-objects.  This means that object files with LTO
9046           information can be linked as normal object files; if -fno-lto is
9047           passed to the linker, no interprocedural optimizations are applied.
9048           Note that when -fno-fat-lto-objects is enabled the compile stage is
9049           faster but you cannot perform a regular, non-LTO link on them.
9050
9051           When producing the final binary, GCC only applies link-time
9052           optimizations to those files that contain bytecode.  Therefore, you
9053           can mix and match object files and libraries with GIMPLE bytecodes
9054           and final object code.  GCC automatically selects which files to
9055           optimize in LTO mode and which files to link without further
9056           processing.
9057
9058           Generally, options specified at link time override those specified
9059           at compile time, although in some cases GCC attempts to infer link-
9060           time options from the settings used to compile the input files.
9061
9062           If you do not specify an optimization level option -O at link time,
9063           then GCC uses the highest optimization level used when compiling
9064           the object files.  Note that it is generally ineffective to specify
9065           an optimization level option only at link time and not at compile
9066           time, for two reasons.  First, compiling without optimization
9067           suppresses compiler passes that gather information needed for
9068           effective optimization at link time.  Second, some early
9069           optimization passes can be performed only at compile time and not
9070           at link time.
9071
9072           There are some code generation flags preserved by GCC when
9073           generating bytecodes, as they need to be used during the final
9074           link.  Currently, the following options and their settings are
9075           taken from the first object file that explicitly specifies them:
9076           -fPIC, -fpic, -fpie, -fcommon, -fexceptions, -fnon-call-exceptions,
9077           -fgnu-tm and all the -m target flags.
9078
9079           Certain ABI-changing flags are required to match in all compilation
9080           units, and trying to override this at link time with a conflicting
9081           value is ignored.  This includes options such as
9082           -freg-struct-return and -fpcc-struct-return.
9083
9084           Other options such as -ffp-contract, -fno-strict-overflow, -fwrapv,
9085           -fno-trapv or -fno-strict-aliasing are passed through to the link
9086           stage and merged conservatively for conflicting translation units.
9087           Specifically -fno-strict-overflow, -fwrapv and -fno-trapv take
9088           precedence; and for example -ffp-contract=off takes precedence over
9089           -ffp-contract=fast.  You can override them at link time.
9090
9091           Diagnostic options such as -Wstringop-overflow are passed through
9092           to the link stage and their setting matches that of the compile-
9093           step at function granularity.  Note that this matters only for
9094           diagnostics emitted during optimization.  Note that code transforms
9095           such as inlining can lead to warnings being enabled or disabled for
9096           regions if code not consistent with the setting at compile time.
9097
9098           When you need to pass options to the assembler via -Wa or
9099           -Xassembler make sure to either compile such translation units with
9100           -fno-lto or consistently use the same assembler options on all
9101           translation units.  You can alternatively also specify assembler
9102           options at LTO link time.
9103
9104           To enable debug info generation you need to supply -g at compile
9105           time.  If any of the input files at link time were built with debug
9106           info generation enabled the link will enable debug info generation
9107           as well.  Any elaborate debug info settings like the dwarf level
9108           -gdwarf-5 need to be explicitly repeated at the linker command line
9109           and mixing different settings in different translation units is
9110           discouraged.
9111
9112           If LTO encounters objects with C linkage declared with incompatible
9113           types in separate translation units to be linked together
9114           (undefined behavior according to ISO C99 6.2.7), a non-fatal
9115           diagnostic may be issued.  The behavior is still undefined at run
9116           time.  Similar diagnostics may be raised for other languages.
9117
9118           Another feature of LTO is that it is possible to apply
9119           interprocedural optimizations on files written in different
9120           languages:
9121
9122                   gcc -c -flto foo.c
9123                   g++ -c -flto bar.cc
9124                   gfortran -c -flto baz.f90
9125                   g++ -o myprog -flto -O3 foo.o bar.o baz.o -lgfortran
9126
9127           Notice that the final link is done with g++ to get the C++ runtime
9128           libraries and -lgfortran is added to get the Fortran runtime
9129           libraries.  In general, when mixing languages in LTO mode, you
9130           should use the same link command options as when mixing languages
9131           in a regular (non-LTO) compilation.
9132
9133           If object files containing GIMPLE bytecode are stored in a library
9134           archive, say libfoo.a, it is possible to extract and use them in an
9135           LTO link if you are using a linker with plugin support.  To create
9136           static libraries suitable for LTO, use gcc-ar and gcc-ranlib
9137           instead of ar and ranlib; to show the symbols of object files with
9138           GIMPLE bytecode, use gcc-nm.  Those commands require that ar,
9139           ranlib and nm have been compiled with plugin support.  At link
9140           time, use the flag -fuse-linker-plugin to ensure that the library
9141           participates in the LTO optimization process:
9142
9143                   gcc -o myprog -O2 -flto -fuse-linker-plugin a.o b.o -lfoo
9144
9145           With the linker plugin enabled, the linker extracts the needed
9146           GIMPLE files from libfoo.a and passes them on to the running GCC to
9147           make them part of the aggregated GIMPLE image to be optimized.
9148
9149           If you are not using a linker with plugin support and/or do not
9150           enable the linker plugin, then the objects inside libfoo.a are
9151           extracted and linked as usual, but they do not participate in the
9152           LTO optimization process.  In order to make a static library
9153           suitable for both LTO optimization and usual linkage, compile its
9154           object files with -flto -ffat-lto-objects.
9155
9156           Link-time optimizations do not require the presence of the whole
9157           program to operate.  If the program does not require any symbols to
9158           be exported, it is possible to combine -flto and -fwhole-program to
9159           allow the interprocedural optimizers to use more aggressive
9160           assumptions which may lead to improved optimization opportunities.
9161           Use of -fwhole-program is not needed when linker plugin is active
9162           (see -fuse-linker-plugin).
9163
9164           The current implementation of LTO makes no attempt to generate
9165           bytecode that is portable between different types of hosts.  The
9166           bytecode files are versioned and there is a strict version check,
9167           so bytecode files generated in one version of GCC do not work with
9168           an older or newer version of GCC.
9169
9170           Link-time optimization does not work well with generation of
9171           debugging information on systems other than those using a
9172           combination of ELF and DWARF.
9173
9174           If you specify the optional n, the optimization and code generation
9175           done at link time is executed in parallel using n parallel jobs by
9176           utilizing an installed make program.  The environment variable MAKE
9177           may be used to override the program used.
9178
9179           You can also specify -flto=jobserver to use GNU make's job server
9180           mode to determine the number of parallel jobs. This is useful when
9181           the Makefile calling GCC is already executing in parallel.  You
9182           must prepend a + to the command recipe in the parent Makefile for
9183           this to work.  This option likely only works if MAKE is GNU make.
9184           Even without the option value, GCC tries to automatically detect a
9185           running GNU make's job server.
9186
9187           Use -flto=auto to use GNU make's job server, if available, or
9188           otherwise fall back to autodetection of the number of CPU threads
9189           present in your system.
9190
9191       -flto-partition=alg
9192           Specify the partitioning algorithm used by the link-time optimizer.
9193           The value is either 1to1 to specify a partitioning mirroring the
9194           original source files or balanced to specify partitioning into
9195           equally sized chunks (whenever possible) or max to create new
9196           partition for every symbol where possible.  Specifying none as an
9197           algorithm disables partitioning and streaming completely.  The
9198           default value is balanced. While 1to1 can be used as an workaround
9199           for various code ordering issues, the max partitioning is intended
9200           for internal testing only.  The value one specifies that exactly
9201           one partition should be used while the value none bypasses
9202           partitioning and executes the link-time optimization step directly
9203           from the WPA phase.
9204
9205       -flto-compression-level=n
9206           This option specifies the level of compression used for
9207           intermediate language written to LTO object files, and is only
9208           meaningful in conjunction with LTO mode (-flto).  Valid values are
9209           0 (no compression) to 9 (maximum compression).  Values outside this
9210           range are clamped to either 0 or 9.  If the option is not given, a
9211           default balanced compression setting is used.
9212
9213       -fuse-linker-plugin
9214           Enables the use of a linker plugin during link-time optimization.
9215           This option relies on plugin support in the linker, which is
9216           available in gold or in GNU ld 2.21 or newer.
9217
9218           This option enables the extraction of object files with GIMPLE
9219           bytecode out of library archives. This improves the quality of
9220           optimization by exposing more code to the link-time optimizer.
9221           This information specifies what symbols can be accessed externally
9222           (by non-LTO object or during dynamic linking).  Resulting code
9223           quality improvements on binaries (and shared libraries that use
9224           hidden visibility) are similar to -fwhole-program.  See -flto for a
9225           description of the effect of this flag and how to use it.
9226
9227           This option is enabled by default when LTO support in GCC is
9228           enabled and GCC was configured for use with a linker supporting
9229           plugins (GNU ld 2.21 or newer or gold).
9230
9231       -ffat-lto-objects
9232           Fat LTO objects are object files that contain both the intermediate
9233           language and the object code. This makes them usable for both LTO
9234           linking and normal linking. This option is effective only when
9235           compiling with -flto and is ignored at link time.
9236
9237           -fno-fat-lto-objects improves compilation time over plain LTO, but
9238           requires the complete toolchain to be aware of LTO. It requires a
9239           linker with linker plugin support for basic functionality.
9240           Additionally, nm, ar and ranlib need to support linker plugins to
9241           allow a full-featured build environment (capable of building static
9242           libraries etc).  GCC provides the gcc-ar, gcc-nm, gcc-ranlib
9243           wrappers to pass the right options to these tools. With non fat LTO
9244           makefiles need to be modified to use them.
9245
9246           Note that modern binutils provide plugin auto-load mechanism.
9247           Installing the linker plugin into $libdir/bfd-plugins has the same
9248           effect as usage of the command wrappers (gcc-ar, gcc-nm and gcc-
9249           ranlib).
9250
9251           The default is -fno-fat-lto-objects on targets with linker plugin
9252           support.
9253
9254       -fcompare-elim
9255           After register allocation and post-register allocation instruction
9256           splitting, identify arithmetic instructions that compute processor
9257           flags similar to a comparison operation based on that arithmetic.
9258           If possible, eliminate the explicit comparison operation.
9259
9260           This pass only applies to certain targets that cannot explicitly
9261           represent the comparison operation before register allocation is
9262           complete.
9263
9264           Enabled at levels -O, -O2, -O3, -Os.
9265
9266       -fcprop-registers
9267           After register allocation and post-register allocation instruction
9268           splitting, perform a copy-propagation pass to try to reduce
9269           scheduling dependencies and occasionally eliminate the copy.
9270
9271           Enabled at levels -O, -O2, -O3, -Os.
9272
9273       -fprofile-correction
9274           Profiles collected using an instrumented binary for multi-threaded
9275           programs may be inconsistent due to missed counter updates. When
9276           this option is specified, GCC uses heuristics to correct or smooth
9277           out such inconsistencies. By default, GCC emits an error message
9278           when an inconsistent profile is detected.
9279
9280           This option is enabled by -fauto-profile.
9281
9282       -fprofile-partial-training
9283           With "-fprofile-use" all portions of programs not executed during
9284           train run are optimized agressively for size rather than speed.  In
9285           some cases it is not practical to train all possible hot paths in
9286           the program. (For example, program may contain functions specific
9287           for a given hardware and trianing may not cover all hardware
9288           configurations program is run on.)  With
9289           "-fprofile-partial-training" profile feedback will be ignored for
9290           all functions not executed during the train run leading them to be
9291           optimized as if they were compiled without profile feedback. This
9292           leads to better performance when train run is not representative
9293           but also leads to significantly bigger code.
9294
9295       -fprofile-use
9296       -fprofile-use=path
9297           Enable profile feedback-directed optimizations, and the following
9298           optimizations, many of which are generally profitable only with
9299           profile feedback available:
9300
9301           -fbranch-probabilities  -fprofile-values -funroll-loops
9302           -fpeel-loops  -ftracer  -fvpt -finline-functions  -fipa-cp
9303           -fipa-cp-clone  -fipa-bit-cp -fpredictive-commoning  -fsplit-loops
9304           -funswitch-loops -fgcse-after-reload  -ftree-loop-vectorize
9305           -ftree-slp-vectorize -fvect-cost-model=dynamic
9306           -ftree-loop-distribute-patterns -fprofile-reorder-functions
9307
9308           Before you can use this option, you must first generate profiling
9309           information.
9310
9311           By default, GCC emits an error message if the feedback profiles do
9312           not match the source code.  This error can be turned into a warning
9313           by using -Wno-error=coverage-mismatch.  Note this may result in
9314           poorly optimized code.  Additionally, by default, GCC also emits a
9315           warning message if the feedback profiles do not exist (see
9316           -Wmissing-profile).
9317
9318           If path is specified, GCC looks at the path to find the profile
9319           feedback data files. See -fprofile-dir.
9320
9321       -fauto-profile
9322       -fauto-profile=path
9323           Enable sampling-based feedback-directed optimizations, and the
9324           following optimizations, many of which are generally profitable
9325           only with profile feedback available:
9326
9327           -fbranch-probabilities  -fprofile-values -funroll-loops
9328           -fpeel-loops  -ftracer  -fvpt -finline-functions  -fipa-cp
9329           -fipa-cp-clone  -fipa-bit-cp -fpredictive-commoning  -fsplit-loops
9330           -funswitch-loops -fgcse-after-reload  -ftree-loop-vectorize
9331           -ftree-slp-vectorize -fvect-cost-model=dynamic
9332           -ftree-loop-distribute-patterns -fprofile-correction
9333
9334           path is the name of a file containing AutoFDO profile information.
9335           If omitted, it defaults to fbdata.afdo in the current directory.
9336
9337           Producing an AutoFDO profile data file requires running your
9338           program with the perf utility on a supported GNU/Linux target
9339           system.  For more information, see <https://perf.wiki.kernel.org/>.
9340
9341           E.g.
9342
9343                   perf record -e br_inst_retired:near_taken -b -o perf.data \
9344                       -- your_program
9345
9346           Then use the create_gcov tool to convert the raw profile data to a
9347           format that can be used by GCC.  You must also supply the
9348           unstripped binary for your program to this tool.  See
9349           <https://github.com/google/autofdo>.
9350
9351           E.g.
9352
9353                   create_gcov --binary=your_program.unstripped --profile=perf.data \
9354                       --gcov=profile.afdo
9355
9356       The following options control compiler behavior regarding floating-
9357       point arithmetic.  These options trade off between speed and
9358       correctness.  All must be specifically enabled.
9359
9360       -ffloat-store
9361           Do not store floating-point variables in registers, and inhibit
9362           other options that might change whether a floating-point value is
9363           taken from a register or memory.
9364
9365           This option prevents undesirable excess precision on machines such
9366           as the 68000 where the floating registers (of the 68881) keep more
9367           precision than a "double" is supposed to have.  Similarly for the
9368           x86 architecture.  For most programs, the excess precision does
9369           only good, but a few programs rely on the precise definition of
9370           IEEE floating point.  Use -ffloat-store for such programs, after
9371           modifying them to store all pertinent intermediate computations
9372           into variables.
9373
9374       -fexcess-precision=style
9375           This option allows further control over excess precision on
9376           machines where floating-point operations occur in a format with
9377           more precision or range than the IEEE standard and interchange
9378           floating-point types.  By default, -fexcess-precision=fast is in
9379           effect; this means that operations may be carried out in a wider
9380           precision than the types specified in the source if that would
9381           result in faster code, and it is unpredictable when rounding to the
9382           types specified in the source code takes place.  When compiling C,
9383           if -fexcess-precision=standard is specified then excess precision
9384           follows the rules specified in ISO C99; in particular, both casts
9385           and assignments cause values to be rounded to their semantic types
9386           (whereas -ffloat-store only affects assignments).  This option is
9387           enabled by default for C if a strict conformance option such as
9388           -std=c99 is used.  -ffast-math enables -fexcess-precision=fast by
9389           default regardless of whether a strict conformance option is used.
9390
9391           -fexcess-precision=standard is not implemented for languages other
9392           than C.  On the x86, it has no effect if -mfpmath=sse or
9393           -mfpmath=sse+387 is specified; in the former case, IEEE semantics
9394           apply without excess precision, and in the latter, rounding is
9395           unpredictable.
9396
9397       -ffast-math
9398           Sets the options -fno-math-errno, -funsafe-math-optimizations,
9399           -ffinite-math-only, -fno-rounding-math, -fno-signaling-nans,
9400           -fcx-limited-range and -fexcess-precision=fast.
9401
9402           This option causes the preprocessor macro "__FAST_MATH__" to be
9403           defined.
9404
9405           This option is not turned on by any -O option besides -Ofast since
9406           it can result in incorrect output for programs that depend on an
9407           exact implementation of IEEE or ISO rules/specifications for math
9408           functions. It may, however, yield faster code for programs that do
9409           not require the guarantees of these specifications.
9410
9411       -fno-math-errno
9412           Do not set "errno" after calling math functions that are executed
9413           with a single instruction, e.g., "sqrt".  A program that relies on
9414           IEEE exceptions for math error handling may want to use this flag
9415           for speed while maintaining IEEE arithmetic compatibility.
9416
9417           This option is not turned on by any -O option since it can result
9418           in incorrect output for programs that depend on an exact
9419           implementation of IEEE or ISO rules/specifications for math
9420           functions. It may, however, yield faster code for programs that do
9421           not require the guarantees of these specifications.
9422
9423           The default is -fmath-errno.
9424
9425           On Darwin systems, the math library never sets "errno".  There is
9426           therefore no reason for the compiler to consider the possibility
9427           that it might, and -fno-math-errno is the default.
9428
9429       -funsafe-math-optimizations
9430           Allow optimizations for floating-point arithmetic that (a) assume
9431           that arguments and results are valid and (b) may violate IEEE or
9432           ANSI standards.  When used at link time, it may include libraries
9433           or startup files that change the default FPU control word or other
9434           similar optimizations.
9435
9436           This option is not turned on by any -O option since it can result
9437           in incorrect output for programs that depend on an exact
9438           implementation of IEEE or ISO rules/specifications for math
9439           functions. It may, however, yield faster code for programs that do
9440           not require the guarantees of these specifications.  Enables
9441           -fno-signed-zeros, -fno-trapping-math, -fassociative-math and
9442           -freciprocal-math.
9443
9444           The default is -fno-unsafe-math-optimizations.
9445
9446       -fassociative-math
9447           Allow re-association of operands in series of floating-point
9448           operations.  This violates the ISO C and C++ language standard by
9449           possibly changing computation result.  NOTE: re-ordering may change
9450           the sign of zero as well as ignore NaNs and inhibit or create
9451           underflow or overflow (and thus cannot be used on code that relies
9452           on rounding behavior like "(x + 2**52) - 2**52".  May also reorder
9453           floating-point comparisons and thus may not be used when ordered
9454           comparisons are required.  This option requires that both
9455           -fno-signed-zeros and -fno-trapping-math be in effect.  Moreover,
9456           it doesn't make much sense with -frounding-math. For Fortran the
9457           option is automatically enabled when both -fno-signed-zeros and
9458           -fno-trapping-math are in effect.
9459
9460           The default is -fno-associative-math.
9461
9462       -freciprocal-math
9463           Allow the reciprocal of a value to be used instead of dividing by
9464           the value if this enables optimizations.  For example "x / y" can
9465           be replaced with "x * (1/y)", which is useful if "(1/y)" is subject
9466           to common subexpression elimination.  Note that this loses
9467           precision and increases the number of flops operating on the value.
9468
9469           The default is -fno-reciprocal-math.
9470
9471       -ffinite-math-only
9472           Allow optimizations for floating-point arithmetic that assume that
9473           arguments and results are not NaNs or +-Infs.
9474
9475           This option is not turned on by any -O option since it can result
9476           in incorrect output for programs that depend on an exact
9477           implementation of IEEE or ISO rules/specifications for math
9478           functions. It may, however, yield faster code for programs that do
9479           not require the guarantees of these specifications.
9480
9481           The default is -fno-finite-math-only.
9482
9483       -fno-signed-zeros
9484           Allow optimizations for floating-point arithmetic that ignore the
9485           signedness of zero.  IEEE arithmetic specifies the behavior of
9486           distinct +0.0 and -0.0 values, which then prohibits simplification
9487           of expressions such as x+0.0 or 0.0*x (even with
9488           -ffinite-math-only).  This option implies that the sign of a zero
9489           result isn't significant.
9490
9491           The default is -fsigned-zeros.
9492
9493       -fno-trapping-math
9494           Compile code assuming that floating-point operations cannot
9495           generate user-visible traps.  These traps include division by zero,
9496           overflow, underflow, inexact result and invalid operation.  This
9497           option requires that -fno-signaling-nans be in effect.  Setting
9498           this option may allow faster code if one relies on "non-stop" IEEE
9499           arithmetic, for example.
9500
9501           This option should never be turned on by any -O option since it can
9502           result in incorrect output for programs that depend on an exact
9503           implementation of IEEE or ISO rules/specifications for math
9504           functions.
9505
9506           The default is -ftrapping-math.
9507
9508       -frounding-math
9509           Disable transformations and optimizations that assume default
9510           floating-point rounding behavior.  This is round-to-zero for all
9511           floating point to integer conversions, and round-to-nearest for all
9512           other arithmetic truncations.  This option should be specified for
9513           programs that change the FP rounding mode dynamically, or that may
9514           be executed with a non-default rounding mode.  This option disables
9515           constant folding of floating-point expressions at compile time
9516           (which may be affected by rounding mode) and arithmetic
9517           transformations that are unsafe in the presence of sign-dependent
9518           rounding modes.
9519
9520           The default is -fno-rounding-math.
9521
9522           This option is experimental and does not currently guarantee to
9523           disable all GCC optimizations that are affected by rounding mode.
9524           Future versions of GCC may provide finer control of this setting
9525           using C99's "FENV_ACCESS" pragma.  This command-line option will be
9526           used to specify the default state for "FENV_ACCESS".
9527
9528       -fsignaling-nans
9529           Compile code assuming that IEEE signaling NaNs may generate user-
9530           visible traps during floating-point operations.  Setting this
9531           option disables optimizations that may change the number of
9532           exceptions visible with signaling NaNs.  This option implies
9533           -ftrapping-math.
9534
9535           This option causes the preprocessor macro "__SUPPORT_SNAN__" to be
9536           defined.
9537
9538           The default is -fno-signaling-nans.
9539
9540           This option is experimental and does not currently guarantee to
9541           disable all GCC optimizations that affect signaling NaN behavior.
9542
9543       -fno-fp-int-builtin-inexact
9544           Do not allow the built-in functions "ceil", "floor", "round" and
9545           "trunc", and their "float" and "long double" variants, to generate
9546           code that raises the "inexact" floating-point exception for
9547           noninteger arguments.  ISO C99 and C11 allow these functions to
9548           raise the "inexact" exception, but ISO/IEC TS 18661-1:2014, the C
9549           bindings to IEEE 754-2008, as integrated into ISO C2X, does not
9550           allow these functions to do so.
9551
9552           The default is -ffp-int-builtin-inexact, allowing the exception to
9553           be raised, unless C2X or a later C standard is selected.  This
9554           option does nothing unless -ftrapping-math is in effect.
9555
9556           Even if -fno-fp-int-builtin-inexact is used, if the functions
9557           generate a call to a library function then the "inexact" exception
9558           may be raised if the library implementation does not follow TS
9559           18661.
9560
9561       -fsingle-precision-constant
9562           Treat floating-point constants as single precision instead of
9563           implicitly converting them to double-precision constants.
9564
9565       -fcx-limited-range
9566           When enabled, this option states that a range reduction step is not
9567           needed when performing complex division.  Also, there is no
9568           checking whether the result of a complex multiplication or division
9569           is "NaN + I*NaN", with an attempt to rescue the situation in that
9570           case.  The default is -fno-cx-limited-range, but is enabled by
9571           -ffast-math.
9572
9573           This option controls the default setting of the ISO C99
9574           "CX_LIMITED_RANGE" pragma.  Nevertheless, the option applies to all
9575           languages.
9576
9577       -fcx-fortran-rules
9578           Complex multiplication and division follow Fortran rules.  Range
9579           reduction is done as part of complex division, but there is no
9580           checking whether the result of a complex multiplication or division
9581           is "NaN + I*NaN", with an attempt to rescue the situation in that
9582           case.
9583
9584           The default is -fno-cx-fortran-rules.
9585
9586       The following options control optimizations that may improve
9587       performance, but are not enabled by any -O options.  This section
9588       includes experimental options that may produce broken code.
9589
9590       -fbranch-probabilities
9591           After running a program compiled with -fprofile-arcs, you can
9592           compile it a second time using -fbranch-probabilities, to improve
9593           optimizations based on the number of times each branch was taken.
9594           When a program compiled with -fprofile-arcs exits, it saves arc
9595           execution counts to a file called sourcename.gcda for each source
9596           file.  The information in this data file is very dependent on the
9597           structure of the generated code, so you must use the same source
9598           code and the same optimization options for both compilations.
9599
9600           With -fbranch-probabilities, GCC puts a REG_BR_PROB note on each
9601           JUMP_INSN and CALL_INSN.  These can be used to improve
9602           optimization.  Currently, they are only used in one place: in
9603           reorg.c, instead of guessing which path a branch is most likely to
9604           take, the REG_BR_PROB values are used to exactly determine which
9605           path is taken more often.
9606
9607           Enabled by -fprofile-use and -fauto-profile.
9608
9609       -fprofile-values
9610           If combined with -fprofile-arcs, it adds code so that some data
9611           about values of expressions in the program is gathered.
9612
9613           With -fbranch-probabilities, it reads back the data gathered from
9614           profiling values of expressions for usage in optimizations.
9615
9616           Enabled by -fprofile-generate, -fprofile-use, and -fauto-profile.
9617
9618       -fprofile-reorder-functions
9619           Function reordering based on profile instrumentation collects first
9620           time of execution of a function and orders these functions in
9621           ascending order.
9622
9623           Enabled with -fprofile-use.
9624
9625       -fvpt
9626           If combined with -fprofile-arcs, this option instructs the compiler
9627           to add code to gather information about values of expressions.
9628
9629           With -fbranch-probabilities, it reads back the data gathered and
9630           actually performs the optimizations based on them.  Currently the
9631           optimizations include specialization of division operations using
9632           the knowledge about the value of the denominator.
9633
9634           Enabled with -fprofile-use and -fauto-profile.
9635
9636       -frename-registers
9637           Attempt to avoid false dependencies in scheduled code by making use
9638           of registers left over after register allocation.  This
9639           optimization most benefits processors with lots of registers.
9640           Depending on the debug information format adopted by the target,
9641           however, it can make debugging impossible, since variables no
9642           longer stay in a "home register".
9643
9644           Enabled by default with -funroll-loops.
9645
9646       -fschedule-fusion
9647           Performs a target dependent pass over the instruction stream to
9648           schedule instructions of same type together because target machine
9649           can execute them more efficiently if they are adjacent to each
9650           other in the instruction flow.
9651
9652           Enabled at levels -O2, -O3, -Os.
9653
9654       -ftracer
9655           Perform tail duplication to enlarge superblock size.  This
9656           transformation simplifies the control flow of the function allowing
9657           other optimizations to do a better job.
9658
9659           Enabled by -fprofile-use and -fauto-profile.
9660
9661       -funroll-loops
9662           Unroll loops whose number of iterations can be determined at
9663           compile time or upon entry to the loop.  -funroll-loops implies
9664           -frerun-cse-after-loop, -fweb and -frename-registers.  It also
9665           turns on complete loop peeling (i.e. complete removal of loops with
9666           a small constant number of iterations).  This option makes code
9667           larger, and may or may not make it run faster.
9668
9669           Enabled by -fprofile-use and -fauto-profile.
9670
9671       -funroll-all-loops
9672           Unroll all loops, even if their number of iterations is uncertain
9673           when the loop is entered.  This usually makes programs run more
9674           slowly.  -funroll-all-loops implies the same options as
9675           -funroll-loops.
9676
9677       -fpeel-loops
9678           Peels loops for which there is enough information that they do not
9679           roll much (from profile feedback or static analysis).  It also
9680           turns on complete loop peeling (i.e. complete removal of loops with
9681           small constant number of iterations).
9682
9683           Enabled by -O3, -fprofile-use, and -fauto-profile.
9684
9685       -fmove-loop-invariants
9686           Enables the loop invariant motion pass in the RTL loop optimizer.
9687           Enabled at level -O1 and higher, except for -Og.
9688
9689       -fsplit-loops
9690           Split a loop into two if it contains a condition that's always true
9691           for one side of the iteration space and false for the other.
9692
9693           Enabled by -fprofile-use and -fauto-profile.
9694
9695       -funswitch-loops
9696           Move branches with loop invariant conditions out of the loop, with
9697           duplicates of the loop on both branches (modified according to
9698           result of the condition).
9699
9700           Enabled by -fprofile-use and -fauto-profile.
9701
9702       -fversion-loops-for-strides
9703           If a loop iterates over an array with a variable stride, create
9704           another version of the loop that assumes the stride is always one.
9705           For example:
9706
9707                   for (int i = 0; i < n; ++i)
9708                     x[i * stride] = ...;
9709
9710           becomes:
9711
9712                   if (stride == 1)
9713                     for (int i = 0; i < n; ++i)
9714                       x[i] = ...;
9715                   else
9716                     for (int i = 0; i < n; ++i)
9717                       x[i * stride] = ...;
9718
9719           This is particularly useful for assumed-shape arrays in Fortran
9720           where (for example) it allows better vectorization assuming
9721           contiguous accesses.  This flag is enabled by default at -O3.  It
9722           is also enabled by -fprofile-use and -fauto-profile.
9723
9724       -ffunction-sections
9725       -fdata-sections
9726           Place each function or data item into its own section in the output
9727           file if the target supports arbitrary sections.  The name of the
9728           function or the name of the data item determines the section's name
9729           in the output file.
9730
9731           Use these options on systems where the linker can perform
9732           optimizations to improve locality of reference in the instruction
9733           space.  Most systems using the ELF object format have linkers with
9734           such optimizations.  On AIX, the linker rearranges sections
9735           (CSECTs) based on the call graph.  The performance impact varies.
9736
9737           Together with a linker garbage collection (linker --gc-sections
9738           option) these options may lead to smaller statically-linked
9739           executables (after stripping).
9740
9741           On ELF/DWARF systems these options do not degenerate the quality of
9742           the debug information.  There could be issues with other object
9743           files/debug info formats.
9744
9745           Only use these options when there are significant benefits from
9746           doing so.  When you specify these options, the assembler and linker
9747           create larger object and executable files and are also slower.
9748           These options affect code generation.  They prevent optimizations
9749           by the compiler and assembler using relative locations inside a
9750           translation unit since the locations are unknown until link time.
9751           An example of such an optimization is relaxing calls to short call
9752           instructions.
9753
9754       -fstdarg-opt
9755           Optimize the prologue of variadic argument functions with respect
9756           to usage of those arguments.
9757
9758       -fsection-anchors
9759           Try to reduce the number of symbolic address calculations by using
9760           shared "anchor" symbols to address nearby objects.  This
9761           transformation can help to reduce the number of GOT entries and GOT
9762           accesses on some targets.
9763
9764           For example, the implementation of the following function "foo":
9765
9766                   static int a, b, c;
9767                   int foo (void) { return a + b + c; }
9768
9769           usually calculates the addresses of all three variables, but if you
9770           compile it with -fsection-anchors, it accesses the variables from a
9771           common anchor point instead.  The effect is similar to the
9772           following pseudocode (which isn't valid C):
9773
9774                   int foo (void)
9775                   {
9776                     register int *xr = &x;
9777                     return xr[&a - &x] + xr[&b - &x] + xr[&c - &x];
9778                   }
9779
9780           Not all targets support this option.
9781
9782       --param name=value
9783           In some places, GCC uses various constants to control the amount of
9784           optimization that is done.  For example, GCC does not inline
9785           functions that contain more than a certain number of instructions.
9786           You can control some of these constants on the command line using
9787           the --param option.
9788
9789           The names of specific parameters, and the meaning of the values,
9790           are tied to the internals of the compiler, and are subject to
9791           change without notice in future releases.
9792
9793           In order to get minimal, maximal and default value of a parameter,
9794           one can use --help=param -Q options.
9795
9796           In each case, the value is an integer.  The following choices of
9797           name are recognized for all targets:
9798
9799           predictable-branch-outcome
9800               When branch is predicted to be taken with probability lower
9801               than this threshold (in percent), then it is considered well
9802               predictable.
9803
9804           max-rtl-if-conversion-insns
9805               RTL if-conversion tries to remove conditional branches around a
9806               block and replace them with conditionally executed
9807               instructions.  This parameter gives the maximum number of
9808               instructions in a block which should be considered for if-
9809               conversion.  The compiler will also use other heuristics to
9810               decide whether if-conversion is likely to be profitable.
9811
9812           max-rtl-if-conversion-predictable-cost
9813           max-rtl-if-conversion-unpredictable-cost
9814               RTL if-conversion will try to remove conditional branches
9815               around a block and replace them with conditionally executed
9816               instructions.  These parameters give the maximum permissible
9817               cost for the sequence that would be generated by if-conversion
9818               depending on whether the branch is statically determined to be
9819               predictable or not.  The units for this parameter are the same
9820               as those for the GCC internal seq_cost metric.  The compiler
9821               will try to provide a reasonable default for this parameter
9822               using the BRANCH_COST target macro.
9823
9824           max-crossjump-edges
9825               The maximum number of incoming edges to consider for cross-
9826               jumping.  The algorithm used by -fcrossjumping is O(N^2) in the
9827               number of edges incoming to each block.  Increasing values mean
9828               more aggressive optimization, making the compilation time
9829               increase with probably small improvement in executable size.
9830
9831           min-crossjump-insns
9832               The minimum number of instructions that must be matched at the
9833               end of two blocks before cross-jumping is performed on them.
9834               This value is ignored in the case where all instructions in the
9835               block being cross-jumped from are matched.
9836
9837           max-grow-copy-bb-insns
9838               The maximum code size expansion factor when copying basic
9839               blocks instead of jumping.  The expansion is relative to a jump
9840               instruction.
9841
9842           max-goto-duplication-insns
9843               The maximum number of instructions to duplicate to a block that
9844               jumps to a computed goto.  To avoid O(N^2) behavior in a number
9845               of passes, GCC factors computed gotos early in the compilation
9846               process, and unfactors them as late as possible.  Only computed
9847               jumps at the end of a basic blocks with no more than max-goto-
9848               duplication-insns are unfactored.
9849
9850           max-delay-slot-insn-search
9851               The maximum number of instructions to consider when looking for
9852               an instruction to fill a delay slot.  If more than this
9853               arbitrary number of instructions are searched, the time savings
9854               from filling the delay slot are minimal, so stop searching.
9855               Increasing values mean more aggressive optimization, making the
9856               compilation time increase with probably small improvement in
9857               execution time.
9858
9859           max-delay-slot-live-search
9860               When trying to fill delay slots, the maximum number of
9861               instructions to consider when searching for a block with valid
9862               live register information.  Increasing this arbitrarily chosen
9863               value means more aggressive optimization, increasing the
9864               compilation time.  This parameter should be removed when the
9865               delay slot code is rewritten to maintain the control-flow
9866               graph.
9867
9868           max-gcse-memory
9869               The approximate maximum amount of memory that can be allocated
9870               in order to perform the global common subexpression elimination
9871               optimization.  If more memory than specified is required, the
9872               optimization is not done.
9873
9874           max-gcse-insertion-ratio
9875               If the ratio of expression insertions to deletions is larger
9876               than this value for any expression, then RTL PRE inserts or
9877               removes the expression and thus leaves partially redundant
9878               computations in the instruction stream.
9879
9880           max-pending-list-length
9881               The maximum number of pending dependencies scheduling allows
9882               before flushing the current state and starting over.  Large
9883               functions with few branches or calls can create excessively
9884               large lists which needlessly consume memory and resources.
9885
9886           max-modulo-backtrack-attempts
9887               The maximum number of backtrack attempts the scheduler should
9888               make when modulo scheduling a loop.  Larger values can
9889               exponentially increase compilation time.
9890
9891           max-inline-insns-single
9892               Several parameters control the tree inliner used in GCC.  This
9893               number sets the maximum number of instructions (counted in
9894               GCC's internal representation) in a single function that the
9895               tree inliner considers for inlining.  This only affects
9896               functions declared inline and methods implemented in a class
9897               declaration (C++).
9898
9899           max-inline-insns-auto
9900               When you use -finline-functions (included in -O3), a lot of
9901               functions that would otherwise not be considered for inlining
9902               by the compiler are investigated.  To those functions, a
9903               different (more restrictive) limit compared to functions
9904               declared inline can be applied (--param max-inline-insns-auto).
9905
9906           max-inline-insns-small
9907               This is bound applied to calls which are considered relevant
9908               with -finline-small-functions.
9909
9910           max-inline-insns-size
9911               This is bound applied to calls which are optimized for size.
9912               Small growth may be desirable to anticipate optimization
9913               oppurtunities exposed by inlining.
9914
9915           uninlined-function-insns
9916               Number of instructions accounted by inliner for function
9917               overhead such as function prologue and epilogue.
9918
9919           uninlined-function-time
9920               Extra time accounted by inliner for function overhead such as
9921               time needed to execute function prologue and epilogue
9922
9923           inline-heuristics-hint-percent
9924               The scale (in percents) applied to inline-insns-single,
9925               inline-insns-single-O2, inline-insns-auto when inline
9926               heuristics hints that inlining is very profitable (will enable
9927               later optimizations).
9928
9929           uninlined-thunk-insns
9930           uninlined-thunk-time
9931               Same as --param uninlined-function-insns and --param uninlined-
9932               function-time but applied to function thunks
9933
9934           inline-min-speedup
9935               When estimated performance improvement of caller + callee
9936               runtime exceeds this threshold (in percent), the function can
9937               be inlined regardless of the limit on --param max-inline-insns-
9938               single and --param max-inline-insns-auto.
9939
9940           large-function-insns
9941               The limit specifying really large functions.  For functions
9942               larger than this limit after inlining, inlining is constrained
9943               by --param large-function-growth.  This parameter is useful
9944               primarily to avoid extreme compilation time caused by non-
9945               linear algorithms used by the back end.
9946
9947           large-function-growth
9948               Specifies maximal growth of large function caused by inlining
9949               in percents.  For example, parameter value 100 limits large
9950               function growth to 2.0 times the original size.
9951
9952           large-unit-insns
9953               The limit specifying large translation unit.  Growth caused by
9954               inlining of units larger than this limit is limited by --param
9955               inline-unit-growth.  For small units this might be too tight.
9956               For example, consider a unit consisting of function A that is
9957               inline and B that just calls A three times.  If B is small
9958               relative to A, the growth of unit is 300\% and yet such
9959               inlining is very sane.  For very large units consisting of
9960               small inlineable functions, however, the overall unit growth
9961               limit is needed to avoid exponential explosion of code size.
9962               Thus for smaller units, the size is increased to --param large-
9963               unit-insns before applying --param inline-unit-growth.
9964
9965           inline-unit-growth
9966               Specifies maximal overall growth of the compilation unit caused
9967               by inlining.  For example, parameter value 20 limits unit
9968               growth to 1.2 times the original size. Cold functions (either
9969               marked cold via an attribute or by profile feedback) are not
9970               accounted into the unit size.
9971
9972           ipa-cp-unit-growth
9973               Specifies maximal overall growth of the compilation unit caused
9974               by interprocedural constant propagation.  For example,
9975               parameter value 10 limits unit growth to 1.1 times the original
9976               size.
9977
9978           large-stack-frame
9979               The limit specifying large stack frames.  While inlining the
9980               algorithm is trying to not grow past this limit too much.
9981
9982           large-stack-frame-growth
9983               Specifies maximal growth of large stack frames caused by
9984               inlining in percents.  For example, parameter value 1000 limits
9985               large stack frame growth to 11 times the original size.
9986
9987           max-inline-insns-recursive
9988           max-inline-insns-recursive-auto
9989               Specifies the maximum number of instructions an out-of-line
9990               copy of a self-recursive inline function can grow into by
9991               performing recursive inlining.
9992
9993               --param max-inline-insns-recursive applies to functions
9994               declared inline.  For functions not declared inline, recursive
9995               inlining happens only when -finline-functions (included in -O3)
9996               is enabled; --param max-inline-insns-recursive-auto applies
9997               instead.
9998
9999           max-inline-recursive-depth
10000           max-inline-recursive-depth-auto
10001               Specifies the maximum recursion depth used for recursive
10002               inlining.
10003
10004               --param max-inline-recursive-depth applies to functions
10005               declared inline.  For functions not declared inline, recursive
10006               inlining happens only when -finline-functions (included in -O3)
10007               is enabled; --param max-inline-recursive-depth-auto applies
10008               instead.
10009
10010           min-inline-recursive-probability
10011               Recursive inlining is profitable only for function having deep
10012               recursion in average and can hurt for function having little
10013               recursion depth by increasing the prologue size or complexity
10014               of function body to other optimizers.
10015
10016               When profile feedback is available (see -fprofile-generate) the
10017               actual recursion depth can be guessed from the probability that
10018               function recurses via a given call expression.  This parameter
10019               limits inlining only to call expressions whose probability
10020               exceeds the given threshold (in percents).
10021
10022           early-inlining-insns
10023               Specify growth that the early inliner can make.  In effect it
10024               increases the amount of inlining for code having a large
10025               abstraction penalty.
10026
10027           max-early-inliner-iterations
10028               Limit of iterations of the early inliner.  This basically
10029               bounds the number of nested indirect calls the early inliner
10030               can resolve.  Deeper chains are still handled by late inlining.
10031
10032           comdat-sharing-probability
10033               Probability (in percent) that C++ inline function with comdat
10034               visibility are shared across multiple compilation units.
10035
10036           profile-func-internal-id
10037               A parameter to control whether to use function internal id in
10038               profile database lookup. If the value is 0, the compiler uses
10039               an id that is based on function assembler name and filename,
10040               which makes old profile data more tolerant to source changes
10041               such as function reordering etc.
10042
10043           min-vect-loop-bound
10044               The minimum number of iterations under which loops are not
10045               vectorized when -ftree-vectorize is used.  The number of
10046               iterations after vectorization needs to be greater than the
10047               value specified by this option to allow vectorization.
10048
10049           gcse-cost-distance-ratio
10050               Scaling factor in calculation of maximum distance an expression
10051               can be moved by GCSE optimizations.  This is currently
10052               supported only in the code hoisting pass.  The bigger the
10053               ratio, the more aggressive code hoisting is with simple
10054               expressions, i.e., the expressions that have cost less than
10055               gcse-unrestricted-cost.  Specifying 0 disables hoisting of
10056               simple expressions.
10057
10058           gcse-unrestricted-cost
10059               Cost, roughly measured as the cost of a single typical machine
10060               instruction, at which GCSE optimizations do not constrain the
10061               distance an expression can travel.  This is currently supported
10062               only in the code hoisting pass.  The lesser the cost, the more
10063               aggressive code hoisting is.  Specifying 0 allows all
10064               expressions to travel unrestricted distances.
10065
10066           max-hoist-depth
10067               The depth of search in the dominator tree for expressions to
10068               hoist.  This is used to avoid quadratic behavior in hoisting
10069               algorithm.  The value of 0 does not limit on the search, but
10070               may slow down compilation of huge functions.
10071
10072           max-tail-merge-comparisons
10073               The maximum amount of similar bbs to compare a bb with.  This
10074               is used to avoid quadratic behavior in tree tail merging.
10075
10076           max-tail-merge-iterations
10077               The maximum amount of iterations of the pass over the function.
10078               This is used to limit compilation time in tree tail merging.
10079
10080           store-merging-allow-unaligned
10081               Allow the store merging pass to introduce unaligned stores if
10082               it is legal to do so.
10083
10084           max-stores-to-merge
10085               The maximum number of stores to attempt to merge into wider
10086               stores in the store merging pass.
10087
10088           max-unrolled-insns
10089               The maximum number of instructions that a loop may have to be
10090               unrolled.  If a loop is unrolled, this parameter also
10091               determines how many times the loop code is unrolled.
10092
10093           max-average-unrolled-insns
10094               The maximum number of instructions biased by probabilities of
10095               their execution that a loop may have to be unrolled.  If a loop
10096               is unrolled, this parameter also determines how many times the
10097               loop code is unrolled.
10098
10099           max-unroll-times
10100               The maximum number of unrollings of a single loop.
10101
10102           max-peeled-insns
10103               The maximum number of instructions that a loop may have to be
10104               peeled.  If a loop is peeled, this parameter also determines
10105               how many times the loop code is peeled.
10106
10107           max-peel-times
10108               The maximum number of peelings of a single loop.
10109
10110           max-peel-branches
10111               The maximum number of branches on the hot path through the
10112               peeled sequence.
10113
10114           max-completely-peeled-insns
10115               The maximum number of insns of a completely peeled loop.
10116
10117           max-completely-peel-times
10118               The maximum number of iterations of a loop to be suitable for
10119               complete peeling.
10120
10121           max-completely-peel-loop-nest-depth
10122               The maximum depth of a loop nest suitable for complete peeling.
10123
10124           max-unswitch-insns
10125               The maximum number of insns of an unswitched loop.
10126
10127           max-unswitch-level
10128               The maximum number of branches unswitched in a single loop.
10129
10130           lim-expensive
10131               The minimum cost of an expensive expression in the loop
10132               invariant motion.
10133
10134           min-loop-cond-split-prob
10135               When FDO profile information is available, min-loop-cond-split-
10136               prob specifies minimum threshold for probability of semi-
10137               invariant condition statement to trigger loop split.
10138
10139           iv-consider-all-candidates-bound
10140               Bound on number of candidates for induction variables, below
10141               which all candidates are considered for each use in induction
10142               variable optimizations.  If there are more candidates than
10143               this, only the most relevant ones are considered to avoid
10144               quadratic time complexity.
10145
10146           iv-max-considered-uses
10147               The induction variable optimizations give up on loops that
10148               contain more induction variable uses.
10149
10150           iv-always-prune-cand-set-bound
10151               If the number of candidates in the set is smaller than this
10152               value, always try to remove unnecessary ivs from the set when
10153               adding a new one.
10154
10155           avg-loop-niter
10156               Average number of iterations of a loop.
10157
10158           dse-max-object-size
10159               Maximum size (in bytes) of objects tracked bytewise by dead
10160               store elimination.  Larger values may result in larger
10161               compilation times.
10162
10163           dse-max-alias-queries-per-store
10164               Maximum number of queries into the alias oracle per store.
10165               Larger values result in larger compilation times and may result
10166               in more removed dead stores.
10167
10168           scev-max-expr-size
10169               Bound on size of expressions used in the scalar evolutions
10170               analyzer.  Large expressions slow the analyzer.
10171
10172           scev-max-expr-complexity
10173               Bound on the complexity of the expressions in the scalar
10174               evolutions analyzer.  Complex expressions slow the analyzer.
10175
10176           max-tree-if-conversion-phi-args
10177               Maximum number of arguments in a PHI supported by TREE if
10178               conversion unless the loop is marked with simd pragma.
10179
10180           vect-max-version-for-alignment-checks
10181               The maximum number of run-time checks that can be performed
10182               when doing loop versioning for alignment in the vectorizer.
10183
10184           vect-max-version-for-alias-checks
10185               The maximum number of run-time checks that can be performed
10186               when doing loop versioning for alias in the vectorizer.
10187
10188           vect-max-peeling-for-alignment
10189               The maximum number of loop peels to enhance access alignment
10190               for vectorizer. Value -1 means no limit.
10191
10192           max-iterations-to-track
10193               The maximum number of iterations of a loop the brute-force
10194               algorithm for analysis of the number of iterations of the loop
10195               tries to evaluate.
10196
10197           hot-bb-count-fraction
10198               The denominator n of fraction 1/n of the maximal execution
10199               count of a basic block in the entire program that a basic block
10200               needs to at least have in order to be considered hot.  The
10201               default is 10000, which means that a basic block is considered
10202               hot if its execution count is greater than 1/10000 of the
10203               maximal execution count.  0 means that it is never considered
10204               hot.  Used in non-LTO mode.
10205
10206           hot-bb-count-ws-permille
10207               The number of most executed permilles, ranging from 0 to 1000,
10208               of the profiled execution of the entire program to which the
10209               execution count of a basic block must be part of in order to be
10210               considered hot.  The default is 990, which means that a basic
10211               block is considered hot if its execution count contributes to
10212               the upper 990 permilles, or 99.0%, of the profiled execution of
10213               the entire program.  0 means that it is never considered hot.
10214               Used in LTO mode.
10215
10216           hot-bb-frequency-fraction
10217               The denominator n of fraction 1/n of the execution frequency of
10218               the entry block of a function that a basic block of this
10219               function needs to at least have in order to be considered hot.
10220               The default is 1000, which means that a basic block is
10221               considered hot in a function if it is executed more frequently
10222               than 1/1000 of the frequency of the entry block of the
10223               function.  0 means that it is never considered hot.
10224
10225           unlikely-bb-count-fraction
10226               The denominator n of fraction 1/n of the number of profiled
10227               runs of the entire program below which the execution count of a
10228               basic block must be in order for the basic block to be
10229               considered unlikely executed.  The default is 20, which means
10230               that a basic block is considered unlikely executed if it is
10231               executed in fewer than 1/20, or 5%, of the runs of the program.
10232               0 means that it is always considered unlikely executed.
10233
10234           max-predicted-iterations
10235               The maximum number of loop iterations we predict statically.
10236               This is useful in cases where a function contains a single loop
10237               with known bound and another loop with unknown bound.  The
10238               known number of iterations is predicted correctly, while the
10239               unknown number of iterations average to roughly 10.  This means
10240               that the loop without bounds appears artificially cold relative
10241               to the other one.
10242
10243           builtin-expect-probability
10244               Control the probability of the expression having the specified
10245               value. This parameter takes a percentage (i.e. 0 ... 100) as
10246               input.
10247
10248           builtin-string-cmp-inline-length
10249               The maximum length of a constant string for a builtin string
10250               cmp call eligible for inlining.
10251
10252           align-threshold
10253               Select fraction of the maximal frequency of executions of a
10254               basic block in a function to align the basic block.
10255
10256           align-loop-iterations
10257               A loop expected to iterate at least the selected number of
10258               iterations is aligned.
10259
10260           tracer-dynamic-coverage
10261           tracer-dynamic-coverage-feedback
10262               This value is used to limit superblock formation once the given
10263               percentage of executed instructions is covered.  This limits
10264               unnecessary code size expansion.
10265
10266               The tracer-dynamic-coverage-feedback parameter is used only
10267               when profile feedback is available.  The real profiles (as
10268               opposed to statically estimated ones) are much less balanced
10269               allowing the threshold to be larger value.
10270
10271           tracer-max-code-growth
10272               Stop tail duplication once code growth has reached given
10273               percentage.  This is a rather artificial limit, as most of the
10274               duplicates are eliminated later in cross jumping, so it may be
10275               set to much higher values than is the desired code growth.
10276
10277           tracer-min-branch-ratio
10278               Stop reverse growth when the reverse probability of best edge
10279               is less than this threshold (in percent).
10280
10281           tracer-min-branch-probability
10282           tracer-min-branch-probability-feedback
10283               Stop forward growth if the best edge has probability lower than
10284               this threshold.
10285
10286               Similarly to tracer-dynamic-coverage two parameters are
10287               provided.  tracer-min-branch-probability-feedback is used for
10288               compilation with profile feedback and tracer-min-branch-
10289               probability compilation without.  The value for compilation
10290               with profile feedback needs to be more conservative (higher) in
10291               order to make tracer effective.
10292
10293           stack-clash-protection-guard-size
10294               Specify the size of the operating system provided stack guard
10295               as 2 raised to num bytes.  Higher values may reduce the number
10296               of explicit probes, but a value larger than the operating
10297               system provided guard will leave code vulnerable to stack clash
10298               style attacks.
10299
10300           stack-clash-protection-probe-interval
10301               Stack clash protection involves probing stack space as it is
10302               allocated.  This param controls the maximum distance between
10303               probes into the stack as 2 raised to num bytes.  Higher values
10304               may reduce the number of explicit probes, but a value larger
10305               than the operating system provided guard will leave code
10306               vulnerable to stack clash style attacks.
10307
10308           max-cse-path-length
10309               The maximum number of basic blocks on path that CSE considers.
10310
10311           max-cse-insns
10312               The maximum number of instructions CSE processes before
10313               flushing.
10314
10315           ggc-min-expand
10316               GCC uses a garbage collector to manage its own memory
10317               allocation.  This parameter specifies the minimum percentage by
10318               which the garbage collector's heap should be allowed to expand
10319               between collections.  Tuning this may improve compilation
10320               speed; it has no effect on code generation.
10321
10322               The default is 30% + 70% * (RAM/1GB) with an upper bound of
10323               100% when RAM >= 1GB.  If "getrlimit" is available, the notion
10324               of "RAM" is the smallest of actual RAM and "RLIMIT_DATA" or
10325               "RLIMIT_AS".  If GCC is not able to calculate RAM on a
10326               particular platform, the lower bound of 30% is used.  Setting
10327               this parameter and ggc-min-heapsize to zero causes a full
10328               collection to occur at every opportunity.  This is extremely
10329               slow, but can be useful for debugging.
10330
10331           ggc-min-heapsize
10332               Minimum size of the garbage collector's heap before it begins
10333               bothering to collect garbage.  The first collection occurs
10334               after the heap expands by ggc-min-expand% beyond ggc-min-
10335               heapsize.  Again, tuning this may improve compilation speed,
10336               and has no effect on code generation.
10337
10338               The default is the smaller of RAM/8, RLIMIT_RSS, or a limit
10339               that tries to ensure that RLIMIT_DATA or RLIMIT_AS are not
10340               exceeded, but with a lower bound of 4096 (four megabytes) and
10341               an upper bound of 131072 (128 megabytes).  If GCC is not able
10342               to calculate RAM on a particular platform, the lower bound is
10343               used.  Setting this parameter very large effectively disables
10344               garbage collection.  Setting this parameter and ggc-min-expand
10345               to zero causes a full collection to occur at every opportunity.
10346
10347           max-reload-search-insns
10348               The maximum number of instruction reload should look backward
10349               for equivalent register.  Increasing values mean more
10350               aggressive optimization, making the compilation time increase
10351               with probably slightly better performance.
10352
10353           max-cselib-memory-locations
10354               The maximum number of memory locations cselib should take into
10355               account.  Increasing values mean more aggressive optimization,
10356               making the compilation time increase with probably slightly
10357               better performance.
10358
10359           max-sched-ready-insns
10360               The maximum number of instructions ready to be issued the
10361               scheduler should consider at any given time during the first
10362               scheduling pass.  Increasing values mean more thorough
10363               searches, making the compilation time increase with probably
10364               little benefit.
10365
10366           max-sched-region-blocks
10367               The maximum number of blocks in a region to be considered for
10368               interblock scheduling.
10369
10370           max-pipeline-region-blocks
10371               The maximum number of blocks in a region to be considered for
10372               pipelining in the selective scheduler.
10373
10374           max-sched-region-insns
10375               The maximum number of insns in a region to be considered for
10376               interblock scheduling.
10377
10378           max-pipeline-region-insns
10379               The maximum number of insns in a region to be considered for
10380               pipelining in the selective scheduler.
10381
10382           min-spec-prob
10383               The minimum probability (in percents) of reaching a source
10384               block for interblock speculative scheduling.
10385
10386           max-sched-extend-regions-iters
10387               The maximum number of iterations through CFG to extend regions.
10388               A value of 0 disables region extensions.
10389
10390           max-sched-insn-conflict-delay
10391               The maximum conflict delay for an insn to be considered for
10392               speculative motion.
10393
10394           sched-spec-prob-cutoff
10395               The minimal probability of speculation success (in percents),
10396               so that speculative insns are scheduled.
10397
10398           sched-state-edge-prob-cutoff
10399               The minimum probability an edge must have for the scheduler to
10400               save its state across it.
10401
10402           sched-mem-true-dep-cost
10403               Minimal distance (in CPU cycles) between store and load
10404               targeting same memory locations.
10405
10406           selsched-max-lookahead
10407               The maximum size of the lookahead window of selective
10408               scheduling.  It is a depth of search for available
10409               instructions.
10410
10411           selsched-max-sched-times
10412               The maximum number of times that an instruction is scheduled
10413               during selective scheduling.  This is the limit on the number
10414               of iterations through which the instruction may be pipelined.
10415
10416           selsched-insns-to-rename
10417               The maximum number of best instructions in the ready list that
10418               are considered for renaming in the selective scheduler.
10419
10420           sms-min-sc
10421               The minimum value of stage count that swing modulo scheduler
10422               generates.
10423
10424           max-last-value-rtl
10425               The maximum size measured as number of RTLs that can be
10426               recorded in an expression in combiner for a pseudo register as
10427               last known value of that register.
10428
10429           max-combine-insns
10430               The maximum number of instructions the RTL combiner tries to
10431               combine.
10432
10433           integer-share-limit
10434               Small integer constants can use a shared data structure,
10435               reducing the compiler's memory usage and increasing its speed.
10436               This sets the maximum value of a shared integer constant.
10437
10438           ssp-buffer-size
10439               The minimum size of buffers (i.e. arrays) that receive stack
10440               smashing protection when -fstack-protection is used.
10441
10442           min-size-for-stack-sharing
10443               The minimum size of variables taking part in stack slot sharing
10444               when not optimizing.
10445
10446           max-jump-thread-duplication-stmts
10447               Maximum number of statements allowed in a block that needs to
10448               be duplicated when threading jumps.
10449
10450           max-fields-for-field-sensitive
10451               Maximum number of fields in a structure treated in a field
10452               sensitive manner during pointer analysis.
10453
10454           prefetch-latency
10455               Estimate on average number of instructions that are executed
10456               before prefetch finishes.  The distance prefetched ahead is
10457               proportional to this constant.  Increasing this number may also
10458               lead to less streams being prefetched (see simultaneous-
10459               prefetches).
10460
10461           simultaneous-prefetches
10462               Maximum number of prefetches that can run at the same time.
10463
10464           l1-cache-line-size
10465               The size of cache line in L1 data cache, in bytes.
10466
10467           l1-cache-size
10468               The size of L1 data cache, in kilobytes.
10469
10470           l2-cache-size
10471               The size of L2 data cache, in kilobytes.
10472
10473           prefetch-dynamic-strides
10474               Whether the loop array prefetch pass should issue software
10475               prefetch hints for strides that are non-constant.  In some
10476               cases this may be beneficial, though the fact the stride is
10477               non-constant may make it hard to predict when there is clear
10478               benefit to issuing these hints.
10479
10480               Set to 1 if the prefetch hints should be issued for non-
10481               constant strides.  Set to 0 if prefetch hints should be issued
10482               only for strides that are known to be constant and below
10483               prefetch-minimum-stride.
10484
10485           prefetch-minimum-stride
10486               Minimum constant stride, in bytes, to start using prefetch
10487               hints for.  If the stride is less than this threshold, prefetch
10488               hints will not be issued.
10489
10490               This setting is useful for processors that have hardware
10491               prefetchers, in which case there may be conflicts between the
10492               hardware prefetchers and the software prefetchers.  If the
10493               hardware prefetchers have a maximum stride they can handle, it
10494               should be used here to improve the use of software prefetchers.
10495
10496               A value of -1 means we don't have a threshold and therefore
10497               prefetch hints can be issued for any constant stride.
10498
10499               This setting is only useful for strides that are known and
10500               constant.
10501
10502           loop-interchange-max-num-stmts
10503               The maximum number of stmts in a loop to be interchanged.
10504
10505           loop-interchange-stride-ratio
10506               The minimum ratio between stride of two loops for interchange
10507               to be profitable.
10508
10509           min-insn-to-prefetch-ratio
10510               The minimum ratio between the number of instructions and the
10511               number of prefetches to enable prefetching in a loop.
10512
10513           prefetch-min-insn-to-mem-ratio
10514               The minimum ratio between the number of instructions and the
10515               number of memory references to enable prefetching in a loop.
10516
10517           use-canonical-types
10518               Whether the compiler should use the "canonical" type system.
10519               Should always be 1, which uses a more efficient internal
10520               mechanism for comparing types in C++ and Objective-C++.
10521               However, if bugs in the canonical type system are causing
10522               compilation failures, set this value to 0 to disable canonical
10523               types.
10524
10525           switch-conversion-max-branch-ratio
10526               Switch initialization conversion refuses to create arrays that
10527               are bigger than switch-conversion-max-branch-ratio times the
10528               number of branches in the switch.
10529
10530           max-partial-antic-length
10531               Maximum length of the partial antic set computed during the
10532               tree partial redundancy elimination optimization (-ftree-pre)
10533               when optimizing at -O3 and above.  For some sorts of source
10534               code the enhanced partial redundancy elimination optimization
10535               can run away, consuming all of the memory available on the host
10536               machine.  This parameter sets a limit on the length of the sets
10537               that are computed, which prevents the runaway behavior.
10538               Setting a value of 0 for this parameter allows an unlimited set
10539               length.
10540
10541           rpo-vn-max-loop-depth
10542               Maximum loop depth that is value-numbered optimistically.  When
10543               the limit hits the innermost rpo-vn-max-loop-depth loops and
10544               the outermost loop in the loop nest are value-numbered
10545               optimistically and the remaining ones not.
10546
10547           sccvn-max-alias-queries-per-access
10548               Maximum number of alias-oracle queries we perform when looking
10549               for redundancies for loads and stores.  If this limit is hit
10550               the search is aborted and the load or store is not considered
10551               redundant.  The number of queries is algorithmically limited to
10552               the number of stores on all paths from the load to the function
10553               entry.
10554
10555           ira-max-loops-num
10556               IRA uses regional register allocation by default.  If a
10557               function contains more loops than the number given by this
10558               parameter, only at most the given number of the most
10559               frequently-executed loops form regions for regional register
10560               allocation.
10561
10562           ira-max-conflict-table-size
10563               Although IRA uses a sophisticated algorithm to compress the
10564               conflict table, the table can still require excessive amounts
10565               of memory for huge functions.  If the conflict table for a
10566               function could be more than the size in MB given by this
10567               parameter, the register allocator instead uses a faster,
10568               simpler, and lower-quality algorithm that does not require
10569               building a pseudo-register conflict table.
10570
10571           ira-loop-reserved-regs
10572               IRA can be used to evaluate more accurate register pressure in
10573               loops for decisions to move loop invariants (see -O3).  The
10574               number of available registers reserved for some other purposes
10575               is given by this parameter.  Default of the parameter is the
10576               best found from numerous experiments.
10577
10578           lra-inheritance-ebb-probability-cutoff
10579               LRA tries to reuse values reloaded in registers in subsequent
10580               insns.  This optimization is called inheritance.  EBB is used
10581               as a region to do this optimization.  The parameter defines a
10582               minimal fall-through edge probability in percentage used to add
10583               BB to inheritance EBB in LRA.  The default value was chosen
10584               from numerous runs of SPEC2000 on x86-64.
10585
10586           loop-invariant-max-bbs-in-loop
10587               Loop invariant motion can be very expensive, both in
10588               compilation time and in amount of needed compile-time memory,
10589               with very large loops.  Loops with more basic blocks than this
10590               parameter won't have loop invariant motion optimization
10591               performed on them.
10592
10593           loop-max-datarefs-for-datadeps
10594               Building data dependencies is expensive for very large loops.
10595               This parameter limits the number of data references in loops
10596               that are considered for data dependence analysis.  These large
10597               loops are no handled by the optimizations using loop data
10598               dependencies.
10599
10600           max-vartrack-size
10601               Sets a maximum number of hash table slots to use during
10602               variable tracking dataflow analysis of any function.  If this
10603               limit is exceeded with variable tracking at assignments
10604               enabled, analysis for that function is retried without it,
10605               after removing all debug insns from the function.  If the limit
10606               is exceeded even without debug insns, var tracking analysis is
10607               completely disabled for the function.  Setting the parameter to
10608               zero makes it unlimited.
10609
10610           max-vartrack-expr-depth
10611               Sets a maximum number of recursion levels when attempting to
10612               map variable names or debug temporaries to value expressions.
10613               This trades compilation time for more complete debug
10614               information.  If this is set too low, value expressions that
10615               are available and could be represented in debug information may
10616               end up not being used; setting this higher may enable the
10617               compiler to find more complex debug expressions, but compile
10618               time and memory use may grow.
10619
10620           max-debug-marker-count
10621               Sets a threshold on the number of debug markers (e.g. begin
10622               stmt markers) to avoid complexity explosion at inlining or
10623               expanding to RTL.  If a function has more such gimple stmts
10624               than the set limit, such stmts will be dropped from the inlined
10625               copy of a function, and from its RTL expansion.
10626
10627           min-nondebug-insn-uid
10628               Use uids starting at this parameter for nondebug insns.  The
10629               range below the parameter is reserved exclusively for debug
10630               insns created by -fvar-tracking-assignments, but debug insns
10631               may get (non-overlapping) uids above it if the reserved range
10632               is exhausted.
10633
10634           ipa-sra-ptr-growth-factor
10635               IPA-SRA replaces a pointer to an aggregate with one or more new
10636               parameters only when their cumulative size is less or equal to
10637               ipa-sra-ptr-growth-factor times the size of the original
10638               pointer parameter.
10639
10640           ipa-sra-max-replacements
10641               Maximum pieces of an aggregate that IPA-SRA tracks.  As a
10642               consequence, it is also the maximum number of replacements of a
10643               formal parameter.
10644
10645           sra-max-scalarization-size-Ospeed
10646           sra-max-scalarization-size-Osize
10647               The two Scalar Reduction of Aggregates passes (SRA and IPA-SRA)
10648               aim to replace scalar parts of aggregates with uses of
10649               independent scalar variables.  These parameters control the
10650               maximum size, in storage units, of aggregate which is
10651               considered for replacement when compiling for speed (sra-max-
10652               scalarization-size-Ospeed) or size (sra-max-scalarization-size-
10653               Osize) respectively.
10654
10655           sra-max-propagations
10656               The maximum number of artificial accesses that Scalar
10657               Replacement of Aggregates (SRA) will track, per one local
10658               variable, in order to facilitate copy propagation.
10659
10660           tm-max-aggregate-size
10661               When making copies of thread-local variables in a transaction,
10662               this parameter specifies the size in bytes after which
10663               variables are saved with the logging functions as opposed to
10664               save/restore code sequence pairs.  This option only applies
10665               when using -fgnu-tm.
10666
10667           graphite-max-nb-scop-params
10668               To avoid exponential effects in the Graphite loop transforms,
10669               the number of parameters in a Static Control Part (SCoP) is
10670               bounded.  A value of zero can be used to lift the bound.  A
10671               variable whose value is unknown at compilation time and defined
10672               outside a SCoP is a parameter of the SCoP.
10673
10674           loop-block-tile-size
10675               Loop blocking or strip mining transforms, enabled with
10676               -floop-block or -floop-strip-mine, strip mine each loop in the
10677               loop nest by a given number of iterations.  The strip length
10678               can be changed using the loop-block-tile-size parameter.
10679
10680           ipa-cp-value-list-size
10681               IPA-CP attempts to track all possible values and types passed
10682               to a function's parameter in order to propagate them and
10683               perform devirtualization.  ipa-cp-value-list-size is the
10684               maximum number of values and types it stores per one formal
10685               parameter of a function.
10686
10687           ipa-cp-eval-threshold
10688               IPA-CP calculates its own score of cloning profitability
10689               heuristics and performs those cloning opportunities with scores
10690               that exceed ipa-cp-eval-threshold.
10691
10692           ipa-cp-max-recursive-depth
10693               Maximum depth of recursive cloning for self-recursive function.
10694
10695           ipa-cp-min-recursive-probability
10696               Recursive cloning only when the probability of call being
10697               executed exceeds the parameter.
10698
10699           ipa-cp-recursion-penalty
10700               Percentage penalty the recursive functions will receive when
10701               they are evaluated for cloning.
10702
10703           ipa-cp-single-call-penalty
10704               Percentage penalty functions containing a single call to
10705               another function will receive when they are evaluated for
10706               cloning.
10707
10708           ipa-max-agg-items
10709               IPA-CP is also capable to propagate a number of scalar values
10710               passed in an aggregate. ipa-max-agg-items controls the maximum
10711               number of such values per one parameter.
10712
10713           ipa-cp-loop-hint-bonus
10714               When IPA-CP determines that a cloning candidate would make the
10715               number of iterations of a loop known, it adds a bonus of ipa-
10716               cp-loop-hint-bonus to the profitability score of the candidate.
10717
10718           ipa-max-aa-steps
10719               During its analysis of function bodies, IPA-CP employs alias
10720               analysis in order to track values pointed to by function
10721               parameters.  In order not spend too much time analyzing huge
10722               functions, it gives up and consider all memory clobbered after
10723               examining ipa-max-aa-steps statements modifying memory.
10724
10725           ipa-max-switch-predicate-bounds
10726               Maximal number of boundary endpoints of case ranges of switch
10727               statement.  For switch exceeding this limit, IPA-CP will not
10728               construct cloning cost predicate, which is used to estimate
10729               cloning benefit, for default case of the switch statement.
10730
10731           ipa-max-param-expr-ops
10732               IPA-CP will analyze conditional statement that references some
10733               function parameter to estimate benefit for cloning upon certain
10734               constant value.  But if number of operations in a parameter
10735               expression exceeds ipa-max-param-expr-ops, the expression is
10736               treated as complicated one, and is not handled by IPA analysis.
10737
10738           lto-partitions
10739               Specify desired number of partitions produced during WHOPR
10740               compilation.  The number of partitions should exceed the number
10741               of CPUs used for compilation.
10742
10743           lto-min-partition
10744               Size of minimal partition for WHOPR (in estimated
10745               instructions).  This prevents expenses of splitting very small
10746               programs into too many partitions.
10747
10748           lto-max-partition
10749               Size of max partition for WHOPR (in estimated instructions).
10750               to provide an upper bound for individual size of partition.
10751               Meant to be used only with balanced partitioning.
10752
10753           lto-max-streaming-parallelism
10754               Maximal number of parallel processes used for LTO streaming.
10755
10756           cxx-max-namespaces-for-diagnostic-help
10757               The maximum number of namespaces to consult for suggestions
10758               when C++ name lookup fails for an identifier.
10759
10760           sink-frequency-threshold
10761               The maximum relative execution frequency (in percents) of the
10762               target block relative to a statement's original block to allow
10763               statement sinking of a statement.  Larger numbers result in
10764               more aggressive statement sinking.  A small positive adjustment
10765               is applied for statements with memory operands as those are
10766               even more profitable so sink.
10767
10768           max-stores-to-sink
10769               The maximum number of conditional store pairs that can be sunk.
10770               Set to 0 if either vectorization (-ftree-vectorize) or if-
10771               conversion (-ftree-loop-if-convert) is disabled.
10772
10773           case-values-threshold
10774               The smallest number of different values for which it is best to
10775               use a jump-table instead of a tree of conditional branches.  If
10776               the value is 0, use the default for the machine.
10777
10778           jump-table-max-growth-ratio-for-size
10779               The maximum code size growth ratio when expanding into a jump
10780               table (in percent).  The parameter is used when optimizing for
10781               size.
10782
10783           jump-table-max-growth-ratio-for-speed
10784               The maximum code size growth ratio when expanding into a jump
10785               table (in percent).  The parameter is used when optimizing for
10786               speed.
10787
10788           tree-reassoc-width
10789               Set the maximum number of instructions executed in parallel in
10790               reassociated tree. This parameter overrides target dependent
10791               heuristics used by default if has non zero value.
10792
10793           sched-pressure-algorithm
10794               Choose between the two available implementations of
10795               -fsched-pressure.  Algorithm 1 is the original implementation
10796               and is the more likely to prevent instructions from being
10797               reordered.  Algorithm 2 was designed to be a compromise between
10798               the relatively conservative approach taken by algorithm 1 and
10799               the rather aggressive approach taken by the default scheduler.
10800               It relies more heavily on having a regular register file and
10801               accurate register pressure classes.  See haifa-sched.c in the
10802               GCC sources for more details.
10803
10804               The default choice depends on the target.
10805
10806           max-slsr-cand-scan
10807               Set the maximum number of existing candidates that are
10808               considered when seeking a basis for a new straight-line
10809               strength reduction candidate.
10810
10811           asan-globals
10812               Enable buffer overflow detection for global objects.  This kind
10813               of protection is enabled by default if you are using
10814               -fsanitize=address option.  To disable global objects
10815               protection use --param asan-globals=0.
10816
10817           asan-stack
10818               Enable buffer overflow detection for stack objects.  This kind
10819               of protection is enabled by default when using
10820               -fsanitize=address.  To disable stack protection use --param
10821               asan-stack=0 option.
10822
10823           asan-instrument-reads
10824               Enable buffer overflow detection for memory reads.  This kind
10825               of protection is enabled by default when using
10826               -fsanitize=address.  To disable memory reads protection use
10827               --param asan-instrument-reads=0.
10828
10829           asan-instrument-writes
10830               Enable buffer overflow detection for memory writes.  This kind
10831               of protection is enabled by default when using
10832               -fsanitize=address.  To disable memory writes protection use
10833               --param asan-instrument-writes=0 option.
10834
10835           asan-memintrin
10836               Enable detection for built-in functions.  This kind of
10837               protection is enabled by default when using -fsanitize=address.
10838               To disable built-in functions protection use --param
10839               asan-memintrin=0.
10840
10841           asan-use-after-return
10842               Enable detection of use-after-return.  This kind of protection
10843               is enabled by default when using the -fsanitize=address option.
10844               To disable it use --param asan-use-after-return=0.
10845
10846               Note: By default the check is disabled at run time.  To enable
10847               it, add "detect_stack_use_after_return=1" to the environment
10848               variable ASAN_OPTIONS.
10849
10850           asan-instrumentation-with-call-threshold
10851               If number of memory accesses in function being instrumented is
10852               greater or equal to this number, use callbacks instead of
10853               inline checks.  E.g. to disable inline code use --param
10854               asan-instrumentation-with-call-threshold=0.
10855
10856           use-after-scope-direct-emission-threshold
10857               If the size of a local variable in bytes is smaller or equal to
10858               this number, directly poison (or unpoison) shadow memory
10859               instead of using run-time callbacks.
10860
10861           max-fsm-thread-path-insns
10862               Maximum number of instructions to copy when duplicating blocks
10863               on a finite state automaton jump thread path.
10864
10865           max-fsm-thread-length
10866               Maximum number of basic blocks on a finite state automaton jump
10867               thread path.
10868
10869           max-fsm-thread-paths
10870               Maximum number of new jump thread paths to create for a finite
10871               state automaton.
10872
10873           parloops-chunk-size
10874               Chunk size of omp schedule for loops parallelized by parloops.
10875
10876           parloops-schedule
10877               Schedule type of omp schedule for loops parallelized by
10878               parloops (static, dynamic, guided, auto, runtime).
10879
10880           parloops-min-per-thread
10881               The minimum number of iterations per thread of an innermost
10882               parallelized loop for which the parallelized variant is
10883               preferred over the single threaded one.  Note that for a
10884               parallelized loop nest the minimum number of iterations of the
10885               outermost loop per thread is two.
10886
10887           max-ssa-name-query-depth
10888               Maximum depth of recursion when querying properties of SSA
10889               names in things like fold routines.  One level of recursion
10890               corresponds to following a use-def chain.
10891
10892           hsa-gen-debug-stores
10893               Enable emission of special debug stores within HSA kernels
10894               which are then read and reported by libgomp plugin.  Generation
10895               of these stores is disabled by default, use --param
10896               hsa-gen-debug-stores=1 to enable it.
10897
10898           max-speculative-devirt-maydefs
10899               The maximum number of may-defs we analyze when looking for a
10900               must-def specifying the dynamic type of an object that invokes
10901               a virtual call we may be able to devirtualize speculatively.
10902
10903           max-vrp-switch-assertions
10904               The maximum number of assertions to add along the default edge
10905               of a switch statement during VRP.
10906
10907           unroll-jam-min-percent
10908               The minimum percentage of memory references that must be
10909               optimized away for the unroll-and-jam transformation to be
10910               considered profitable.
10911
10912           unroll-jam-max-unroll
10913               The maximum number of times the outer loop should be unrolled
10914               by the unroll-and-jam transformation.
10915
10916           max-rtl-if-conversion-unpredictable-cost
10917               Maximum permissible cost for the sequence that would be
10918               generated by the RTL if-conversion pass for a branch that is
10919               considered unpredictable.
10920
10921           max-variable-expansions-in-unroller
10922               If -fvariable-expansion-in-unroller is used, the maximum number
10923               of times that an individual variable will be expanded during
10924               loop unrolling.
10925
10926           tracer-min-branch-probability-feedback
10927               Stop forward growth if the probability of best edge is less
10928               than this threshold (in percent). Used when profile feedback is
10929               available.
10930
10931           partial-inlining-entry-probability
10932               Maximum probability of the entry BB of split region (in percent
10933               relative to entry BB of the function) to make partial inlining
10934               happen.
10935
10936           max-tracked-strlens
10937               Maximum number of strings for which strlen optimization pass
10938               will track string lengths.
10939
10940           gcse-after-reload-partial-fraction
10941               The threshold ratio for performing partial redundancy
10942               elimination after reload.
10943
10944           gcse-after-reload-critical-fraction
10945               The threshold ratio of critical edges execution count that
10946               permit performing redundancy elimination after reload.
10947
10948           max-loop-header-insns
10949               The maximum number of insns in loop header duplicated by the
10950               copy loop headers pass.
10951
10952           vect-epilogues-nomask
10953               Enable loop epilogue vectorization using smaller vector size.
10954
10955           slp-max-insns-in-bb
10956               Maximum number of instructions in basic block to be considered
10957               for SLP vectorization.
10958
10959           avoid-fma-max-bits
10960               Maximum number of bits for which we avoid creating FMAs.
10961
10962           sms-loop-average-count-threshold
10963               A threshold on the average loop count considered by the swing
10964               modulo scheduler.
10965
10966           sms-dfa-history
10967               The number of cycles the swing modulo scheduler considers when
10968               checking conflicts using DFA.
10969
10970           max-inline-insns-recursive-auto
10971               The maximum number of instructions non-inline function can grow
10972               to via recursive inlining.
10973
10974           graphite-allow-codegen-errors
10975               Whether codegen errors should be ICEs when -fchecking.
10976
10977           sms-max-ii-factor
10978               A factor for tuning the upper bound that swing modulo scheduler
10979               uses for scheduling a loop.
10980
10981           lra-max-considered-reload-pseudos
10982               The max number of reload pseudos which are considered during
10983               spilling a non-reload pseudo.
10984
10985           max-pow-sqrt-depth
10986               Maximum depth of sqrt chains to use when synthesizing
10987               exponentiation by a real constant.
10988
10989           max-dse-active-local-stores
10990               Maximum number of active local stores in RTL dead store
10991               elimination.
10992
10993           asan-instrument-allocas
10994               Enable asan allocas/VLAs protection.
10995
10996           max-iterations-computation-cost
10997               Bound on the cost of an expression to compute the number of
10998               iterations.
10999
11000           max-isl-operations
11001               Maximum number of isl operations, 0 means unlimited.
11002
11003           graphite-max-arrays-per-scop
11004               Maximum number of arrays per scop.
11005
11006           max-vartrack-reverse-op-size
11007               Max. size of loc list for which reverse ops should be added.
11008
11009           tracer-dynamic-coverage-feedback
11010               The percentage of function, weighted by execution frequency,
11011               that must be covered by trace formation.  Used when profile
11012               feedback is available.
11013
11014           max-inline-recursive-depth-auto
11015               The maximum depth of recursive inlining for non-inline
11016               functions.
11017
11018           fsm-scale-path-stmts
11019               Scale factor to apply to the number of statements in a
11020               threading path when comparing to the number of (scaled) blocks.
11021
11022           fsm-maximum-phi-arguments
11023               Maximum number of arguments a PHI may have before the FSM
11024               threader will not try to thread through its block.
11025
11026           uninit-control-dep-attempts
11027               Maximum number of nested calls to search for control
11028               dependencies during uninitialized variable analysis.
11029
11030           sra-max-scalarization-size-Osize
11031               Maximum size, in storage units, of an aggregate which should be
11032               considered for scalarization when compiling for size.
11033
11034           fsm-scale-path-blocks
11035               Scale factor to apply to the number of blocks in a threading
11036               path when comparing to the number of (scaled) statements.
11037
11038           sched-autopref-queue-depth
11039               Hardware autoprefetcher scheduler model control flag.  Number
11040               of lookahead cycles the model looks into; at ' ' only enable
11041               instruction sorting heuristic.
11042
11043           loop-versioning-max-inner-insns
11044               The maximum number of instructions that an inner loop can have
11045               before the loop versioning pass considers it too big to copy.
11046
11047           loop-versioning-max-outer-insns
11048               The maximum number of instructions that an outer loop can have
11049               before the loop versioning pass considers it too big to copy,
11050               discounting any instructions in inner loops that directly
11051               benefit from versioning.
11052
11053           ssa-name-def-chain-limit
11054               The maximum number of SSA_NAME assignments to follow in
11055               determining a property of a variable such as its value.  This
11056               limits the number of iterations or recursive calls GCC performs
11057               when optimizing certain statements or when determining their
11058               validity prior to issuing diagnostics.
11059
11060           store-merging-max-size
11061               Maximum size of a single store merging region in bytes.
11062
11063           hash-table-verification-limit
11064               The number of elements for which hash table verification is
11065               done for each searched element.
11066
11067           max-find-base-term-values
11068               Maximum number of VALUEs handled during a single find_base_term
11069               call.
11070
11071           analyzer-max-enodes-per-program-point
11072               The maximum number of exploded nodes per program point within
11073               the analyzer, before terminating analysis of that point.
11074
11075           analyzer-min-snodes-for-call-summary
11076               The minimum number of supernodes within a function for the
11077               analyzer to consider summarizing its effects at call sites.
11078
11079           analyzer-max-recursion-depth
11080               The maximum number of times a callsite can appear in a call
11081               stack within the analyzer, before terminating analysis of a
11082               call that would recurse deeper.
11083
11084           gimple-fe-computed-hot-bb-threshold
11085               The number of executions of a basic block which is considered
11086               hot.  The parameter is used only in GIMPLE FE.
11087
11088           analyzer-bb-explosion-factor
11089               The maximum number of 'after supernode' exploded nodes within
11090               the analyzer per supernode, before terminating analysis.
11091
11092           The following choices of name are available on AArch64 targets:
11093
11094           aarch64-sve-compare-costs
11095               When vectorizing for SVE, consider using "unpacked" vectors for
11096               smaller elements and use the cost model to pick the cheapest
11097               approach.  Also use the cost model to choose between SVE and
11098               Advanced SIMD vectorization.
11099
11100               Using unpacked vectors includes storing smaller elements in
11101               larger containers and accessing elements with extending loads
11102               and truncating stores.
11103
11104           aarch64-float-recp-precision
11105               The number of Newton iterations for calculating the reciprocal
11106               for float type.  The precision of division is proportional to
11107               this param when division approximation is enabled.  The default
11108               value is 1.
11109
11110           aarch64-double-recp-precision
11111               The number of Newton iterations for calculating the reciprocal
11112               for double type.  The precision of division is propotional to
11113               this param when division approximation is enabled.  The default
11114               value is 2.
11115
11116   Program Instrumentation Options
11117       GCC supports a number of command-line options that control adding run-
11118       time instrumentation to the code it normally generates.  For example,
11119       one purpose of instrumentation is collect profiling statistics for use
11120       in finding program hot spots, code coverage analysis, or profile-guided
11121       optimizations.  Another class of program instrumentation is adding run-
11122       time checking to detect programming errors like invalid pointer
11123       dereferences or out-of-bounds array accesses, as well as deliberately
11124       hostile attacks such as stack smashing or C++ vtable hijacking.  There
11125       is also a general hook which can be used to implement other forms of
11126       tracing or function-level instrumentation for debug or program analysis
11127       purposes.
11128
11129       -p
11130       -pg Generate extra code to write profile information suitable for the
11131           analysis program prof (for -p) or gprof (for -pg).  You must use
11132           this option when compiling the source files you want data about,
11133           and you must also use it when linking.
11134
11135           You can use the function attribute "no_instrument_function" to
11136           suppress profiling of individual functions when compiling with
11137           these options.
11138
11139       -fprofile-arcs
11140           Add code so that program flow arcs are instrumented.  During
11141           execution the program records how many times each branch and call
11142           is executed and how many times it is taken or returns.  On targets
11143           that support constructors with priority support, profiling properly
11144           handles constructors, destructors and C++ constructors (and
11145           destructors) of classes which are used as a type of a global
11146           variable.
11147
11148           When the compiled program exits it saves this data to a file called
11149           auxname.gcda for each source file.  The data may be used for
11150           profile-directed optimizations (-fbranch-probabilities), or for
11151           test coverage analysis (-ftest-coverage).  Each object file's
11152           auxname is generated from the name of the output file, if
11153           explicitly specified and it is not the final executable, otherwise
11154           it is the basename of the source file.  In both cases any suffix is
11155           removed (e.g. foo.gcda for input file dir/foo.c, or dir/foo.gcda
11156           for output file specified as -o dir/foo.o).
11157
11158       --coverage
11159           This option is used to compile and link code instrumented for
11160           coverage analysis.  The option is a synonym for -fprofile-arcs
11161           -ftest-coverage (when compiling) and -lgcov (when linking).  See
11162           the documentation for those options for more details.
11163
11164           *   Compile the source files with -fprofile-arcs plus optimization
11165               and code generation options.  For test coverage analysis, use
11166               the additional -ftest-coverage option.  You do not need to
11167               profile every source file in a program.
11168
11169           *   Compile the source files additionally with -fprofile-abs-path
11170               to create absolute path names in the .gcno files.  This allows
11171               gcov to find the correct sources in projects where compilations
11172               occur with different working directories.
11173
11174           *   Link your object files with -lgcov or -fprofile-arcs (the
11175               latter implies the former).
11176
11177           *   Run the program on a representative workload to generate the
11178               arc profile information.  This may be repeated any number of
11179               times.  You can run concurrent instances of your program, and
11180               provided that the file system supports locking, the data files
11181               will be correctly updated.  Unless a strict ISO C dialect
11182               option is in effect, "fork" calls are detected and correctly
11183               handled without double counting.
11184
11185           *   For profile-directed optimizations, compile the source files
11186               again with the same optimization and code generation options
11187               plus -fbranch-probabilities.
11188
11189           *   For test coverage analysis, use gcov to produce human readable
11190               information from the .gcno and .gcda files.  Refer to the gcov
11191               documentation for further information.
11192
11193           With -fprofile-arcs, for each function of your program GCC creates
11194           a program flow graph, then finds a spanning tree for the graph.
11195           Only arcs that are not on the spanning tree have to be
11196           instrumented: the compiler adds code to count the number of times
11197           that these arcs are executed.  When an arc is the only exit or only
11198           entrance to a block, the instrumentation code can be added to the
11199           block; otherwise, a new basic block must be created to hold the
11200           instrumentation code.
11201
11202       -ftest-coverage
11203           Produce a notes file that the gcov code-coverage utility can use to
11204           show program coverage.  Each source file's note file is called
11205           auxname.gcno.  Refer to the -fprofile-arcs option above for a
11206           description of auxname and instructions on how to generate test
11207           coverage data.  Coverage data matches the source files more closely
11208           if you do not optimize.
11209
11210       -fprofile-abs-path
11211           Automatically convert relative source file names to absolute path
11212           names in the .gcno files.  This allows gcov to find the correct
11213           sources in projects where compilations occur with different working
11214           directories.
11215
11216       -fprofile-dir=path
11217           Set the directory to search for the profile data files in to path.
11218           This option affects only the profile data generated by
11219           -fprofile-generate, -ftest-coverage, -fprofile-arcs and used by
11220           -fprofile-use and -fbranch-probabilities and its related options.
11221           Both absolute and relative paths can be used.  By default, GCC uses
11222           the current directory as path, thus the profile data file appears
11223           in the same directory as the object file.  In order to prevent the
11224           file name clashing, if the object file name is not an absolute
11225           path, we mangle the absolute path of the sourcename.gcda file and
11226           use it as the file name of a .gcda file.  See similar option
11227           -fprofile-note.
11228
11229           When an executable is run in a massive parallel environment, it is
11230           recommended to save profile to different folders.  That can be done
11231           with variables in path that are exported during run-time:
11232
11233           %p  process ID.
11234
11235           %q{VAR}
11236               value of environment variable VAR
11237
11238       -fprofile-generate
11239       -fprofile-generate=path
11240           Enable options usually used for instrumenting application to
11241           produce profile useful for later recompilation with profile
11242           feedback based optimization.  You must use -fprofile-generate both
11243           when compiling and when linking your program.
11244
11245           The following options are enabled: -fprofile-arcs,
11246           -fprofile-values, -finline-functions, and -fipa-bit-cp.
11247
11248           If path is specified, GCC looks at the path to find the profile
11249           feedback data files. See -fprofile-dir.
11250
11251           To optimize the program based on the collected profile information,
11252           use -fprofile-use.
11253
11254       -fprofile-note=path
11255           If path is specified, GCC saves .gcno file into path location.  If
11256           you combine the option with multiple source files, the .gcno file
11257           will be overwritten.
11258
11259       -fprofile-prefix-path=path
11260           This option can be used in combination with
11261           profile-generate=profile_dir and profile-use=profile_dir to inform
11262           GCC where is the base directory of built source tree.  By default
11263           profile_dir will contain files with mangled absolute paths of all
11264           object files in the built project.  This is not desirable when
11265           directory used to build the instrumented binary differs from the
11266           directory used to build the binary optimized with profile feedback
11267           because the profile data will not be found during the optimized
11268           build.  In such setups -fprofile-prefix-path=path with path
11269           pointing to the base directory of the build can be used to strip
11270           the irrelevant part of the path and keep all file names relative to
11271           the main build directory.
11272
11273       -fprofile-update=method
11274           Alter the update method for an application instrumented for profile
11275           feedback based optimization.  The method argument should be one of
11276           single, atomic or prefer-atomic.  The first one is useful for
11277           single-threaded applications, while the second one prevents profile
11278           corruption by emitting thread-safe code.
11279
11280           Warning: When an application does not properly join all threads (or
11281           creates an detached thread), a profile file can be still corrupted.
11282
11283           Using prefer-atomic would be transformed either to atomic, when
11284           supported by a target, or to single otherwise.  The GCC driver
11285           automatically selects prefer-atomic when -pthread is present in the
11286           command line.
11287
11288       -fprofile-filter-files=regex
11289           Instrument only functions from files where names match any regular
11290           expression (separated by a semi-colon).
11291
11292           For example, -fprofile-filter-files=main.c;module.*.c will
11293           instrument only main.c and all C files starting with 'module'.
11294
11295       -fprofile-exclude-files=regex
11296           Instrument only functions from files where names do not match all
11297           the regular expressions (separated by a semi-colon).
11298
11299           For example, -fprofile-exclude-files=/usr/* will prevent
11300           instrumentation of all files that are located in /usr/ folder.
11301
11302       -fprofile-reproducible=[multithreaded|parallel-runs|serial]
11303           Control level of reproducibility of profile gathered by
11304           "-fprofile-generate".  This makes it possible to rebuild program
11305           with same outcome which is useful, for example, for distribution
11306           packages.
11307
11308           With -fprofile-reproducible=serial the profile gathered by
11309           -fprofile-generate is reproducible provided the trained program
11310           behaves the same at each invocation of the train run, it is not
11311           multi-threaded and profile data streaming is always done in the
11312           same order.  Note that profile streaming happens at the end of
11313           program run but also before "fork" function is invoked.
11314
11315           Note that it is quite common that execution counts of some part of
11316           programs depends, for example, on length of temporary file names or
11317           memory space randomization (that may affect hash-table collision
11318           rate).  Such non-reproducible part of programs may be annotated by
11319           "no_instrument_function" function attribute. "gcov-dump" with -l
11320           can be used to dump gathered data and verify that they are indeed
11321           reproducible.
11322
11323           With -fprofile-reproducible=parallel-runs collected profile stays
11324           reproducible regardless the order of streaming of the data into
11325           gcda files.  This setting makes it possible to run multiple
11326           instances of instrumented program in parallel (such as with "make
11327           -j"). This reduces quality of gathered data, in particular of
11328           indirect call profiling.
11329
11330       -fsanitize=address
11331           Enable AddressSanitizer, a fast memory error detector.  Memory
11332           access instructions are instrumented to detect out-of-bounds and
11333           use-after-free bugs.  The option enables
11334           -fsanitize-address-use-after-scope.  See
11335           <https://github.com/google/sanitizers/wiki/AddressSanitizer> for
11336           more details.  The run-time behavior can be influenced using the
11337           ASAN_OPTIONS environment variable.  When set to "help=1", the
11338           available options are shown at startup of the instrumented program.
11339           See
11340           <https://github.com/google/sanitizers/wiki/AddressSanitizerFlags#run-time-flags>
11341           for a list of supported options.  The option cannot be combined
11342           with -fsanitize=thread.
11343
11344       -fsanitize=kernel-address
11345           Enable AddressSanitizer for Linux kernel.  See
11346           <https://github.com/google/kasan/wiki> for more details.
11347
11348       -fsanitize=pointer-compare
11349           Instrument comparison operation (<, <=, >, >=) with pointer
11350           operands.  The option must be combined with either
11351           -fsanitize=kernel-address or -fsanitize=address The option cannot
11352           be combined with -fsanitize=thread.  Note: By default the check is
11353           disabled at run time.  To enable it, add
11354           "detect_invalid_pointer_pairs=2" to the environment variable
11355           ASAN_OPTIONS. Using "detect_invalid_pointer_pairs=1" detects
11356           invalid operation only when both pointers are non-null.
11357
11358       -fsanitize=pointer-subtract
11359           Instrument subtraction with pointer operands.  The option must be
11360           combined with either -fsanitize=kernel-address or
11361           -fsanitize=address The option cannot be combined with
11362           -fsanitize=thread.  Note: By default the check is disabled at run
11363           time.  To enable it, add "detect_invalid_pointer_pairs=2" to the
11364           environment variable ASAN_OPTIONS. Using
11365           "detect_invalid_pointer_pairs=1" detects invalid operation only
11366           when both pointers are non-null.
11367
11368       -fsanitize=thread
11369           Enable ThreadSanitizer, a fast data race detector.  Memory access
11370           instructions are instrumented to detect data race bugs.  See
11371           <https://github.com/google/sanitizers/wiki#threadsanitizer> for
11372           more details. The run-time behavior can be influenced using the
11373           TSAN_OPTIONS environment variable; see
11374           <https://github.com/google/sanitizers/wiki/ThreadSanitizerFlags>
11375           for a list of supported options.  The option cannot be combined
11376           with -fsanitize=address, -fsanitize=leak.
11377
11378           Note that sanitized atomic builtins cannot throw exceptions when
11379           operating on invalid memory addresses with non-call exceptions
11380           (-fnon-call-exceptions).
11381
11382       -fsanitize=leak
11383           Enable LeakSanitizer, a memory leak detector.  This option only
11384           matters for linking of executables and the executable is linked
11385           against a library that overrides "malloc" and other allocator
11386           functions.  See
11387           <https://github.com/google/sanitizers/wiki/AddressSanitizerLeakSanitizer>
11388           for more details.  The run-time behavior can be influenced using
11389           the LSAN_OPTIONS environment variable.  The option cannot be
11390           combined with -fsanitize=thread.
11391
11392       -fsanitize=undefined
11393           Enable UndefinedBehaviorSanitizer, a fast undefined behavior
11394           detector.  Various computations are instrumented to detect
11395           undefined behavior at runtime.  Current suboptions are:
11396
11397           -fsanitize=shift
11398               This option enables checking that the result of a shift
11399               operation is not undefined.  Note that what exactly is
11400               considered undefined differs slightly between C and C++, as
11401               well as between ISO C90 and C99, etc.  This option has two
11402               suboptions, -fsanitize=shift-base and
11403               -fsanitize=shift-exponent.
11404
11405           -fsanitize=shift-exponent
11406               This option enables checking that the second argument of a
11407               shift operation is not negative and is smaller than the
11408               precision of the promoted first argument.
11409
11410           -fsanitize=shift-base
11411               If the second argument of a shift operation is within range,
11412               check that the result of a shift operation is not undefined.
11413               Note that what exactly is considered undefined differs slightly
11414               between C and C++, as well as between ISO C90 and C99, etc.
11415
11416           -fsanitize=integer-divide-by-zero
11417               Detect integer division by zero as well as "INT_MIN / -1"
11418               division.
11419
11420           -fsanitize=unreachable
11421               With this option, the compiler turns the
11422               "__builtin_unreachable" call into a diagnostics message call
11423               instead.  When reaching the "__builtin_unreachable" call, the
11424               behavior is undefined.
11425
11426           -fsanitize=vla-bound
11427               This option instructs the compiler to check that the size of a
11428               variable length array is positive.
11429
11430           -fsanitize=null
11431               This option enables pointer checking.  Particularly, the
11432               application built with this option turned on will issue an
11433               error message when it tries to dereference a NULL pointer, or
11434               if a reference (possibly an rvalue reference) is bound to a
11435               NULL pointer, or if a method is invoked on an object pointed by
11436               a NULL pointer.
11437
11438           -fsanitize=return
11439               This option enables return statement checking.  Programs built
11440               with this option turned on will issue an error message when the
11441               end of a non-void function is reached without actually
11442               returning a value.  This option works in C++ only.
11443
11444           -fsanitize=signed-integer-overflow
11445               This option enables signed integer overflow checking.  We check
11446               that the result of "+", "*", and both unary and binary "-" does
11447               not overflow in the signed arithmetics.  Note, integer
11448               promotion rules must be taken into account.  That is, the
11449               following is not an overflow:
11450
11451                       signed char a = SCHAR_MAX;
11452                       a++;
11453
11454           -fsanitize=bounds
11455               This option enables instrumentation of array bounds.  Various
11456               out of bounds accesses are detected.  Flexible array members,
11457               flexible array member-like arrays, and initializers of
11458               variables with static storage are not instrumented.
11459
11460           -fsanitize=bounds-strict
11461               This option enables strict instrumentation of array bounds.
11462               Most out of bounds accesses are detected, including flexible
11463               array members and flexible array member-like arrays.
11464               Initializers of variables with static storage are not
11465               instrumented.
11466
11467           -fsanitize=alignment
11468               This option enables checking of alignment of pointers when they
11469               are dereferenced, or when a reference is bound to
11470               insufficiently aligned target, or when a method or constructor
11471               is invoked on insufficiently aligned object.
11472
11473           -fsanitize=object-size
11474               This option enables instrumentation of memory references using
11475               the "__builtin_object_size" function.  Various out of bounds
11476               pointer accesses are detected.
11477
11478           -fsanitize=float-divide-by-zero
11479               Detect floating-point division by zero.  Unlike other similar
11480               options, -fsanitize=float-divide-by-zero is not enabled by
11481               -fsanitize=undefined, since floating-point division by zero can
11482               be a legitimate way of obtaining infinities and NaNs.
11483
11484           -fsanitize=float-cast-overflow
11485               This option enables floating-point type to integer conversion
11486               checking.  We check that the result of the conversion does not
11487               overflow.  Unlike other similar options,
11488               -fsanitize=float-cast-overflow is not enabled by
11489               -fsanitize=undefined.  This option does not work well with
11490               "FE_INVALID" exceptions enabled.
11491
11492           -fsanitize=nonnull-attribute
11493               This option enables instrumentation of calls, checking whether
11494               null values are not passed to arguments marked as requiring a
11495               non-null value by the "nonnull" function attribute.
11496
11497           -fsanitize=returns-nonnull-attribute
11498               This option enables instrumentation of return statements in
11499               functions marked with "returns_nonnull" function attribute, to
11500               detect returning of null values from such functions.
11501
11502           -fsanitize=bool
11503               This option enables instrumentation of loads from bool.  If a
11504               value other than 0/1 is loaded, a run-time error is issued.
11505
11506           -fsanitize=enum
11507               This option enables instrumentation of loads from an enum type.
11508               If a value outside the range of values for the enum type is
11509               loaded, a run-time error is issued.
11510
11511           -fsanitize=vptr
11512               This option enables instrumentation of C++ member function
11513               calls, member accesses and some conversions between pointers to
11514               base and derived classes, to verify the referenced object has
11515               the correct dynamic type.
11516
11517           -fsanitize=pointer-overflow
11518               This option enables instrumentation of pointer arithmetics.  If
11519               the pointer arithmetics overflows, a run-time error is issued.
11520
11521           -fsanitize=builtin
11522               This option enables instrumentation of arguments to selected
11523               builtin functions.  If an invalid value is passed to such
11524               arguments, a run-time error is issued.  E.g. passing 0 as the
11525               argument to "__builtin_ctz" or "__builtin_clz" invokes
11526               undefined behavior and is diagnosed by this option.
11527
11528           While -ftrapv causes traps for signed overflows to be emitted,
11529           -fsanitize=undefined gives a diagnostic message.  This currently
11530           works only for the C family of languages.
11531
11532       -fno-sanitize=all
11533           This option disables all previously enabled sanitizers.
11534           -fsanitize=all is not allowed, as some sanitizers cannot be used
11535           together.
11536
11537       -fasan-shadow-offset=number
11538           This option forces GCC to use custom shadow offset in
11539           AddressSanitizer checks.  It is useful for experimenting with
11540           different shadow memory layouts in Kernel AddressSanitizer.
11541
11542       -fsanitize-sections=s1,s2,...
11543           Sanitize global variables in selected user-defined sections.  si
11544           may contain wildcards.
11545
11546       -fsanitize-recover[=opts]
11547           -fsanitize-recover= controls error recovery mode for sanitizers
11548           mentioned in comma-separated list of opts.  Enabling this option
11549           for a sanitizer component causes it to attempt to continue running
11550           the program as if no error happened.  This means multiple runtime
11551           errors can be reported in a single program run, and the exit code
11552           of the program may indicate success even when errors have been
11553           reported.  The -fno-sanitize-recover= option can be used to alter
11554           this behavior: only the first detected error is reported and
11555           program then exits with a non-zero exit code.
11556
11557           Currently this feature only works for -fsanitize=undefined (and its
11558           suboptions except for -fsanitize=unreachable and
11559           -fsanitize=return), -fsanitize=float-cast-overflow,
11560           -fsanitize=float-divide-by-zero, -fsanitize=bounds-strict,
11561           -fsanitize=kernel-address and -fsanitize=address.  For these
11562           sanitizers error recovery is turned on by default, except
11563           -fsanitize=address, for which this feature is experimental.
11564           -fsanitize-recover=all and -fno-sanitize-recover=all is also
11565           accepted, the former enables recovery for all sanitizers that
11566           support it, the latter disables recovery for all sanitizers that
11567           support it.
11568
11569           Even if a recovery mode is turned on the compiler side, it needs to
11570           be also enabled on the runtime library side, otherwise the failures
11571           are still fatal.  The runtime library defaults to "halt_on_error=0"
11572           for ThreadSanitizer and UndefinedBehaviorSanitizer, while default
11573           value for AddressSanitizer is "halt_on_error=1". This can be
11574           overridden through setting the "halt_on_error" flag in the
11575           corresponding environment variable.
11576
11577           Syntax without an explicit opts parameter is deprecated.  It is
11578           equivalent to specifying an opts list of:
11579
11580                   undefined,float-cast-overflow,float-divide-by-zero,bounds-strict
11581
11582       -fsanitize-address-use-after-scope
11583           Enable sanitization of local variables to detect use-after-scope
11584           bugs.  The option sets -fstack-reuse to none.
11585
11586       -fsanitize-undefined-trap-on-error
11587           The -fsanitize-undefined-trap-on-error option instructs the
11588           compiler to report undefined behavior using "__builtin_trap" rather
11589           than a "libubsan" library routine.  The advantage of this is that
11590           the "libubsan" library is not needed and is not linked in, so this
11591           is usable even in freestanding environments.
11592
11593       -fsanitize-coverage=trace-pc
11594           Enable coverage-guided fuzzing code instrumentation.  Inserts a
11595           call to "__sanitizer_cov_trace_pc" into every basic block.
11596
11597       -fsanitize-coverage=trace-cmp
11598           Enable dataflow guided fuzzing code instrumentation.  Inserts a
11599           call to "__sanitizer_cov_trace_cmp1", "__sanitizer_cov_trace_cmp2",
11600           "__sanitizer_cov_trace_cmp4" or "__sanitizer_cov_trace_cmp8" for
11601           integral comparison with both operands variable or
11602           "__sanitizer_cov_trace_const_cmp1",
11603           "__sanitizer_cov_trace_const_cmp2",
11604           "__sanitizer_cov_trace_const_cmp4" or
11605           "__sanitizer_cov_trace_const_cmp8" for integral comparison with one
11606           operand constant, "__sanitizer_cov_trace_cmpf" or
11607           "__sanitizer_cov_trace_cmpd" for float or double comparisons and
11608           "__sanitizer_cov_trace_switch" for switch statements.
11609
11610       -fcf-protection=[full|branch|return|none|check]
11611           Enable code instrumentation of control-flow transfers to increase
11612           program security by checking that target addresses of control-flow
11613           transfer instructions (such as indirect function call, function
11614           return, indirect jump) are valid.  This prevents diverting the flow
11615           of control to an unexpected target.  This is intended to protect
11616           against such threats as Return-oriented Programming (ROP), and
11617           similarly call/jmp-oriented programming (COP/JOP).
11618
11619           The value "branch" tells the compiler to implement checking of
11620           validity of control-flow transfer at the point of indirect branch
11621           instructions, i.e. call/jmp instructions.  The value "return"
11622           implements checking of validity at the point of returning from a
11623           function.  The value "full" is an alias for specifying both
11624           "branch" and "return". The value "none" turns off instrumentation.
11625
11626           The value "check" is used for the final link with link-time
11627           optimization (LTO).  An error is issued if LTO object files are
11628           compiled with different -fcf-protection values.  The value "check"
11629           is ignored at the compile time.
11630
11631           The macro "__CET__" is defined when -fcf-protection is used.  The
11632           first bit of "__CET__" is set to 1 for the value "branch" and the
11633           second bit of "__CET__" is set to 1 for the "return".
11634
11635           You can also use the "nocf_check" attribute to identify which
11636           functions and calls should be skipped from instrumentation.
11637
11638           Currently the x86 GNU/Linux target provides an implementation based
11639           on Intel Control-flow Enforcement Technology (CET).
11640
11641       -fstack-protector
11642           Emit extra code to check for buffer overflows, such as stack
11643           smashing attacks.  This is done by adding a guard variable to
11644           functions with vulnerable objects.  This includes functions that
11645           call "alloca", and functions with buffers larger than or equal to 8
11646           bytes.  The guards are initialized when a function is entered and
11647           then checked when the function exits.  If a guard check fails, an
11648           error message is printed and the program exits.  Only variables
11649           that are actually allocated on the stack are considered, optimized
11650           away variables or variables allocated in registers don't count.
11651
11652       -fstack-protector-all
11653           Like -fstack-protector except that all functions are protected.
11654
11655       -fstack-protector-strong
11656           Like -fstack-protector but includes additional functions to be
11657           protected --- those that have local array definitions, or have
11658           references to local frame addresses.  Only variables that are
11659           actually allocated on the stack are considered, optimized away
11660           variables or variables allocated in registers don't count.
11661
11662       -fstack-protector-explicit
11663           Like -fstack-protector but only protects those functions which have
11664           the "stack_protect" attribute.
11665
11666       -fstack-check
11667           Generate code to verify that you do not go beyond the boundary of
11668           the stack.  You should specify this flag if you are running in an
11669           environment with multiple threads, but you only rarely need to
11670           specify it in a single-threaded environment since stack overflow is
11671           automatically detected on nearly all systems if there is only one
11672           stack.
11673
11674           Note that this switch does not actually cause checking to be done;
11675           the operating system or the language runtime must do that.  The
11676           switch causes generation of code to ensure that they see the stack
11677           being extended.
11678
11679           You can additionally specify a string parameter: no means no
11680           checking, generic means force the use of old-style checking,
11681           specific means use the best checking method and is equivalent to
11682           bare -fstack-check.
11683
11684           Old-style checking is a generic mechanism that requires no specific
11685           target support in the compiler but comes with the following
11686           drawbacks:
11687
11688           1.  Modified allocation strategy for large objects: they are always
11689               allocated dynamically if their size exceeds a fixed threshold.
11690               Note this may change the semantics of some code.
11691
11692           2.  Fixed limit on the size of the static frame of functions: when
11693               it is topped by a particular function, stack checking is not
11694               reliable and a warning is issued by the compiler.
11695
11696           3.  Inefficiency: because of both the modified allocation strategy
11697               and the generic implementation, code performance is hampered.
11698
11699           Note that old-style stack checking is also the fallback method for
11700           specific if no target support has been added in the compiler.
11701
11702           -fstack-check= is designed for Ada's needs to detect infinite
11703           recursion and stack overflows.  specific is an excellent choice
11704           when compiling Ada code.  It is not generally sufficient to protect
11705           against stack-clash attacks.  To protect against those you want
11706           -fstack-clash-protection.
11707
11708       -fstack-clash-protection
11709           Generate code to prevent stack clash style attacks.  When this
11710           option is enabled, the compiler will only allocate one page of
11711           stack space at a time and each page is accessed immediately after
11712           allocation.  Thus, it prevents allocations from jumping over any
11713           stack guard page provided by the operating system.
11714
11715           Most targets do not fully support stack clash protection.  However,
11716           on those targets -fstack-clash-protection will protect dynamic
11717           stack allocations.  -fstack-clash-protection may also provide
11718           limited protection for static stack allocations if the target
11719           supports -fstack-check=specific.
11720
11721       -fstack-limit-register=reg
11722       -fstack-limit-symbol=sym
11723       -fno-stack-limit
11724           Generate code to ensure that the stack does not grow beyond a
11725           certain value, either the value of a register or the address of a
11726           symbol.  If a larger stack is required, a signal is raised at run
11727           time.  For most targets, the signal is raised before the stack
11728           overruns the boundary, so it is possible to catch the signal
11729           without taking special precautions.
11730
11731           For instance, if the stack starts at absolute address 0x80000000
11732           and grows downwards, you can use the flags
11733           -fstack-limit-symbol=__stack_limit and
11734           -Wl,--defsym,__stack_limit=0x7ffe0000 to enforce a stack limit of
11735           128KB.  Note that this may only work with the GNU linker.
11736
11737           You can locally override stack limit checking by using the
11738           "no_stack_limit" function attribute.
11739
11740       -fsplit-stack
11741           Generate code to automatically split the stack before it overflows.
11742           The resulting program has a discontiguous stack which can only
11743           overflow if the program is unable to allocate any more memory.
11744           This is most useful when running threaded programs, as it is no
11745           longer necessary to calculate a good stack size to use for each
11746           thread.  This is currently only implemented for the x86 targets
11747           running GNU/Linux.
11748
11749           When code compiled with -fsplit-stack calls code compiled without
11750           -fsplit-stack, there may not be much stack space available for the
11751           latter code to run.  If compiling all code, including library code,
11752           with -fsplit-stack is not an option, then the linker can fix up
11753           these calls so that the code compiled without -fsplit-stack always
11754           has a large stack.  Support for this is implemented in the gold
11755           linker in GNU binutils release 2.21 and later.
11756
11757       -fvtable-verify=[std|preinit|none]
11758           This option is only available when compiling C++ code.  It turns on
11759           (or off, if using -fvtable-verify=none) the security feature that
11760           verifies at run time, for every virtual call, that the vtable
11761           pointer through which the call is made is valid for the type of the
11762           object, and has not been corrupted or overwritten.  If an invalid
11763           vtable pointer is detected at run time, an error is reported and
11764           execution of the program is immediately halted.
11765
11766           This option causes run-time data structures to be built at program
11767           startup, which are used for verifying the vtable pointers.  The
11768           options std and preinit control the timing of when these data
11769           structures are built.  In both cases the data structures are built
11770           before execution reaches "main".  Using -fvtable-verify=std causes
11771           the data structures to be built after shared libraries have been
11772           loaded and initialized.  -fvtable-verify=preinit causes them to be
11773           built before shared libraries have been loaded and initialized.
11774
11775           If this option appears multiple times in the command line with
11776           different values specified, none takes highest priority over both
11777           std and preinit; preinit takes priority over std.
11778
11779       -fvtv-debug
11780           When used in conjunction with -fvtable-verify=std or
11781           -fvtable-verify=preinit, causes debug versions of the runtime
11782           functions for the vtable verification feature to be called.  This
11783           flag also causes the compiler to log information about which vtable
11784           pointers it finds for each class.  This information is written to a
11785           file named vtv_set_ptr_data.log in the directory named by the
11786           environment variable VTV_LOGS_DIR if that is defined or the current
11787           working directory otherwise.
11788
11789           Note:  This feature appends data to the log file. If you want a
11790           fresh log file, be sure to delete any existing one.
11791
11792       -fvtv-counts
11793           This is a debugging flag.  When used in conjunction with
11794           -fvtable-verify=std or -fvtable-verify=preinit, this causes the
11795           compiler to keep track of the total number of virtual calls it
11796           encounters and the number of verifications it inserts.  It also
11797           counts the number of calls to certain run-time library functions
11798           that it inserts and logs this information for each compilation
11799           unit.  The compiler writes this information to a file named
11800           vtv_count_data.log in the directory named by the environment
11801           variable VTV_LOGS_DIR if that is defined or the current working
11802           directory otherwise.  It also counts the size of the vtable pointer
11803           sets for each class, and writes this information to
11804           vtv_class_set_sizes.log in the same directory.
11805
11806           Note:  This feature appends data to the log files.  To get fresh
11807           log files, be sure to delete any existing ones.
11808
11809       -finstrument-functions
11810           Generate instrumentation calls for entry and exit to functions.
11811           Just after function entry and just before function exit, the
11812           following profiling functions are called with the address of the
11813           current function and its call site.  (On some platforms,
11814           "__builtin_return_address" does not work beyond the current
11815           function, so the call site information may not be available to the
11816           profiling functions otherwise.)
11817
11818                   void __cyg_profile_func_enter (void *this_fn,
11819                                                  void *call_site);
11820                   void __cyg_profile_func_exit  (void *this_fn,
11821                                                  void *call_site);
11822
11823           The first argument is the address of the start of the current
11824           function, which may be looked up exactly in the symbol table.
11825
11826           This instrumentation is also done for functions expanded inline in
11827           other functions.  The profiling calls indicate where, conceptually,
11828           the inline function is entered and exited.  This means that
11829           addressable versions of such functions must be available.  If all
11830           your uses of a function are expanded inline, this may mean an
11831           additional expansion of code size.  If you use "extern inline" in
11832           your C code, an addressable version of such functions must be
11833           provided.  (This is normally the case anyway, but if you get lucky
11834           and the optimizer always expands the functions inline, you might
11835           have gotten away without providing static copies.)
11836
11837           A function may be given the attribute "no_instrument_function", in
11838           which case this instrumentation is not done.  This can be used, for
11839           example, for the profiling functions listed above, high-priority
11840           interrupt routines, and any functions from which the profiling
11841           functions cannot safely be called (perhaps signal handlers, if the
11842           profiling routines generate output or allocate memory).
11843
11844       -finstrument-functions-exclude-file-list=file,file,...
11845           Set the list of functions that are excluded from instrumentation
11846           (see the description of -finstrument-functions).  If the file that
11847           contains a function definition matches with one of file, then that
11848           function is not instrumented.  The match is done on substrings: if
11849           the file parameter is a substring of the file name, it is
11850           considered to be a match.
11851
11852           For example:
11853
11854                   -finstrument-functions-exclude-file-list=/bits/stl,include/sys
11855
11856           excludes any inline function defined in files whose pathnames
11857           contain /bits/stl or include/sys.
11858
11859           If, for some reason, you want to include letter , in one of sym,
11860           write ,. For example,
11861           -finstrument-functions-exclude-file-list=',,tmp' (note the single
11862           quote surrounding the option).
11863
11864       -finstrument-functions-exclude-function-list=sym,sym,...
11865           This is similar to -finstrument-functions-exclude-file-list, but
11866           this option sets the list of function names to be excluded from
11867           instrumentation.  The function name to be matched is its user-
11868           visible name, such as "vector<int> blah(const vector<int> &)", not
11869           the internal mangled name (e.g., "_Z4blahRSt6vectorIiSaIiEE").  The
11870           match is done on substrings: if the sym parameter is a substring of
11871           the function name, it is considered to be a match.  For C99 and C++
11872           extended identifiers, the function name must be given in UTF-8, not
11873           using universal character names.
11874
11875       -fpatchable-function-entry=N[,M]
11876           Generate N NOPs right at the beginning of each function, with the
11877           function entry point before the Mth NOP.  If M is omitted, it
11878           defaults to 0 so the function entry points to the address just at
11879           the first NOP.  The NOP instructions reserve extra space which can
11880           be used to patch in any desired instrumentation at run time,
11881           provided that the code segment is writable.  The amount of space is
11882           controllable indirectly via the number of NOPs; the NOP instruction
11883           used corresponds to the instruction emitted by the internal GCC
11884           back-end interface "gen_nop".  This behavior is target-specific and
11885           may also depend on the architecture variant and/or other
11886           compilation options.
11887
11888           For run-time identification, the starting addresses of these areas,
11889           which correspond to their respective function entries minus M, are
11890           additionally collected in the "__patchable_function_entries"
11891           section of the resulting binary.
11892
11893           Note that the value of "__attribute__ ((patchable_function_entry
11894           (N,M)))" takes precedence over command-line option
11895           -fpatchable-function-entry=N,M.  This can be used to increase the
11896           area size or to remove it completely on a single function.  If
11897           "N=0", no pad location is recorded.
11898
11899           The NOP instructions are inserted at---and maybe before, depending
11900           on M---the function entry address, even before the prologue.
11901
11902   Options Controlling the Preprocessor
11903       These options control the C preprocessor, which is run on each C source
11904       file before actual compilation.
11905
11906       If you use the -E option, nothing is done except preprocessing.  Some
11907       of these options make sense only together with -E because they cause
11908       the preprocessor output to be unsuitable for actual compilation.
11909
11910       In addition to the options listed here, there are a number of options
11911       to control search paths for include files documented in Directory
11912       Options.  Options to control preprocessor diagnostics are listed in
11913       Warning Options.
11914
11915       -D name
11916           Predefine name as a macro, with definition 1.
11917
11918       -D name=definition
11919           The contents of definition are tokenized and processed as if they
11920           appeared during translation phase three in a #define directive.  In
11921           particular, the definition is truncated by embedded newline
11922           characters.
11923
11924           If you are invoking the preprocessor from a shell or shell-like
11925           program you may need to use the shell's quoting syntax to protect
11926           characters such as spaces that have a meaning in the shell syntax.
11927
11928           If you wish to define a function-like macro on the command line,
11929           write its argument list with surrounding parentheses before the
11930           equals sign (if any).  Parentheses are meaningful to most shells,
11931           so you should quote the option.  With sh and csh,
11932           -D'name(args...)=definition' works.
11933
11934           -D and -U options are processed in the order they are given on the
11935           command line.  All -imacros file and -include file options are
11936           processed after all -D and -U options.
11937
11938       -U name
11939           Cancel any previous definition of name, either built in or provided
11940           with a -D option.
11941
11942       -include file
11943           Process file as if "#include "file"" appeared as the first line of
11944           the primary source file.  However, the first directory searched for
11945           file is the preprocessor's working directory instead of the
11946           directory containing the main source file.  If not found there, it
11947           is searched for in the remainder of the "#include "..."" search
11948           chain as normal.
11949
11950           If multiple -include options are given, the files are included in
11951           the order they appear on the command line.
11952
11953       -imacros file
11954           Exactly like -include, except that any output produced by scanning
11955           file is thrown away.  Macros it defines remain defined.  This
11956           allows you to acquire all the macros from a header without also
11957           processing its declarations.
11958
11959           All files specified by -imacros are processed before all files
11960           specified by -include.
11961
11962       -undef
11963           Do not predefine any system-specific or GCC-specific macros.  The
11964           standard predefined macros remain defined.
11965
11966       -pthread
11967           Define additional macros required for using the POSIX threads
11968           library.  You should use this option consistently for both
11969           compilation and linking.  This option is supported on GNU/Linux
11970           targets, most other Unix derivatives, and also on x86 Cygwin and
11971           MinGW targets.
11972
11973       -M  Instead of outputting the result of preprocessing, output a rule
11974           suitable for make describing the dependencies of the main source
11975           file.  The preprocessor outputs one make rule containing the object
11976           file name for that source file, a colon, and the names of all the
11977           included files, including those coming from -include or -imacros
11978           command-line options.
11979
11980           Unless specified explicitly (with -MT or -MQ), the object file name
11981           consists of the name of the source file with any suffix replaced
11982           with object file suffix and with any leading directory parts
11983           removed.  If there are many included files then the rule is split
11984           into several lines using \-newline.  The rule has no commands.
11985
11986           This option does not suppress the preprocessor's debug output, such
11987           as -dM.  To avoid mixing such debug output with the dependency
11988           rules you should explicitly specify the dependency output file with
11989           -MF, or use an environment variable like DEPENDENCIES_OUTPUT.
11990           Debug output is still sent to the regular output stream as normal.
11991
11992           Passing -M to the driver implies -E, and suppresses warnings with
11993           an implicit -w.
11994
11995       -MM Like -M but do not mention header files that are found in system
11996           header directories, nor header files that are included, directly or
11997           indirectly, from such a header.
11998
11999           This implies that the choice of angle brackets or double quotes in
12000           an #include directive does not in itself determine whether that
12001           header appears in -MM dependency output.
12002
12003       -MF file
12004           When used with -M or -MM, specifies a file to write the
12005           dependencies to.  If no -MF switch is given the preprocessor sends
12006           the rules to the same place it would send preprocessed output.
12007
12008           When used with the driver options -MD or -MMD, -MF overrides the
12009           default dependency output file.
12010
12011           If file is -, then the dependencies are written to stdout.
12012
12013       -MG In conjunction with an option such as -M requesting dependency
12014           generation, -MG assumes missing header files are generated files
12015           and adds them to the dependency list without raising an error.  The
12016           dependency filename is taken directly from the "#include" directive
12017           without prepending any path.  -MG also suppresses preprocessed
12018           output, as a missing header file renders this useless.
12019
12020           This feature is used in automatic updating of makefiles.
12021
12022       -MP This option instructs CPP to add a phony target for each dependency
12023           other than the main file, causing each to depend on nothing.  These
12024           dummy rules work around errors make gives if you remove header
12025           files without updating the Makefile to match.
12026
12027           This is typical output:
12028
12029                   test.o: test.c test.h
12030
12031                   test.h:
12032
12033       -MT target
12034           Change the target of the rule emitted by dependency generation.  By
12035           default CPP takes the name of the main input file, deletes any
12036           directory components and any file suffix such as .c, and appends
12037           the platform's usual object suffix.  The result is the target.
12038
12039           An -MT option sets the target to be exactly the string you specify.
12040           If you want multiple targets, you can specify them as a single
12041           argument to -MT, or use multiple -MT options.
12042
12043           For example, -MT '$(objpfx)foo.o' might give
12044
12045                   $(objpfx)foo.o: foo.c
12046
12047       -MQ target
12048           Same as -MT, but it quotes any characters which are special to
12049           Make.  -MQ '$(objpfx)foo.o' gives
12050
12051                   $$(objpfx)foo.o: foo.c
12052
12053           The default target is automatically quoted, as if it were given
12054           with -MQ.
12055
12056       -MD -MD is equivalent to -M -MF file, except that -E is not implied.
12057           The driver determines file based on whether an -o option is given.
12058           If it is, the driver uses its argument but with a suffix of .d,
12059           otherwise it takes the name of the input file, removes any
12060           directory components and suffix, and applies a .d suffix.
12061
12062           If -MD is used in conjunction with -E, any -o switch is understood
12063           to specify the dependency output file, but if used without -E, each
12064           -o is understood to specify a target object file.
12065
12066           Since -E is not implied, -MD can be used to generate a dependency
12067           output file as a side effect of the compilation process.
12068
12069       -MMD
12070           Like -MD except mention only user header files, not system header
12071           files.
12072
12073       -fpreprocessed
12074           Indicate to the preprocessor that the input file has already been
12075           preprocessed.  This suppresses things like macro expansion,
12076           trigraph conversion, escaped newline splicing, and processing of
12077           most directives.  The preprocessor still recognizes and removes
12078           comments, so that you can pass a file preprocessed with -C to the
12079           compiler without problems.  In this mode the integrated
12080           preprocessor is little more than a tokenizer for the front ends.
12081
12082           -fpreprocessed is implicit if the input file has one of the
12083           extensions .i, .ii or .mi.  These are the extensions that GCC uses
12084           for preprocessed files created by -save-temps.
12085
12086       -fdirectives-only
12087           When preprocessing, handle directives, but do not expand macros.
12088
12089           The option's behavior depends on the -E and -fpreprocessed options.
12090
12091           With -E, preprocessing is limited to the handling of directives
12092           such as "#define", "#ifdef", and "#error".  Other preprocessor
12093           operations, such as macro expansion and trigraph conversion are not
12094           performed.  In addition, the -dD option is implicitly enabled.
12095
12096           With -fpreprocessed, predefinition of command line and most builtin
12097           macros is disabled.  Macros such as "__LINE__", which are
12098           contextually dependent, are handled normally.  This enables
12099           compilation of files previously preprocessed with "-E
12100           -fdirectives-only".
12101
12102           With both -E and -fpreprocessed, the rules for -fpreprocessed take
12103           precedence.  This enables full preprocessing of files previously
12104           preprocessed with "-E -fdirectives-only".
12105
12106       -fdollars-in-identifiers
12107           Accept $ in identifiers.
12108
12109       -fextended-identifiers
12110           Accept universal character names and extended characters in
12111           identifiers.  This option is enabled by default for C99 (and later
12112           C standard versions) and C++.
12113
12114       -fno-canonical-system-headers
12115           When preprocessing, do not shorten system header paths with
12116           canonicalization.
12117
12118       -fmax-include-depth=depth
12119           Set the maximum depth of the nested #include. The default is 200.
12120
12121       -ftabstop=width
12122           Set the distance between tab stops.  This helps the preprocessor
12123           report correct column numbers in warnings or errors, even if tabs
12124           appear on the line.  If the value is less than 1 or greater than
12125           100, the option is ignored.  The default is 8.
12126
12127       -ftrack-macro-expansion[=level]
12128           Track locations of tokens across macro expansions. This allows the
12129           compiler to emit diagnostic about the current macro expansion stack
12130           when a compilation error occurs in a macro expansion. Using this
12131           option makes the preprocessor and the compiler consume more memory.
12132           The level parameter can be used to choose the level of precision of
12133           token location tracking thus decreasing the memory consumption if
12134           necessary. Value 0 of level de-activates this option. Value 1
12135           tracks tokens locations in a degraded mode for the sake of minimal
12136           memory overhead. In this mode all tokens resulting from the
12137           expansion of an argument of a function-like macro have the same
12138           location. Value 2 tracks tokens locations completely. This value is
12139           the most memory hungry.  When this option is given no argument, the
12140           default parameter value is 2.
12141
12142           Note that "-ftrack-macro-expansion=2" is activated by default.
12143
12144       -fmacro-prefix-map=old=new
12145           When preprocessing files residing in directory old, expand the
12146           "__FILE__" and "__BASE_FILE__" macros as if the files resided in
12147           directory new instead.  This can be used to change an absolute path
12148           to a relative path by using . for new which can result in more
12149           reproducible builds that are location independent.  This option
12150           also affects "__builtin_FILE()" during compilation.  See also
12151           -ffile-prefix-map.
12152
12153       -fexec-charset=charset
12154           Set the execution character set, used for string and character
12155           constants.  The default is UTF-8.  charset can be any encoding
12156           supported by the system's "iconv" library routine.
12157
12158       -fwide-exec-charset=charset
12159           Set the wide execution character set, used for wide string and
12160           character constants.  The default is UTF-32 or UTF-16, whichever
12161           corresponds to the width of "wchar_t".  As with -fexec-charset,
12162           charset can be any encoding supported by the system's "iconv"
12163           library routine; however, you will have problems with encodings
12164           that do not fit exactly in "wchar_t".
12165
12166       -finput-charset=charset
12167           Set the input character set, used for translation from the
12168           character set of the input file to the source character set used by
12169           GCC.  If the locale does not specify, or GCC cannot get this
12170           information from the locale, the default is UTF-8.  This can be
12171           overridden by either the locale or this command-line option.
12172           Currently the command-line option takes precedence if there's a
12173           conflict.  charset can be any encoding supported by the system's
12174           "iconv" library routine.
12175
12176       -fpch-deps
12177           When using precompiled headers, this flag causes the dependency-
12178           output flags to also list the files from the precompiled header's
12179           dependencies.  If not specified, only the precompiled header are
12180           listed and not the files that were used to create it, because those
12181           files are not consulted when a precompiled header is used.
12182
12183       -fpch-preprocess
12184           This option allows use of a precompiled header together with -E.
12185           It inserts a special "#pragma", "#pragma GCC pch_preprocess
12186           "filename"" in the output to mark the place where the precompiled
12187           header was found, and its filename.  When -fpreprocessed is in use,
12188           GCC recognizes this "#pragma" and loads the PCH.
12189
12190           This option is off by default, because the resulting preprocessed
12191           output is only really suitable as input to GCC.  It is switched on
12192           by -save-temps.
12193
12194           You should not write this "#pragma" in your own code, but it is
12195           safe to edit the filename if the PCH file is available in a
12196           different location.  The filename may be absolute or it may be
12197           relative to GCC's current directory.
12198
12199       -fworking-directory
12200           Enable generation of linemarkers in the preprocessor output that
12201           let the compiler know the current working directory at the time of
12202           preprocessing.  When this option is enabled, the preprocessor
12203           emits, after the initial linemarker, a second linemarker with the
12204           current working directory followed by two slashes.  GCC uses this
12205           directory, when it's present in the preprocessed input, as the
12206           directory emitted as the current working directory in some
12207           debugging information formats.  This option is implicitly enabled
12208           if debugging information is enabled, but this can be inhibited with
12209           the negated form -fno-working-directory.  If the -P flag is present
12210           in the command line, this option has no effect, since no "#line"
12211           directives are emitted whatsoever.
12212
12213       -A predicate=answer
12214           Make an assertion with the predicate predicate and answer answer.
12215           This form is preferred to the older form -A predicate(answer),
12216           which is still supported, because it does not use shell special
12217           characters.
12218
12219       -A -predicate=answer
12220           Cancel an assertion with the predicate predicate and answer answer.
12221
12222       -C  Do not discard comments.  All comments are passed through to the
12223           output file, except for comments in processed directives, which are
12224           deleted along with the directive.
12225
12226           You should be prepared for side effects when using -C; it causes
12227           the preprocessor to treat comments as tokens in their own right.
12228           For example, comments appearing at the start of what would be a
12229           directive line have the effect of turning that line into an
12230           ordinary source line, since the first token on the line is no
12231           longer a #.
12232
12233       -CC Do not discard comments, including during macro expansion.  This is
12234           like -C, except that comments contained within macros are also
12235           passed through to the output file where the macro is expanded.
12236
12237           In addition to the side effects of the -C option, the -CC option
12238           causes all C++-style comments inside a macro to be converted to
12239           C-style comments.  This is to prevent later use of that macro from
12240           inadvertently commenting out the remainder of the source line.
12241
12242           The -CC option is generally used to support lint comments.
12243
12244       -P  Inhibit generation of linemarkers in the output from the
12245           preprocessor.  This might be useful when running the preprocessor
12246           on something that is not C code, and will be sent to a program
12247           which might be confused by the linemarkers.
12248
12249       -traditional
12250       -traditional-cpp
12251           Try to imitate the behavior of pre-standard C preprocessors, as
12252           opposed to ISO C preprocessors.  See the GNU CPP manual for
12253           details.
12254
12255           Note that GCC does not otherwise attempt to emulate a pre-standard
12256           C compiler, and these options are only supported with the -E
12257           switch, or when invoking CPP explicitly.
12258
12259       -trigraphs
12260           Support ISO C trigraphs.  These are three-character sequences, all
12261           starting with ??, that are defined by ISO C to stand for single
12262           characters.  For example, ??/ stands for \, so '??/n' is a
12263           character constant for a newline.
12264
12265           The nine trigraphs and their replacements are
12266
12267                   Trigraph:       ??(  ??)  ??<  ??>  ??=  ??/  ??'  ??!  ??-
12268                   Replacement:      [    ]    {    }    #    \    ^    |    ~
12269
12270           By default, GCC ignores trigraphs, but in standard-conforming modes
12271           it converts them.  See the -std and -ansi options.
12272
12273       -remap
12274           Enable special code to work around file systems which only permit
12275           very short file names, such as MS-DOS.
12276
12277       -H  Print the name of each header file used, in addition to other
12278           normal activities.  Each name is indented to show how deep in the
12279           #include stack it is.  Precompiled header files are also printed,
12280           even if they are found to be invalid; an invalid precompiled header
12281           file is printed with ...x and a valid one with ...! .
12282
12283       -dletters
12284           Says to make debugging dumps during compilation as specified by
12285           letters.  The flags documented here are those relevant to the
12286           preprocessor.  Other letters are interpreted by the compiler
12287           proper, or reserved for future versions of GCC, and so are silently
12288           ignored.  If you specify letters whose behavior conflicts, the
12289           result is undefined.
12290
12291           -dM Instead of the normal output, generate a list of #define
12292               directives for all the macros defined during the execution of
12293               the preprocessor, including predefined macros.  This gives you
12294               a way of finding out what is predefined in your version of the
12295               preprocessor.  Assuming you have no file foo.h, the command
12296
12297                       touch foo.h; cpp -dM foo.h
12298
12299               shows all the predefined macros.
12300
12301               If you use -dM without the -E option, -dM is interpreted as a
12302               synonym for -fdump-rtl-mach.
12303
12304           -dD Like -dM except in two respects: it does not include the
12305               predefined macros, and it outputs both the #define directives
12306               and the result of preprocessing.  Both kinds of output go to
12307               the standard output file.
12308
12309           -dN Like -dD, but emit only the macro names, not their expansions.
12310
12311           -dI Output #include directives in addition to the result of
12312               preprocessing.
12313
12314           -dU Like -dD except that only macros that are expanded, or whose
12315               definedness is tested in preprocessor directives, are output;
12316               the output is delayed until the use or test of the macro; and
12317               #undef directives are also output for macros tested but
12318               undefined at the time.
12319
12320       -fdebug-cpp
12321           This option is only useful for debugging GCC.  When used from CPP
12322           or with -E, it dumps debugging information about location maps.
12323           Every token in the output is preceded by the dump of the map its
12324           location belongs to.
12325
12326           When used from GCC without -E, this option has no effect.
12327
12328       -Wp,option
12329           You can use -Wp,option to bypass the compiler driver and pass
12330           option directly through to the preprocessor.  If option contains
12331           commas, it is split into multiple options at the commas.  However,
12332           many options are modified, translated or interpreted by the
12333           compiler driver before being passed to the preprocessor, and -Wp
12334           forcibly bypasses this phase.  The preprocessor's direct interface
12335           is undocumented and subject to change, so whenever possible you
12336           should avoid using -Wp and let the driver handle the options
12337           instead.
12338
12339       -Xpreprocessor option
12340           Pass option as an option to the preprocessor.  You can use this to
12341           supply system-specific preprocessor options that GCC does not
12342           recognize.
12343
12344           If you want to pass an option that takes an argument, you must use
12345           -Xpreprocessor twice, once for the option and once for the
12346           argument.
12347
12348       -no-integrated-cpp
12349           Perform preprocessing as a separate pass before compilation.  By
12350           default, GCC performs preprocessing as an integrated part of input
12351           tokenization and parsing.  If this option is provided, the
12352           appropriate language front end (cc1, cc1plus, or cc1obj for C, C++,
12353           and Objective-C, respectively) is instead invoked twice, once for
12354           preprocessing only and once for actual compilation of the
12355           preprocessed input.  This option may be useful in conjunction with
12356           the -B or -wrapper options to specify an alternate preprocessor or
12357           perform additional processing of the program source between normal
12358           preprocessing and compilation.
12359
12360   Passing Options to the Assembler
12361       You can pass options to the assembler.
12362
12363       -Wa,option
12364           Pass option as an option to the assembler.  If option contains
12365           commas, it is split into multiple options at the commas.
12366
12367       -Xassembler option
12368           Pass option as an option to the assembler.  You can use this to
12369           supply system-specific assembler options that GCC does not
12370           recognize.
12371
12372           If you want to pass an option that takes an argument, you must use
12373           -Xassembler twice, once for the option and once for the argument.
12374
12375   Options for Linking
12376       These options come into play when the compiler links object files into
12377       an executable output file.  They are meaningless if the compiler is not
12378       doing a link step.
12379
12380       object-file-name
12381           A file name that does not end in a special recognized suffix is
12382           considered to name an object file or library.  (Object files are
12383           distinguished from libraries by the linker according to the file
12384           contents.)  If linking is done, these object files are used as
12385           input to the linker.
12386
12387       -c
12388       -S
12389       -E  If any of these options is used, then the linker is not run, and
12390           object file names should not be used as arguments.
12391
12392       -flinker-output=type
12393           This option controls code generation of the link-time optimizer.
12394           By default the linker output is automatically determined by the
12395           linker plugin.  For debugging the compiler and if incremental
12396           linking with a non-LTO object file is desired, it may be useful to
12397           control the type manually.
12398
12399           If type is exec, code generation produces a static binary. In this
12400           case -fpic and -fpie are both disabled.
12401
12402           If type is dyn, code generation produces a shared library.  In this
12403           case -fpic or -fPIC is preserved, but not enabled automatically.
12404           This allows to build shared libraries without position-independent
12405           code on architectures where this is possible, i.e. on x86.
12406
12407           If type is pie, code generation produces an -fpie executable. This
12408           results in similar optimizations as exec except that -fpie is not
12409           disabled if specified at compilation time.
12410
12411           If type is rel, the compiler assumes that incremental linking is
12412           done.  The sections containing intermediate code for link-time
12413           optimization are merged, pre-optimized, and output to the resulting
12414           object file. In addition, if -ffat-lto-objects is specified, binary
12415           code is produced for future non-LTO linking. The object file
12416           produced by incremental linking is smaller than a static library
12417           produced from the same object files.  At link time the result of
12418           incremental linking also loads faster than a static library
12419           assuming that the majority of objects in the library are used.
12420
12421           Finally nolto-rel configures the compiler for incremental linking
12422           where code generation is forced, a final binary is produced, and
12423           the intermediate code for later link-time optimization is stripped.
12424           When multiple object files are linked together the resulting code
12425           is better optimized than with link-time optimizations disabled (for
12426           example, cross-module inlining happens), but most of benefits of
12427           whole program optimizations are lost.
12428
12429           During the incremental link (by -r) the linker plugin defaults to
12430           rel. With current interfaces to GNU Binutils it is however not
12431           possible to incrementally link LTO objects and non-LTO objects into
12432           a single mixed object file.  If any of object files in incremental
12433           link cannot be used for link-time optimization, the linker plugin
12434           issues a warning and uses nolto-rel. To maintain whole program
12435           optimization, it is recommended to link such objects into static
12436           library instead. Alternatively it is possible to use H.J. Lu's
12437           binutils with support for mixed objects.
12438
12439       -fuse-ld=bfd
12440           Use the bfd linker instead of the default linker.
12441
12442       -fuse-ld=gold
12443           Use the gold linker instead of the default linker.
12444
12445       -fuse-ld=lld
12446           Use the LLVM lld linker instead of the default linker.
12447
12448       -llibrary
12449       -l library
12450           Search the library named library when linking.  (The second
12451           alternative with the library as a separate argument is only for
12452           POSIX compliance and is not recommended.)
12453
12454           The -l option is passed directly to the linker by GCC.  Refer to
12455           your linker documentation for exact details.  The general
12456           description below applies to the GNU linker.
12457
12458           The linker searches a standard list of directories for the library.
12459           The directories searched include several standard system
12460           directories plus any that you specify with -L.
12461
12462           Static libraries are archives of object files, and have file names
12463           like liblibrary.a.  Some targets also support shared libraries,
12464           which typically have names like liblibrary.so.  If both static and
12465           shared libraries are found, the linker gives preference to linking
12466           with the shared library unless the -static option is used.
12467
12468           It makes a difference where in the command you write this option;
12469           the linker searches and processes libraries and object files in the
12470           order they are specified.  Thus, foo.o -lz bar.o searches library z
12471           after file foo.o but before bar.o.  If bar.o refers to functions in
12472           z, those functions may not be loaded.
12473
12474       -lobjc
12475           You need this special case of the -l option in order to link an
12476           Objective-C or Objective-C++ program.
12477
12478       -nostartfiles
12479           Do not use the standard system startup files when linking.  The
12480           standard system libraries are used normally, unless -nostdlib,
12481           -nolibc, or -nodefaultlibs is used.
12482
12483       -nodefaultlibs
12484           Do not use the standard system libraries when linking.  Only the
12485           libraries you specify are passed to the linker, and options
12486           specifying linkage of the system libraries, such as -static-libgcc
12487           or -shared-libgcc, are ignored.  The standard startup files are
12488           used normally, unless -nostartfiles is used.
12489
12490           The compiler may generate calls to "memcmp", "memset", "memcpy" and
12491           "memmove".  These entries are usually resolved by entries in libc.
12492           These entry points should be supplied through some other mechanism
12493           when this option is specified.
12494
12495       -nolibc
12496           Do not use the C library or system libraries tightly coupled with
12497           it when linking.  Still link with the startup files, libgcc or
12498           toolchain provided language support libraries such as libgnat,
12499           libgfortran or libstdc++ unless options preventing their inclusion
12500           are used as well.  This typically removes -lc from the link command
12501           line, as well as system libraries that normally go with it and
12502           become meaningless when absence of a C library is assumed, for
12503           example -lpthread or -lm in some configurations.  This is intended
12504           for bare-board targets when there is indeed no C library available.
12505
12506       -nostdlib
12507           Do not use the standard system startup files or libraries when
12508           linking.  No startup files and only the libraries you specify are
12509           passed to the linker, and options specifying linkage of the system
12510           libraries, such as -static-libgcc or -shared-libgcc, are ignored.
12511
12512           The compiler may generate calls to "memcmp", "memset", "memcpy" and
12513           "memmove".  These entries are usually resolved by entries in libc.
12514           These entry points should be supplied through some other mechanism
12515           when this option is specified.
12516
12517           One of the standard libraries bypassed by -nostdlib and
12518           -nodefaultlibs is libgcc.a, a library of internal subroutines which
12519           GCC uses to overcome shortcomings of particular machines, or
12520           special needs for some languages.
12521
12522           In most cases, you need libgcc.a even when you want to avoid other
12523           standard libraries.  In other words, when you specify -nostdlib or
12524           -nodefaultlibs you should usually specify -lgcc as well.  This
12525           ensures that you have no unresolved references to internal GCC
12526           library subroutines.  (An example of such an internal subroutine is
12527           "__main", used to ensure C++ constructors are called.)
12528
12529       -e entry
12530       --entry=entry
12531           Specify that the program entry point is entry.  The argument is
12532           interpreted by the linker; the GNU linker accepts either a symbol
12533           name or an address.
12534
12535       -pie
12536           Produce a dynamically linked position independent executable on
12537           targets that support it.  For predictable results, you must also
12538           specify the same set of options used for compilation (-fpie, -fPIE,
12539           or model suboptions) when you specify this linker option.
12540
12541       -no-pie
12542           Don't produce a dynamically linked position independent executable.
12543
12544       -static-pie
12545           Produce a static position independent executable on targets that
12546           support it.  A static position independent executable is similar to
12547           a static executable, but can be loaded at any address without a
12548           dynamic linker.  For predictable results, you must also specify the
12549           same set of options used for compilation (-fpie, -fPIE, or model
12550           suboptions) when you specify this linker option.
12551
12552       -pthread
12553           Link with the POSIX threads library.  This option is supported on
12554           GNU/Linux targets, most other Unix derivatives, and also on x86
12555           Cygwin and MinGW targets.  On some targets this option also sets
12556           flags for the preprocessor, so it should be used consistently for
12557           both compilation and linking.
12558
12559       -r  Produce a relocatable object as output.  This is also known as
12560           partial linking.
12561
12562       -rdynamic
12563           Pass the flag -export-dynamic to the ELF linker, on targets that
12564           support it. This instructs the linker to add all symbols, not only
12565           used ones, to the dynamic symbol table. This option is needed for
12566           some uses of "dlopen" or to allow obtaining backtraces from within
12567           a program.
12568
12569       -s  Remove all symbol table and relocation information from the
12570           executable.
12571
12572       -static
12573           On systems that support dynamic linking, this overrides -pie and
12574           prevents linking with the shared libraries.  On other systems, this
12575           option has no effect.
12576
12577       -shared
12578           Produce a shared object which can then be linked with other objects
12579           to form an executable.  Not all systems support this option.  For
12580           predictable results, you must also specify the same set of options
12581           used for compilation (-fpic, -fPIC, or model suboptions) when you
12582           specify this linker option.[1]
12583
12584       -shared-libgcc
12585       -static-libgcc
12586           On systems that provide libgcc as a shared library, these options
12587           force the use of either the shared or static version, respectively.
12588           If no shared version of libgcc was built when the compiler was
12589           configured, these options have no effect.
12590
12591           There are several situations in which an application should use the
12592           shared libgcc instead of the static version.  The most common of
12593           these is when the application wishes to throw and catch exceptions
12594           across different shared libraries.  In that case, each of the
12595           libraries as well as the application itself should use the shared
12596           libgcc.
12597
12598           Therefore, the G++ driver automatically adds -shared-libgcc
12599           whenever you build a shared library or a main executable, because
12600           C++ programs typically use exceptions, so this is the right thing
12601           to do.
12602
12603           If, instead, you use the GCC driver to create shared libraries, you
12604           may find that they are not always linked with the shared libgcc.
12605           If GCC finds, at its configuration time, that you have a non-GNU
12606           linker or a GNU linker that does not support option --eh-frame-hdr,
12607           it links the shared version of libgcc into shared libraries by
12608           default.  Otherwise, it takes advantage of the linker and optimizes
12609           away the linking with the shared version of libgcc, linking with
12610           the static version of libgcc by default.  This allows exceptions to
12611           propagate through such shared libraries, without incurring
12612           relocation costs at library load time.
12613
12614           However, if a library or main executable is supposed to throw or
12615           catch exceptions, you must link it using the G++ driver, or using
12616           the option -shared-libgcc, such that it is linked with the shared
12617           libgcc.
12618
12619       -static-libasan
12620           When the -fsanitize=address option is used to link a program, the
12621           GCC driver automatically links against libasan.  If libasan is
12622           available as a shared library, and the -static option is not used,
12623           then this links against the shared version of libasan.  The
12624           -static-libasan option directs the GCC driver to link libasan
12625           statically, without necessarily linking other libraries statically.
12626
12627       -static-libtsan
12628           When the -fsanitize=thread option is used to link a program, the
12629           GCC driver automatically links against libtsan.  If libtsan is
12630           available as a shared library, and the -static option is not used,
12631           then this links against the shared version of libtsan.  The
12632           -static-libtsan option directs the GCC driver to link libtsan
12633           statically, without necessarily linking other libraries statically.
12634
12635       -static-liblsan
12636           When the -fsanitize=leak option is used to link a program, the GCC
12637           driver automatically links against liblsan.  If liblsan is
12638           available as a shared library, and the -static option is not used,
12639           then this links against the shared version of liblsan.  The
12640           -static-liblsan option directs the GCC driver to link liblsan
12641           statically, without necessarily linking other libraries statically.
12642
12643       -static-libubsan
12644           When the -fsanitize=undefined option is used to link a program, the
12645           GCC driver automatically links against libubsan.  If libubsan is
12646           available as a shared library, and the -static option is not used,
12647           then this links against the shared version of libubsan.  The
12648           -static-libubsan option directs the GCC driver to link libubsan
12649           statically, without necessarily linking other libraries statically.
12650
12651       -static-libstdc++
12652           When the g++ program is used to link a C++ program, it normally
12653           automatically links against libstdc++.  If libstdc++ is available
12654           as a shared library, and the -static option is not used, then this
12655           links against the shared version of libstdc++.  That is normally
12656           fine.  However, it is sometimes useful to freeze the version of
12657           libstdc++ used by the program without going all the way to a fully
12658           static link.  The -static-libstdc++ option directs the g++ driver
12659           to link libstdc++ statically, without necessarily linking other
12660           libraries statically.
12661
12662       -symbolic
12663           Bind references to global symbols when building a shared object.
12664           Warn about any unresolved references (unless overridden by the link
12665           editor option -Xlinker -z -Xlinker defs).  Only a few systems
12666           support this option.
12667
12668       -T script
12669           Use script as the linker script.  This option is supported by most
12670           systems using the GNU linker.  On some targets, such as bare-board
12671           targets without an operating system, the -T option may be required
12672           when linking to avoid references to undefined symbols.
12673
12674       -Xlinker option
12675           Pass option as an option to the linker.  You can use this to supply
12676           system-specific linker options that GCC does not recognize.
12677
12678           If you want to pass an option that takes a separate argument, you
12679           must use -Xlinker twice, once for the option and once for the
12680           argument.  For example, to pass -assert definitions, you must write
12681           -Xlinker -assert -Xlinker definitions.  It does not work to write
12682           -Xlinker "-assert definitions", because this passes the entire
12683           string as a single argument, which is not what the linker expects.
12684
12685           When using the GNU linker, it is usually more convenient to pass
12686           arguments to linker options using the option=value syntax than as
12687           separate arguments.  For example, you can specify -Xlinker
12688           -Map=output.map rather than -Xlinker -Map -Xlinker output.map.
12689           Other linkers may not support this syntax for command-line options.
12690
12691       -Wl,option
12692           Pass option as an option to the linker.  If option contains commas,
12693           it is split into multiple options at the commas.  You can use this
12694           syntax to pass an argument to the option.  For example,
12695           -Wl,-Map,output.map passes -Map output.map to the linker.  When
12696           using the GNU linker, you can also get the same effect with
12697           -Wl,-Map=output.map.
12698
12699       -u symbol
12700           Pretend the symbol symbol is undefined, to force linking of library
12701           modules to define it.  You can use -u multiple times with different
12702           symbols to force loading of additional library modules.
12703
12704       -z keyword
12705           -z is passed directly on to the linker along with the keyword
12706           keyword. See the section in the documentation of your linker for
12707           permitted values and their meanings.
12708
12709   Options for Directory Search
12710       These options specify directories to search for header files, for
12711       libraries and for parts of the compiler:
12712
12713       -I dir
12714       -iquote dir
12715       -isystem dir
12716       -idirafter dir
12717           Add the directory dir to the list of directories to be searched for
12718           header files during preprocessing.  If dir begins with = or
12719           $SYSROOT, then the = or $SYSROOT is replaced by the sysroot prefix;
12720           see --sysroot and -isysroot.
12721
12722           Directories specified with -iquote apply only to the quote form of
12723           the directive, "#include "file"".  Directories specified with -I,
12724           -isystem, or -idirafter apply to lookup for both the
12725           "#include "file"" and "#include <file>" directives.
12726
12727           You can specify any number or combination of these options on the
12728           command line to search for header files in several directories.
12729           The lookup order is as follows:
12730
12731           1.  For the quote form of the include directive, the directory of
12732               the current file is searched first.
12733
12734           2.  For the quote form of the include directive, the directories
12735               specified by -iquote options are searched in left-to-right
12736               order, as they appear on the command line.
12737
12738           3.  Directories specified with -I options are scanned in left-to-
12739               right order.
12740
12741           4.  Directories specified with -isystem options are scanned in
12742               left-to-right order.
12743
12744           5.  Standard system directories are scanned.
12745
12746           6.  Directories specified with -idirafter options are scanned in
12747               left-to-right order.
12748
12749           You can use -I to override a system header file, substituting your
12750           own version, since these directories are searched before the
12751           standard system header file directories.  However, you should not
12752           use this option to add directories that contain vendor-supplied
12753           system header files; use -isystem for that.
12754
12755           The -isystem and -idirafter options also mark the directory as a
12756           system directory, so that it gets the same special treatment that
12757           is applied to the standard system directories.
12758
12759           If a standard system include directory, or a directory specified
12760           with -isystem, is also specified with -I, the -I option is ignored.
12761           The directory is still searched but as a system directory at its
12762           normal position in the system include chain.  This is to ensure
12763           that GCC's procedure to fix buggy system headers and the ordering
12764           for the "#include_next" directive are not inadvertently changed.
12765           If you really need to change the search order for system
12766           directories, use the -nostdinc and/or -isystem options.
12767
12768       -I- Split the include path.  This option has been deprecated.  Please
12769           use -iquote instead for -I directories before the -I- and remove
12770           the -I- option.
12771
12772           Any directories specified with -I options before -I- are searched
12773           only for headers requested with "#include "file""; they are not
12774           searched for "#include <file>".  If additional directories are
12775           specified with -I options after the -I-, those directories are
12776           searched for all #include directives.
12777
12778           In addition, -I- inhibits the use of the directory of the current
12779           file directory as the first search directory for "#include "file"".
12780           There is no way to override this effect of -I-.
12781
12782       -iprefix prefix
12783           Specify prefix as the prefix for subsequent -iwithprefix options.
12784           If the prefix represents a directory, you should include the final
12785           /.
12786
12787       -iwithprefix dir
12788       -iwithprefixbefore dir
12789           Append dir to the prefix specified previously with -iprefix, and
12790           add the resulting directory to the include search path.
12791           -iwithprefixbefore puts it in the same place -I would; -iwithprefix
12792           puts it where -idirafter would.
12793
12794       -isysroot dir
12795           This option is like the --sysroot option, but applies only to
12796           header files (except for Darwin targets, where it applies to both
12797           header files and libraries).  See the --sysroot option for more
12798           information.
12799
12800       -imultilib dir
12801           Use dir as a subdirectory of the directory containing target-
12802           specific C++ headers.
12803
12804       -nostdinc
12805           Do not search the standard system directories for header files.
12806           Only the directories explicitly specified with -I, -iquote,
12807           -isystem, and/or -idirafter options (and the directory of the
12808           current file, if appropriate) are searched.
12809
12810       -nostdinc++
12811           Do not search for header files in the C++-specific standard
12812           directories, but do still search the other standard directories.
12813           (This option is used when building the C++ library.)
12814
12815       -iplugindir=dir
12816           Set the directory to search for plugins that are passed by
12817           -fplugin=name instead of -fplugin=path/name.so.  This option is not
12818           meant to be used by the user, but only passed by the driver.
12819
12820       -Ldir
12821           Add directory dir to the list of directories to be searched for -l.
12822
12823       -Bprefix
12824           This option specifies where to find the executables, libraries,
12825           include files, and data files of the compiler itself.
12826
12827           The compiler driver program runs one or more of the subprograms
12828           cpp, cc1, as and ld.  It tries prefix as a prefix for each program
12829           it tries to run, both with and without machine/version/ for the
12830           corresponding target machine and compiler version.
12831
12832           For each subprogram to be run, the compiler driver first tries the
12833           -B prefix, if any.  If that name is not found, or if -B is not
12834           specified, the driver tries two standard prefixes, /usr/lib/gcc/
12835           and /usr/local/lib/gcc/.  If neither of those results in a file
12836           name that is found, the unmodified program name is searched for
12837           using the directories specified in your PATH environment variable.
12838
12839           The compiler checks to see if the path provided by -B refers to a
12840           directory, and if necessary it adds a directory separator character
12841           at the end of the path.
12842
12843           -B prefixes that effectively specify directory names also apply to
12844           libraries in the linker, because the compiler translates these
12845           options into -L options for the linker.  They also apply to include
12846           files in the preprocessor, because the compiler translates these
12847           options into -isystem options for the preprocessor.  In this case,
12848           the compiler appends include to the prefix.
12849
12850           The runtime support file libgcc.a can also be searched for using
12851           the -B prefix, if needed.  If it is not found there, the two
12852           standard prefixes above are tried, and that is all.  The file is
12853           left out of the link if it is not found by those means.
12854
12855           Another way to specify a prefix much like the -B prefix is to use
12856           the environment variable GCC_EXEC_PREFIX.
12857
12858           As a special kludge, if the path provided by -B is [dir/]stageN/,
12859           where N is a number in the range 0 to 9, then it is replaced by
12860           [dir/]include.  This is to help with boot-strapping the compiler.
12861
12862       -no-canonical-prefixes
12863           Do not expand any symbolic links, resolve references to /../ or
12864           /./, or make the path absolute when generating a relative prefix.
12865
12866       --sysroot=dir
12867           Use dir as the logical root directory for headers and libraries.
12868           For example, if the compiler normally searches for headers in
12869           /usr/include and libraries in /usr/lib, it instead searches
12870           dir/usr/include and dir/usr/lib.
12871
12872           If you use both this option and the -isysroot option, then the
12873           --sysroot option applies to libraries, but the -isysroot option
12874           applies to header files.
12875
12876           The GNU linker (beginning with version 2.16) has the necessary
12877           support for this option.  If your linker does not support this
12878           option, the header file aspect of --sysroot still works, but the
12879           library aspect does not.
12880
12881       --no-sysroot-suffix
12882           For some targets, a suffix is added to the root directory specified
12883           with --sysroot, depending on the other options used, so that
12884           headers may for example be found in dir/suffix/usr/include instead
12885           of dir/usr/include.  This option disables the addition of such a
12886           suffix.
12887
12888   Options for Code Generation Conventions
12889       These machine-independent options control the interface conventions
12890       used in code generation.
12891
12892       Most of them have both positive and negative forms; the negative form
12893       of -ffoo is -fno-foo.  In the table below, only one of the forms is
12894       listed---the one that is not the default.  You can figure out the other
12895       form by either removing no- or adding it.
12896
12897       -fstack-reuse=reuse-level
12898           This option controls stack space reuse for user declared local/auto
12899           variables and compiler generated temporaries.  reuse_level can be
12900           all, named_vars, or none. all enables stack reuse for all local
12901           variables and temporaries, named_vars enables the reuse only for
12902           user defined local variables with names, and none disables stack
12903           reuse completely. The default value is all. The option is needed
12904           when the program extends the lifetime of a scoped local variable or
12905           a compiler generated temporary beyond the end point defined by the
12906           language.  When a lifetime of a variable ends, and if the variable
12907           lives in memory, the optimizing compiler has the freedom to reuse
12908           its stack space with other temporaries or scoped local variables
12909           whose live range does not overlap with it. Legacy code extending
12910           local lifetime is likely to break with the stack reuse
12911           optimization.
12912
12913           For example,
12914
12915                      int *p;
12916                      {
12917                        int local1;
12918
12919                        p = &local1;
12920                        local1 = 10;
12921                        ....
12922                      }
12923                      {
12924                         int local2;
12925                         local2 = 20;
12926                         ...
12927                      }
12928
12929                      if (*p == 10)  // out of scope use of local1
12930                        {
12931
12932                        }
12933
12934           Another example:
12935
12936                      struct A
12937                      {
12938                          A(int k) : i(k), j(k) { }
12939                          int i;
12940                          int j;
12941                      };
12942
12943                      A *ap;
12944
12945                      void foo(const A& ar)
12946                      {
12947                         ap = &ar;
12948                      }
12949
12950                      void bar()
12951                      {
12952                         foo(A(10)); // temp object's lifetime ends when foo returns
12953
12954                         {
12955                           A a(20);
12956                           ....
12957                         }
12958                         ap->i+= 10;  // ap references out of scope temp whose space
12959                                      // is reused with a. What is the value of ap->i?
12960                      }
12961
12962           The lifetime of a compiler generated temporary is well defined by
12963           the C++ standard. When a lifetime of a temporary ends, and if the
12964           temporary lives in memory, the optimizing compiler has the freedom
12965           to reuse its stack space with other temporaries or scoped local
12966           variables whose live range does not overlap with it. However some
12967           of the legacy code relies on the behavior of older compilers in
12968           which temporaries' stack space is not reused, the aggressive stack
12969           reuse can lead to runtime errors. This option is used to control
12970           the temporary stack reuse optimization.
12971
12972       -ftrapv
12973           This option generates traps for signed overflow on addition,
12974           subtraction, multiplication operations.  The options -ftrapv and
12975           -fwrapv override each other, so using -ftrapv -fwrapv on the
12976           command-line results in -fwrapv being effective.  Note that only
12977           active options override, so using -ftrapv -fwrapv -fno-wrapv on the
12978           command-line results in -ftrapv being effective.
12979
12980       -fwrapv
12981           This option instructs the compiler to assume that signed arithmetic
12982           overflow of addition, subtraction and multiplication wraps around
12983           using twos-complement representation.  This flag enables some
12984           optimizations and disables others.  The options -ftrapv and -fwrapv
12985           override each other, so using -ftrapv -fwrapv on the command-line
12986           results in -fwrapv being effective.  Note that only active options
12987           override, so using -ftrapv -fwrapv -fno-wrapv on the command-line
12988           results in -ftrapv being effective.
12989
12990       -fwrapv-pointer
12991           This option instructs the compiler to assume that pointer
12992           arithmetic overflow on addition and subtraction wraps around using
12993           twos-complement representation.  This flag disables some
12994           optimizations which assume pointer overflow is invalid.
12995
12996       -fstrict-overflow
12997           This option implies -fno-wrapv -fno-wrapv-pointer and when negated
12998           implies -fwrapv -fwrapv-pointer.
12999
13000       -fexceptions
13001           Enable exception handling.  Generates extra code needed to
13002           propagate exceptions.  For some targets, this implies GCC generates
13003           frame unwind information for all functions, which can produce
13004           significant data size overhead, although it does not affect
13005           execution.  If you do not specify this option, GCC enables it by
13006           default for languages like C++ that normally require exception
13007           handling, and disables it for languages like C that do not normally
13008           require it.  However, you may need to enable this option when
13009           compiling C code that needs to interoperate properly with exception
13010           handlers written in C++.  You may also wish to disable this option
13011           if you are compiling older C++ programs that don't use exception
13012           handling.
13013
13014       -fnon-call-exceptions
13015           Generate code that allows trapping instructions to throw
13016           exceptions.  Note that this requires platform-specific runtime
13017           support that does not exist everywhere.  Moreover, it only allows
13018           trapping instructions to throw exceptions, i.e. memory references
13019           or floating-point instructions.  It does not allow exceptions to be
13020           thrown from arbitrary signal handlers such as "SIGALRM".
13021
13022       -fdelete-dead-exceptions
13023           Consider that instructions that may throw exceptions but don't
13024           otherwise contribute to the execution of the program can be
13025           optimized away.  This option is enabled by default for the Ada
13026           front end, as permitted by the Ada language specification.
13027           Optimization passes that cause dead exceptions to be removed are
13028           enabled independently at different optimization levels.
13029
13030       -funwind-tables
13031           Similar to -fexceptions, except that it just generates any needed
13032           static data, but does not affect the generated code in any other
13033           way.  You normally do not need to enable this option; instead, a
13034           language processor that needs this handling enables it on your
13035           behalf.
13036
13037       -fasynchronous-unwind-tables
13038           Generate unwind table in DWARF format, if supported by target
13039           machine.  The table is exact at each instruction boundary, so it
13040           can be used for stack unwinding from asynchronous events (such as
13041           debugger or garbage collector).
13042
13043       -fno-gnu-unique
13044           On systems with recent GNU assembler and C library, the C++
13045           compiler uses the "STB_GNU_UNIQUE" binding to make sure that
13046           definitions of template static data members and static local
13047           variables in inline functions are unique even in the presence of
13048           "RTLD_LOCAL"; this is necessary to avoid problems with a library
13049           used by two different "RTLD_LOCAL" plugins depending on a
13050           definition in one of them and therefore disagreeing with the other
13051           one about the binding of the symbol.  But this causes "dlclose" to
13052           be ignored for affected DSOs; if your program relies on
13053           reinitialization of a DSO via "dlclose" and "dlopen", you can use
13054           -fno-gnu-unique.
13055
13056       -fpcc-struct-return
13057           Return "short" "struct" and "union" values in memory like longer
13058           ones, rather than in registers.  This convention is less efficient,
13059           but it has the advantage of allowing intercallability between GCC-
13060           compiled files and files compiled with other compilers,
13061           particularly the Portable C Compiler (pcc).
13062
13063           The precise convention for returning structures in memory depends
13064           on the target configuration macros.
13065
13066           Short structures and unions are those whose size and alignment
13067           match that of some integer type.
13068
13069           Warning: code compiled with the -fpcc-struct-return switch is not
13070           binary compatible with code compiled with the -freg-struct-return
13071           switch.  Use it to conform to a non-default application binary
13072           interface.
13073
13074       -freg-struct-return
13075           Return "struct" and "union" values in registers when possible.
13076           This is more efficient for small structures than
13077           -fpcc-struct-return.
13078
13079           If you specify neither -fpcc-struct-return nor -freg-struct-return,
13080           GCC defaults to whichever convention is standard for the target.
13081           If there is no standard convention, GCC defaults to
13082           -fpcc-struct-return, except on targets where GCC is the principal
13083           compiler.  In those cases, we can choose the standard, and we chose
13084           the more efficient register return alternative.
13085
13086           Warning: code compiled with the -freg-struct-return switch is not
13087           binary compatible with code compiled with the -fpcc-struct-return
13088           switch.  Use it to conform to a non-default application binary
13089           interface.
13090
13091       -fshort-enums
13092           Allocate to an "enum" type only as many bytes as it needs for the
13093           declared range of possible values.  Specifically, the "enum" type
13094           is equivalent to the smallest integer type that has enough room.
13095
13096           Warning: the -fshort-enums switch causes GCC to generate code that
13097           is not binary compatible with code generated without that switch.
13098           Use it to conform to a non-default application binary interface.
13099
13100       -fshort-wchar
13101           Override the underlying type for "wchar_t" to be "short unsigned
13102           int" instead of the default for the target.  This option is useful
13103           for building programs to run under WINE.
13104
13105           Warning: the -fshort-wchar switch causes GCC to generate code that
13106           is not binary compatible with code generated without that switch.
13107           Use it to conform to a non-default application binary interface.
13108
13109       -fcommon
13110           In C code, this option controls the placement of global variables
13111           defined without an initializer, known as tentative definitions in
13112           the C standard.  Tentative definitions are distinct from
13113           declarations of a variable with the "extern" keyword, which do not
13114           allocate storage.
13115
13116           The default is -fno-common, which specifies that the compiler
13117           places uninitialized global variables in the BSS section of the
13118           object file.  This inhibits the merging of tentative definitions by
13119           the linker so you get a multiple-definition error if the same
13120           variable is accidentally defined in more than one compilation unit.
13121
13122           The -fcommon places uninitialized global variables in a common
13123           block.  This allows the linker to resolve all tentative definitions
13124           of the same variable in different compilation units to the same
13125           object, or to a non-tentative definition.  This behavior is
13126           inconsistent with C++, and on many targets implies a speed and code
13127           size penalty on global variable references.  It is mainly useful to
13128           enable legacy code to link without errors.
13129
13130       -fno-ident
13131           Ignore the "#ident" directive.
13132
13133       -finhibit-size-directive
13134           Don't output a ".size" assembler directive, or anything else that
13135           would cause trouble if the function is split in the middle, and the
13136           two halves are placed at locations far apart in memory.  This
13137           option is used when compiling crtstuff.c; you should not need to
13138           use it for anything else.
13139
13140       -fverbose-asm
13141           Put extra commentary information in the generated assembly code to
13142           make it more readable.  This option is generally only of use to
13143           those who actually need to read the generated assembly code
13144           (perhaps while debugging the compiler itself).
13145
13146           -fno-verbose-asm, the default, causes the extra information to be
13147           omitted and is useful when comparing two assembler files.
13148
13149           The added comments include:
13150
13151           *   information on the compiler version and command-line options,
13152
13153           *   the source code lines associated with the assembly
13154               instructions, in the form FILENAME:LINENUMBER:CONTENT OF LINE,
13155
13156           *   hints on which high-level expressions correspond to the various
13157               assembly instruction operands.
13158
13159           For example, given this C source file:
13160
13161                   int test (int n)
13162                   {
13163                     int i;
13164                     int total = 0;
13165
13166                     for (i = 0; i < n; i++)
13167                       total += i * i;
13168
13169                     return total;
13170                   }
13171
13172           compiling to (x86_64) assembly via -S and emitting the result
13173           direct to stdout via -o -
13174
13175                   gcc -S test.c -fverbose-asm -Os -o -
13176
13177           gives output similar to this:
13178
13179                           .file   "test.c"
13180                   # GNU C11 (GCC) version 7.0.0 20160809 (experimental) (x86_64-pc-linux-gnu)
13181                     [...snip...]
13182                   # options passed:
13183                     [...snip...]
13184
13185                           .text
13186                           .globl  test
13187                           .type   test, @function
13188                   test:
13189                   .LFB0:
13190                           .cfi_startproc
13191                   # test.c:4:   int total = 0;
13192                           xorl    %eax, %eax      # <retval>
13193                   # test.c:6:   for (i = 0; i < n; i++)
13194                           xorl    %edx, %edx      # i
13195                   .L2:
13196                   # test.c:6:   for (i = 0; i < n; i++)
13197                           cmpl    %edi, %edx      # n, i
13198                           jge     .L5     #,
13199                   # test.c:7:     total += i * i;
13200                           movl    %edx, %ecx      # i, tmp92
13201                           imull   %edx, %ecx      # i, tmp92
13202                   # test.c:6:   for (i = 0; i < n; i++)
13203                           incl    %edx    # i
13204                   # test.c:7:     total += i * i;
13205                           addl    %ecx, %eax      # tmp92, <retval>
13206                           jmp     .L2     #
13207                   .L5:
13208                   # test.c:10: }
13209                           ret
13210                           .cfi_endproc
13211                   .LFE0:
13212                           .size   test, .-test
13213                           .ident  "GCC: (GNU) 7.0.0 20160809 (experimental)"
13214                           .section        .note.GNU-stack,"",@progbits
13215
13216           The comments are intended for humans rather than machines and hence
13217           the precise format of the comments is subject to change.
13218
13219       -frecord-gcc-switches
13220           This switch causes the command line used to invoke the compiler to
13221           be recorded into the object file that is being created.  This
13222           switch is only implemented on some targets and the exact format of
13223           the recording is target and binary file format dependent, but it
13224           usually takes the form of a section containing ASCII text.  This
13225           switch is related to the -fverbose-asm switch, but that switch only
13226           records information in the assembler output file as comments, so it
13227           never reaches the object file.  See also -grecord-gcc-switches for
13228           another way of storing compiler options into the object file.
13229
13230       -fpic
13231           Generate position-independent code (PIC) suitable for use in a
13232           shared library, if supported for the target machine.  Such code
13233           accesses all constant addresses through a global offset table
13234           (GOT).  The dynamic loader resolves the GOT entries when the
13235           program starts (the dynamic loader is not part of GCC; it is part
13236           of the operating system).  If the GOT size for the linked
13237           executable exceeds a machine-specific maximum size, you get an
13238           error message from the linker indicating that -fpic does not work;
13239           in that case, recompile with -fPIC instead.  (These maximums are 8k
13240           on the SPARC, 28k on AArch64 and 32k on the m68k and RS/6000.  The
13241           x86 has no such limit.)
13242
13243           Position-independent code requires special support, and therefore
13244           works only on certain machines.  For the x86, GCC supports PIC for
13245           System V but not for the Sun 386i.  Code generated for the IBM
13246           RS/6000 is always position-independent.
13247
13248           When this flag is set, the macros "__pic__" and "__PIC__" are
13249           defined to 1.
13250
13251       -fPIC
13252           If supported for the target machine, emit position-independent
13253           code, suitable for dynamic linking and avoiding any limit on the
13254           size of the global offset table.  This option makes a difference on
13255           AArch64, m68k, PowerPC and SPARC.
13256
13257           Position-independent code requires special support, and therefore
13258           works only on certain machines.
13259
13260           When this flag is set, the macros "__pic__" and "__PIC__" are
13261           defined to 2.
13262
13263       -fpie
13264       -fPIE
13265           These options are similar to -fpic and -fPIC, but the generated
13266           position-independent code can be only linked into executables.
13267           Usually these options are used to compile code that will be linked
13268           using the -pie GCC option.
13269
13270           -fpie and -fPIE both define the macros "__pie__" and "__PIE__".
13271           The macros have the value 1 for -fpie and 2 for -fPIE.
13272
13273       -fno-plt
13274           Do not use the PLT for external function calls in position-
13275           independent code.  Instead, load the callee address at call sites
13276           from the GOT and branch to it.  This leads to more efficient code
13277           by eliminating PLT stubs and exposing GOT loads to optimizations.
13278           On architectures such as 32-bit x86 where PLT stubs expect the GOT
13279           pointer in a specific register, this gives more register allocation
13280           freedom to the compiler.  Lazy binding requires use of the PLT;
13281           with -fno-plt all external symbols are resolved at load time.
13282
13283           Alternatively, the function attribute "noplt" can be used to avoid
13284           calls through the PLT for specific external functions.
13285
13286           In position-dependent code, a few targets also convert calls to
13287           functions that are marked to not use the PLT to use the GOT
13288           instead.
13289
13290       -fno-jump-tables
13291           Do not use jump tables for switch statements even where it would be
13292           more efficient than other code generation strategies.  This option
13293           is of use in conjunction with -fpic or -fPIC for building code that
13294           forms part of a dynamic linker and cannot reference the address of
13295           a jump table.  On some targets, jump tables do not require a GOT
13296           and this option is not needed.
13297
13298       -ffixed-reg
13299           Treat the register named reg as a fixed register; generated code
13300           should never refer to it (except perhaps as a stack pointer, frame
13301           pointer or in some other fixed role).
13302
13303           reg must be the name of a register.  The register names accepted
13304           are machine-specific and are defined in the "REGISTER_NAMES" macro
13305           in the machine description macro file.
13306
13307           This flag does not have a negative form, because it specifies a
13308           three-way choice.
13309
13310       -fcall-used-reg
13311           Treat the register named reg as an allocable register that is
13312           clobbered by function calls.  It may be allocated for temporaries
13313           or variables that do not live across a call.  Functions compiled
13314           this way do not save and restore the register reg.
13315
13316           It is an error to use this flag with the frame pointer or stack
13317           pointer.  Use of this flag for other registers that have fixed
13318           pervasive roles in the machine's execution model produces
13319           disastrous results.
13320
13321           This flag does not have a negative form, because it specifies a
13322           three-way choice.
13323
13324       -fcall-saved-reg
13325           Treat the register named reg as an allocable register saved by
13326           functions.  It may be allocated even for temporaries or variables
13327           that live across a call.  Functions compiled this way save and
13328           restore the register reg if they use it.
13329
13330           It is an error to use this flag with the frame pointer or stack
13331           pointer.  Use of this flag for other registers that have fixed
13332           pervasive roles in the machine's execution model produces
13333           disastrous results.
13334
13335           A different sort of disaster results from the use of this flag for
13336           a register in which function values may be returned.
13337
13338           This flag does not have a negative form, because it specifies a
13339           three-way choice.
13340
13341       -fpack-struct[=n]
13342           Without a value specified, pack all structure members together
13343           without holes.  When a value is specified (which must be a small
13344           power of two), pack structure members according to this value,
13345           representing the maximum alignment (that is, objects with default
13346           alignment requirements larger than this are output potentially
13347           unaligned at the next fitting location.
13348
13349           Warning: the -fpack-struct switch causes GCC to generate code that
13350           is not binary compatible with code generated without that switch.
13351           Additionally, it makes the code suboptimal.  Use it to conform to a
13352           non-default application binary interface.
13353
13354       -fleading-underscore
13355           This option and its counterpart, -fno-leading-underscore, forcibly
13356           change the way C symbols are represented in the object file.  One
13357           use is to help link with legacy assembly code.
13358
13359           Warning: the -fleading-underscore switch causes GCC to generate
13360           code that is not binary compatible with code generated without that
13361           switch.  Use it to conform to a non-default application binary
13362           interface.  Not all targets provide complete support for this
13363           switch.
13364
13365       -ftls-model=model
13366           Alter the thread-local storage model to be used.  The model
13367           argument should be one of global-dynamic, local-dynamic, initial-
13368           exec or local-exec.  Note that the choice is subject to
13369           optimization: the compiler may use a more efficient model for
13370           symbols not visible outside of the translation unit, or if -fpic is
13371           not given on the command line.
13372
13373           The default without -fpic is initial-exec; with -fpic the default
13374           is global-dynamic.
13375
13376       -ftrampolines
13377           For targets that normally need trampolines for nested functions,
13378           always generate them instead of using descriptors.  Otherwise, for
13379           targets that do not need them, like for example HP-PA or IA-64, do
13380           nothing.
13381
13382           A trampoline is a small piece of code that is created at run time
13383           on the stack when the address of a nested function is taken, and is
13384           used to call the nested function indirectly.  Therefore, it
13385           requires the stack to be made executable in order for the program
13386           to work properly.
13387
13388           -fno-trampolines is enabled by default on a language by language
13389           basis to let the compiler avoid generating them, if it computes
13390           that this is safe, and replace them with descriptors.  Descriptors
13391           are made up of data only, but the generated code must be prepared
13392           to deal with them.  As of this writing, -fno-trampolines is enabled
13393           by default only for Ada.
13394
13395           Moreover, code compiled with -ftrampolines and code compiled with
13396           -fno-trampolines are not binary compatible if nested functions are
13397           present.  This option must therefore be used on a program-wide
13398           basis and be manipulated with extreme care.
13399
13400       -fvisibility=[default|internal|hidden|protected]
13401           Set the default ELF image symbol visibility to the specified
13402           option---all symbols are marked with this unless overridden within
13403           the code.  Using this feature can very substantially improve
13404           linking and load times of shared object libraries, produce more
13405           optimized code, provide near-perfect API export and prevent symbol
13406           clashes.  It is strongly recommended that you use this in any
13407           shared objects you distribute.
13408
13409           Despite the nomenclature, default always means public; i.e.,
13410           available to be linked against from outside the shared object.
13411           protected and internal are pretty useless in real-world usage so
13412           the only other commonly used option is hidden.  The default if
13413           -fvisibility isn't specified is default, i.e., make every symbol
13414           public.
13415
13416           A good explanation of the benefits offered by ensuring ELF symbols
13417           have the correct visibility is given by "How To Write Shared
13418           Libraries" by Ulrich Drepper (which can be found at
13419           <https://www.akkadia.org/drepper/>)---however a superior solution
13420           made possible by this option to marking things hidden when the
13421           default is public is to make the default hidden and mark things
13422           public.  This is the norm with DLLs on Windows and with
13423           -fvisibility=hidden and "__attribute__ ((visibility("default")))"
13424           instead of "__declspec(dllexport)" you get almost identical
13425           semantics with identical syntax.  This is a great boon to those
13426           working with cross-platform projects.
13427
13428           For those adding visibility support to existing code, you may find
13429           "#pragma GCC visibility" of use.  This works by you enclosing the
13430           declarations you wish to set visibility for with (for example)
13431           "#pragma GCC visibility push(hidden)" and "#pragma GCC visibility
13432           pop".  Bear in mind that symbol visibility should be viewed as part
13433           of the API interface contract and thus all new code should always
13434           specify visibility when it is not the default; i.e., declarations
13435           only for use within the local DSO should always be marked
13436           explicitly as hidden as so to avoid PLT indirection
13437           overheads---making this abundantly clear also aids readability and
13438           self-documentation of the code.  Note that due to ISO C++
13439           specification requirements, "operator new" and "operator delete"
13440           must always be of default visibility.
13441
13442           Be aware that headers from outside your project, in particular
13443           system headers and headers from any other library you use, may not
13444           be expecting to be compiled with visibility other than the default.
13445           You may need to explicitly say "#pragma GCC visibility
13446           push(default)" before including any such headers.
13447
13448           "extern" declarations are not affected by -fvisibility, so a lot of
13449           code can be recompiled with -fvisibility=hidden with no
13450           modifications.  However, this means that calls to "extern"
13451           functions with no explicit visibility use the PLT, so it is more
13452           effective to use "__attribute ((visibility))" and/or "#pragma GCC
13453           visibility" to tell the compiler which "extern" declarations should
13454           be treated as hidden.
13455
13456           Note that -fvisibility does affect C++ vague linkage entities. This
13457           means that, for instance, an exception class that is be thrown
13458           between DSOs must be explicitly marked with default visibility so
13459           that the type_info nodes are unified between the DSOs.
13460
13461           An overview of these techniques, their benefits and how to use them
13462           is at <http://gcc.gnu.org/wiki/Visibility>.
13463
13464       -fstrict-volatile-bitfields
13465           This option should be used if accesses to volatile bit-fields (or
13466           other structure fields, although the compiler usually honors those
13467           types anyway) should use a single access of the width of the
13468           field's type, aligned to a natural alignment if possible.  For
13469           example, targets with memory-mapped peripheral registers might
13470           require all such accesses to be 16 bits wide; with this flag you
13471           can declare all peripheral bit-fields as "unsigned short" (assuming
13472           short is 16 bits on these targets) to force GCC to use 16-bit
13473           accesses instead of, perhaps, a more efficient 32-bit access.
13474
13475           If this option is disabled, the compiler uses the most efficient
13476           instruction.  In the previous example, that might be a 32-bit load
13477           instruction, even though that accesses bytes that do not contain
13478           any portion of the bit-field, or memory-mapped registers unrelated
13479           to the one being updated.
13480
13481           In some cases, such as when the "packed" attribute is applied to a
13482           structure field, it may not be possible to access the field with a
13483           single read or write that is correctly aligned for the target
13484           machine.  In this case GCC falls back to generating multiple
13485           accesses rather than code that will fault or truncate the result at
13486           run time.
13487
13488           Note:  Due to restrictions of the C/C++11 memory model, write
13489           accesses are not allowed to touch non bit-field members.  It is
13490           therefore recommended to define all bits of the field's type as
13491           bit-field members.
13492
13493           The default value of this option is determined by the application
13494           binary interface for the target processor.
13495
13496       -fsync-libcalls
13497           This option controls whether any out-of-line instance of the
13498           "__sync" family of functions may be used to implement the C++11
13499           "__atomic" family of functions.
13500
13501           The default value of this option is enabled, thus the only useful
13502           form of the option is -fno-sync-libcalls.  This option is used in
13503           the implementation of the libatomic runtime library.
13504
13505   GCC Developer Options
13506       This section describes command-line options that are primarily of
13507       interest to GCC developers, including options to support compiler
13508       testing and investigation of compiler bugs and compile-time performance
13509       problems.  This includes options that produce debug dumps at various
13510       points in the compilation; that print statistics such as memory use and
13511       execution time; and that print information about GCC's configuration,
13512       such as where it searches for libraries.  You should rarely need to use
13513       any of these options for ordinary compilation and linking tasks.
13514
13515       Many developer options that cause GCC to dump output to a file take an
13516       optional =filename suffix. You can specify stdout or - to dump to
13517       standard output, and stderr for standard error.
13518
13519       If =filename is omitted, a default dump file name is constructed by
13520       concatenating the base dump file name, a pass number, phase letter, and
13521       pass name.  The base dump file name is the name of output file produced
13522       by the compiler if explicitly specified and not an executable;
13523       otherwise it is the source file name.  The pass number is determined by
13524       the order passes are registered with the compiler's pass manager.  This
13525       is generally the same as the order of execution, but passes registered
13526       by plugins, target-specific passes, or passes that are otherwise
13527       registered late are numbered higher than the pass named final, even if
13528       they are executed earlier.  The phase letter is one of i (inter-
13529       procedural analysis), l (language-specific), r (RTL), or t (tree).  The
13530       files are created in the directory of the output file.
13531
13532       -fcallgraph-info
13533       -fcallgraph-info=MARKERS
13534           Makes the compiler output callgraph information for the program, on
13535           a per-object-file basis.  The information is generated in the
13536           common VCG format.  It can be decorated with additional, per-node
13537           and/or per-edge information, if a list of comma-separated markers
13538           is additionally specified.  When the "su" marker is specified, the
13539           callgraph is decorated with stack usage information; it is
13540           equivalent to -fstack-usage.  When the "da" marker is specified,
13541           the callgraph is decorated with information about dynamically
13542           allocated objects.
13543
13544           When compiling with -flto, no callgraph information is output along
13545           with the object file.  At LTO link time, -fcallgraph-info may
13546           generate multiple callgraph information files next to intermediate
13547           LTO output files.
13548
13549       -dletters
13550       -fdump-rtl-pass
13551       -fdump-rtl-pass=filename
13552           Says to make debugging dumps during compilation at times specified
13553           by letters.  This is used for debugging the RTL-based passes of the
13554           compiler.
13555
13556           Some -dletters switches have different meaning when -E is used for
13557           preprocessing.
13558
13559           Debug dumps can be enabled with a -fdump-rtl switch or some -d
13560           option letters.  Here are the possible letters for use in pass and
13561           letters, and their meanings:
13562
13563           -fdump-rtl-alignments
13564               Dump after branch alignments have been computed.
13565
13566           -fdump-rtl-asmcons
13567               Dump after fixing rtl statements that have unsatisfied in/out
13568               constraints.
13569
13570           -fdump-rtl-auto_inc_dec
13571               Dump after auto-inc-dec discovery.  This pass is only run on
13572               architectures that have auto inc or auto dec instructions.
13573
13574           -fdump-rtl-barriers
13575               Dump after cleaning up the barrier instructions.
13576
13577           -fdump-rtl-bbpart
13578               Dump after partitioning hot and cold basic blocks.
13579
13580           -fdump-rtl-bbro
13581               Dump after block reordering.
13582
13583           -fdump-rtl-btl1
13584           -fdump-rtl-btl2
13585               -fdump-rtl-btl1 and -fdump-rtl-btl2 enable dumping after the
13586               two branch target load optimization passes.
13587
13588           -fdump-rtl-bypass
13589               Dump after jump bypassing and control flow optimizations.
13590
13591           -fdump-rtl-combine
13592               Dump after the RTL instruction combination pass.
13593
13594           -fdump-rtl-compgotos
13595               Dump after duplicating the computed gotos.
13596
13597           -fdump-rtl-ce1
13598           -fdump-rtl-ce2
13599           -fdump-rtl-ce3
13600               -fdump-rtl-ce1, -fdump-rtl-ce2, and -fdump-rtl-ce3 enable
13601               dumping after the three if conversion passes.
13602
13603           -fdump-rtl-cprop_hardreg
13604               Dump after hard register copy propagation.
13605
13606           -fdump-rtl-csa
13607               Dump after combining stack adjustments.
13608
13609           -fdump-rtl-cse1
13610           -fdump-rtl-cse2
13611               -fdump-rtl-cse1 and -fdump-rtl-cse2 enable dumping after the
13612               two common subexpression elimination passes.
13613
13614           -fdump-rtl-dce
13615               Dump after the standalone dead code elimination passes.
13616
13617           -fdump-rtl-dbr
13618               Dump after delayed branch scheduling.
13619
13620           -fdump-rtl-dce1
13621           -fdump-rtl-dce2
13622               -fdump-rtl-dce1 and -fdump-rtl-dce2 enable dumping after the
13623               two dead store elimination passes.
13624
13625           -fdump-rtl-eh
13626               Dump after finalization of EH handling code.
13627
13628           -fdump-rtl-eh_ranges
13629               Dump after conversion of EH handling range regions.
13630
13631           -fdump-rtl-expand
13632               Dump after RTL generation.
13633
13634           -fdump-rtl-fwprop1
13635           -fdump-rtl-fwprop2
13636               -fdump-rtl-fwprop1 and -fdump-rtl-fwprop2 enable dumping after
13637               the two forward propagation passes.
13638
13639           -fdump-rtl-gcse1
13640           -fdump-rtl-gcse2
13641               -fdump-rtl-gcse1 and -fdump-rtl-gcse2 enable dumping after
13642               global common subexpression elimination.
13643
13644           -fdump-rtl-init-regs
13645               Dump after the initialization of the registers.
13646
13647           -fdump-rtl-initvals
13648               Dump after the computation of the initial value sets.
13649
13650           -fdump-rtl-into_cfglayout
13651               Dump after converting to cfglayout mode.
13652
13653           -fdump-rtl-ira
13654               Dump after iterated register allocation.
13655
13656           -fdump-rtl-jump
13657               Dump after the second jump optimization.
13658
13659           -fdump-rtl-loop2
13660               -fdump-rtl-loop2 enables dumping after the rtl loop
13661               optimization passes.
13662
13663           -fdump-rtl-mach
13664               Dump after performing the machine dependent reorganization
13665               pass, if that pass exists.
13666
13667           -fdump-rtl-mode_sw
13668               Dump after removing redundant mode switches.
13669
13670           -fdump-rtl-rnreg
13671               Dump after register renumbering.
13672
13673           -fdump-rtl-outof_cfglayout
13674               Dump after converting from cfglayout mode.
13675
13676           -fdump-rtl-peephole2
13677               Dump after the peephole pass.
13678
13679           -fdump-rtl-postreload
13680               Dump after post-reload optimizations.
13681
13682           -fdump-rtl-pro_and_epilogue
13683               Dump after generating the function prologues and epilogues.
13684
13685           -fdump-rtl-sched1
13686           -fdump-rtl-sched2
13687               -fdump-rtl-sched1 and -fdump-rtl-sched2 enable dumping after
13688               the basic block scheduling passes.
13689
13690           -fdump-rtl-ree
13691               Dump after sign/zero extension elimination.
13692
13693           -fdump-rtl-seqabstr
13694               Dump after common sequence discovery.
13695
13696           -fdump-rtl-shorten
13697               Dump after shortening branches.
13698
13699           -fdump-rtl-sibling
13700               Dump after sibling call optimizations.
13701
13702           -fdump-rtl-split1
13703           -fdump-rtl-split2
13704           -fdump-rtl-split3
13705           -fdump-rtl-split4
13706           -fdump-rtl-split5
13707               These options enable dumping after five rounds of instruction
13708               splitting.
13709
13710           -fdump-rtl-sms
13711               Dump after modulo scheduling.  This pass is only run on some
13712               architectures.
13713
13714           -fdump-rtl-stack
13715               Dump after conversion from GCC's "flat register file" registers
13716               to the x87's stack-like registers.  This pass is only run on
13717               x86 variants.
13718
13719           -fdump-rtl-subreg1
13720           -fdump-rtl-subreg2
13721               -fdump-rtl-subreg1 and -fdump-rtl-subreg2 enable dumping after
13722               the two subreg expansion passes.
13723
13724           -fdump-rtl-unshare
13725               Dump after all rtl has been unshared.
13726
13727           -fdump-rtl-vartrack
13728               Dump after variable tracking.
13729
13730           -fdump-rtl-vregs
13731               Dump after converting virtual registers to hard registers.
13732
13733           -fdump-rtl-web
13734               Dump after live range splitting.
13735
13736           -fdump-rtl-regclass
13737           -fdump-rtl-subregs_of_mode_init
13738           -fdump-rtl-subregs_of_mode_finish
13739           -fdump-rtl-dfinit
13740           -fdump-rtl-dfinish
13741               These dumps are defined but always produce empty files.
13742
13743           -da
13744           -fdump-rtl-all
13745               Produce all the dumps listed above.
13746
13747           -dA Annotate the assembler output with miscellaneous debugging
13748               information.
13749
13750           -dD Dump all macro definitions, at the end of preprocessing, in
13751               addition to normal output.
13752
13753           -dH Produce a core dump whenever an error occurs.
13754
13755           -dp Annotate the assembler output with a comment indicating which
13756               pattern and alternative is used.  The length and cost of each
13757               instruction are also printed.
13758
13759           -dP Dump the RTL in the assembler output as a comment before each
13760               instruction.  Also turns on -dp annotation.
13761
13762           -dx Just generate RTL for a function instead of compiling it.
13763               Usually used with -fdump-rtl-expand.
13764
13765       -fdump-debug
13766           Dump debugging information generated during the debug generation
13767           phase.
13768
13769       -fdump-earlydebug
13770           Dump debugging information generated during the early debug
13771           generation phase.
13772
13773       -fdump-noaddr
13774           When doing debugging dumps, suppress address output.  This makes it
13775           more feasible to use diff on debugging dumps for compiler
13776           invocations with different compiler binaries and/or different text
13777           / bss / data / heap / stack / dso start locations.
13778
13779       -freport-bug
13780           Collect and dump debug information into a temporary file if an
13781           internal compiler error (ICE) occurs.
13782
13783       -fdump-unnumbered
13784           When doing debugging dumps, suppress instruction numbers and
13785           address output.  This makes it more feasible to use diff on
13786           debugging dumps for compiler invocations with different options, in
13787           particular with and without -g.
13788
13789       -fdump-unnumbered-links
13790           When doing debugging dumps (see -d option above), suppress
13791           instruction numbers for the links to the previous and next
13792           instructions in a sequence.
13793
13794       -fdump-ipa-switch
13795       -fdump-ipa-switch-options
13796           Control the dumping at various stages of inter-procedural analysis
13797           language tree to a file.  The file name is generated by appending a
13798           switch specific suffix to the source file name, and the file is
13799           created in the same directory as the output file.  The following
13800           dumps are possible:
13801
13802           all Enables all inter-procedural analysis dumps.
13803
13804           cgraph
13805               Dumps information about call-graph optimization, unused
13806               function removal, and inlining decisions.
13807
13808           inline
13809               Dump after function inlining.
13810
13811           Additionally, the options -optimized, -missed, -note, and -all can
13812           be provided, with the same meaning as for -fopt-info, defaulting to
13813           -optimized.
13814
13815           For example, -fdump-ipa-inline-optimized-missed will emit
13816           information on callsites that were inlined, along with callsites
13817           that were not inlined.
13818
13819           By default, the dump will contain messages about successful
13820           optimizations (equivalent to -optimized) together with low-level
13821           details about the analysis.
13822
13823       -fdump-lang-all
13824       -fdump-lang-switch
13825       -fdump-lang-switch-options
13826       -fdump-lang-switch-options=filename
13827           Control the dumping of language-specific information.  The options
13828           and filename portions behave as described in the -fdump-tree
13829           option.  The following switch values are accepted:
13830
13831           all Enable all language-specific dumps.
13832
13833           class
13834               Dump class hierarchy information.  Virtual table information is
13835               emitted unless 'slim' is specified.  This option is applicable
13836               to C++ only.
13837
13838           raw Dump the raw internal tree data.  This option is applicable to
13839               C++ only.
13840
13841       -fdump-passes
13842           Print on stderr the list of optimization passes that are turned on
13843           and off by the current command-line options.
13844
13845       -fdump-statistics-option
13846           Enable and control dumping of pass statistics in a separate file.
13847           The file name is generated by appending a suffix ending in
13848           .statistics to the source file name, and the file is created in the
13849           same directory as the output file.  If the -option form is used,
13850           -stats causes counters to be summed over the whole compilation unit
13851           while -details dumps every event as the passes generate them.  The
13852           default with no option is to sum counters for each function
13853           compiled.
13854
13855       -fdump-tree-all
13856       -fdump-tree-switch
13857       -fdump-tree-switch-options
13858       -fdump-tree-switch-options=filename
13859           Control the dumping at various stages of processing the
13860           intermediate language tree to a file.  If the -options form is
13861           used, options is a list of - separated options which control the
13862           details of the dump.  Not all options are applicable to all dumps;
13863           those that are not meaningful are ignored.  The following options
13864           are available
13865
13866           address
13867               Print the address of each node.  Usually this is not meaningful
13868               as it changes according to the environment and source file.
13869               Its primary use is for tying up a dump file with a debug
13870               environment.
13871
13872           asmname
13873               If "DECL_ASSEMBLER_NAME" has been set for a given decl, use
13874               that in the dump instead of "DECL_NAME".  Its primary use is
13875               ease of use working backward from mangled names in the assembly
13876               file.
13877
13878           slim
13879               When dumping front-end intermediate representations, inhibit
13880               dumping of members of a scope or body of a function merely
13881               because that scope has been reached.  Only dump such items when
13882               they are directly reachable by some other path.
13883
13884               When dumping pretty-printed trees, this option inhibits dumping
13885               the bodies of control structures.
13886
13887               When dumping RTL, print the RTL in slim (condensed) form
13888               instead of the default LISP-like representation.
13889
13890           raw Print a raw representation of the tree.  By default, trees are
13891               pretty-printed into a C-like representation.
13892
13893           details
13894               Enable more detailed dumps (not honored by every dump option).
13895               Also include information from the optimization passes.
13896
13897           stats
13898               Enable dumping various statistics about the pass (not honored
13899               by every dump option).
13900
13901           blocks
13902               Enable showing basic block boundaries (disabled in raw dumps).
13903
13904           graph
13905               For each of the other indicated dump files (-fdump-rtl-pass),
13906               dump a representation of the control flow graph suitable for
13907               viewing with GraphViz to file.passid.pass.dot.  Each function
13908               in the file is pretty-printed as a subgraph, so that GraphViz
13909               can render them all in a single plot.
13910
13911               This option currently only works for RTL dumps, and the RTL is
13912               always dumped in slim form.
13913
13914           vops
13915               Enable showing virtual operands for every statement.
13916
13917           lineno
13918               Enable showing line numbers for statements.
13919
13920           uid Enable showing the unique ID ("DECL_UID") for each variable.
13921
13922           verbose
13923               Enable showing the tree dump for each statement.
13924
13925           eh  Enable showing the EH region number holding each statement.
13926
13927           scev
13928               Enable showing scalar evolution analysis details.
13929
13930           optimized
13931               Enable showing optimization information (only available in
13932               certain passes).
13933
13934           missed
13935               Enable showing missed optimization information (only available
13936               in certain passes).
13937
13938           note
13939               Enable other detailed optimization information (only available
13940               in certain passes).
13941
13942           all Turn on all options, except raw, slim, verbose and lineno.
13943
13944           optall
13945               Turn on all optimization options, i.e., optimized, missed, and
13946               note.
13947
13948           To determine what tree dumps are available or find the dump for a
13949           pass of interest follow the steps below.
13950
13951           1.  Invoke GCC with -fdump-passes and in the stderr output look for
13952               a code that corresponds to the pass you are interested in.  For
13953               example, the codes "tree-evrp", "tree-vrp1", and "tree-vrp2"
13954               correspond to the three Value Range Propagation passes.  The
13955               number at the end distinguishes distinct invocations of the
13956               same pass.
13957
13958           2.  To enable the creation of the dump file, append the pass code
13959               to the -fdump- option prefix and invoke GCC with it.  For
13960               example, to enable the dump from the Early Value Range
13961               Propagation pass, invoke GCC with the -fdump-tree-evrp option.
13962               Optionally, you may specify the name of the dump file.  If you
13963               don't specify one, GCC creates as described below.
13964
13965           3.  Find the pass dump in a file whose name is composed of three
13966               components separated by a period: the name of the source file
13967               GCC was invoked to compile, a numeric suffix indicating the
13968               pass number followed by the letter t for tree passes (and the
13969               letter r for RTL passes), and finally the pass code.  For
13970               example, the Early VRP pass dump might be in a file named
13971               myfile.c.038t.evrp in the current working directory.  Note that
13972               the numeric codes are not stable and may change from one
13973               version of GCC to another.
13974
13975       -fopt-info
13976       -fopt-info-options
13977       -fopt-info-options=filename
13978           Controls optimization dumps from various optimization passes. If
13979           the -options form is used, options is a list of - separated option
13980           keywords to select the dump details and optimizations.
13981
13982           The options can be divided into three groups:
13983
13984           1.  options describing what kinds of messages should be emitted,
13985
13986           2.  options describing the verbosity of the dump, and
13987
13988           3.  options describing which optimizations should be included.
13989
13990           The options from each group can be freely mixed as they are non-
13991           overlapping. However, in case of any conflicts, the later options
13992           override the earlier options on the command line.
13993
13994           The following options control which kinds of messages should be
13995           emitted:
13996
13997           optimized
13998               Print information when an optimization is successfully applied.
13999               It is up to a pass to decide which information is relevant. For
14000               example, the vectorizer passes print the source location of
14001               loops which are successfully vectorized.
14002
14003           missed
14004               Print information about missed optimizations. Individual passes
14005               control which information to include in the output.
14006
14007           note
14008               Print verbose information about optimizations, such as certain
14009               transformations, more detailed messages about decisions etc.
14010
14011           all Print detailed optimization information. This includes
14012               optimized, missed, and note.
14013
14014           The following option controls the dump verbosity:
14015
14016           internals
14017               By default, only "high-level" messages are emitted. This option
14018               enables additional, more detailed, messages, which are likely
14019               to only be of interest to GCC developers.
14020
14021           One or more of the following option keywords can be used to
14022           describe a group of optimizations:
14023
14024           ipa Enable dumps from all interprocedural optimizations.
14025
14026           loop
14027               Enable dumps from all loop optimizations.
14028
14029           inline
14030               Enable dumps from all inlining optimizations.
14031
14032           omp Enable dumps from all OMP (Offloading and Multi Processing)
14033               optimizations.
14034
14035           vec Enable dumps from all vectorization optimizations.
14036
14037           optall
14038               Enable dumps from all optimizations. This is a superset of the
14039               optimization groups listed above.
14040
14041           If options is omitted, it defaults to optimized-optall, which means
14042           to dump messages about successful optimizations from all the
14043           passes, omitting messages that are treated as "internals".
14044
14045           If the filename is provided, then the dumps from all the applicable
14046           optimizations are concatenated into the filename.  Otherwise the
14047           dump is output onto stderr. Though multiple -fopt-info options are
14048           accepted, only one of them can include a filename. If other
14049           filenames are provided then all but the first such option are
14050           ignored.
14051
14052           Note that the output filename is overwritten in case of multiple
14053           translation units. If a combined output from multiple translation
14054           units is desired, stderr should be used instead.
14055
14056           In the following example, the optimization info is output to
14057           stderr:
14058
14059                   gcc -O3 -fopt-info
14060
14061           This example:
14062
14063                   gcc -O3 -fopt-info-missed=missed.all
14064
14065           outputs missed optimization report from all the passes into
14066           missed.all, and this one:
14067
14068                   gcc -O2 -ftree-vectorize -fopt-info-vec-missed
14069
14070           prints information about missed optimization opportunities from
14071           vectorization passes on stderr.  Note that -fopt-info-vec-missed is
14072           equivalent to -fopt-info-missed-vec.  The order of the optimization
14073           group names and message types listed after -fopt-info does not
14074           matter.
14075
14076           As another example,
14077
14078                   gcc -O3 -fopt-info-inline-optimized-missed=inline.txt
14079
14080           outputs information about missed optimizations as well as optimized
14081           locations from all the inlining passes into inline.txt.
14082
14083           Finally, consider:
14084
14085                   gcc -fopt-info-vec-missed=vec.miss -fopt-info-loop-optimized=loop.opt
14086
14087           Here the two output filenames vec.miss and loop.opt are in conflict
14088           since only one output file is allowed. In this case, only the first
14089           option takes effect and the subsequent options are ignored. Thus
14090           only vec.miss is produced which contains dumps from the vectorizer
14091           about missed opportunities.
14092
14093       -fsave-optimization-record
14094           Write a SRCFILE.opt-record.json.gz file detailing what
14095           optimizations were performed, for those optimizations that support
14096           -fopt-info.
14097
14098           This option is experimental and the format of the data within the
14099           compressed JSON file is subject to change.
14100
14101           It is roughly equivalent to a machine-readable version of
14102           -fopt-info-all, as a collection of messages with source file, line
14103           number and column number, with the following additional data for
14104           each message:
14105
14106           *   the execution count of the code being optimized, along with
14107               metadata about whether this was from actual profile data, or
14108               just an estimate, allowing consumers to prioritize messages by
14109               code hotness,
14110
14111           *   the function name of the code being optimized, where
14112               applicable,
14113
14114           *   the "inlining chain" for the code being optimized, so that when
14115               a function is inlined into several different places (which
14116               might themselves be inlined), the reader can distinguish
14117               between the copies,
14118
14119           *   objects identifying those parts of the message that refer to
14120               expressions, statements or symbol-table nodes, which of these
14121               categories they are, and, when available, their source code
14122               location,
14123
14124           *   the GCC pass that emitted the message, and
14125
14126           *   the location in GCC's own code from which the message was
14127               emitted
14128
14129           Additionally, some messages are logically nested within other
14130           messages, reflecting implementation details of the optimization
14131           passes.
14132
14133       -fsched-verbose=n
14134           On targets that use instruction scheduling, this option controls
14135           the amount of debugging output the scheduler prints to the dump
14136           files.
14137
14138           For n greater than zero, -fsched-verbose outputs the same
14139           information as -fdump-rtl-sched1 and -fdump-rtl-sched2.  For n
14140           greater than one, it also output basic block probabilities,
14141           detailed ready list information and unit/insn info.  For n greater
14142           than two, it includes RTL at abort point, control-flow and regions
14143           info.  And for n over four, -fsched-verbose also includes
14144           dependence info.
14145
14146       -fenable-kind-pass
14147       -fdisable-kind-pass=range-list
14148           This is a set of options that are used to explicitly disable/enable
14149           optimization passes.  These options are intended for use for
14150           debugging GCC.  Compiler users should use regular options for
14151           enabling/disabling passes instead.
14152
14153           -fdisable-ipa-pass
14154               Disable IPA pass pass. pass is the pass name.  If the same pass
14155               is statically invoked in the compiler multiple times, the pass
14156               name should be appended with a sequential number starting from
14157               1.
14158
14159           -fdisable-rtl-pass
14160           -fdisable-rtl-pass=range-list
14161               Disable RTL pass pass.  pass is the pass name.  If the same
14162               pass is statically invoked in the compiler multiple times, the
14163               pass name should be appended with a sequential number starting
14164               from 1.  range-list is a comma-separated list of function
14165               ranges or assembler names.  Each range is a number pair
14166               separated by a colon.  The range is inclusive in both ends.  If
14167               the range is trivial, the number pair can be simplified as a
14168               single number.  If the function's call graph node's uid falls
14169               within one of the specified ranges, the pass is disabled for
14170               that function.  The uid is shown in the function header of a
14171               dump file, and the pass names can be dumped by using option
14172               -fdump-passes.
14173
14174           -fdisable-tree-pass
14175           -fdisable-tree-pass=range-list
14176               Disable tree pass pass.  See -fdisable-rtl for the description
14177               of option arguments.
14178
14179           -fenable-ipa-pass
14180               Enable IPA pass pass.  pass is the pass name.  If the same pass
14181               is statically invoked in the compiler multiple times, the pass
14182               name should be appended with a sequential number starting from
14183               1.
14184
14185           -fenable-rtl-pass
14186           -fenable-rtl-pass=range-list
14187               Enable RTL pass pass.  See -fdisable-rtl for option argument
14188               description and examples.
14189
14190           -fenable-tree-pass
14191           -fenable-tree-pass=range-list
14192               Enable tree pass pass.  See -fdisable-rtl for the description
14193               of option arguments.
14194
14195           Here are some examples showing uses of these options.
14196
14197                   # disable ccp1 for all functions
14198                      -fdisable-tree-ccp1
14199                   # disable complete unroll for function whose cgraph node uid is 1
14200                      -fenable-tree-cunroll=1
14201                   # disable gcse2 for functions at the following ranges [1,1],
14202                   # [300,400], and [400,1000]
14203                   # disable gcse2 for functions foo and foo2
14204                      -fdisable-rtl-gcse2=foo,foo2
14205                   # disable early inlining
14206                      -fdisable-tree-einline
14207                   # disable ipa inlining
14208                      -fdisable-ipa-inline
14209                   # enable tree full unroll
14210                      -fenable-tree-unroll
14211
14212       -fchecking
14213       -fchecking=n
14214           Enable internal consistency checking.  The default depends on the
14215           compiler configuration.  -fchecking=2 enables further internal
14216           consistency checking that might affect code generation.
14217
14218       -frandom-seed=string
14219           This option provides a seed that GCC uses in place of random
14220           numbers in generating certain symbol names that have to be
14221           different in every compiled file.  It is also used to place unique
14222           stamps in coverage data files and the object files that produce
14223           them.  You can use the -frandom-seed option to produce reproducibly
14224           identical object files.
14225
14226           The string can either be a number (decimal, octal or hex) or an
14227           arbitrary string (in which case it's converted to a number by
14228           computing CRC32).
14229
14230           The string should be different for every file you compile.
14231
14232       -save-temps
14233       -save-temps=cwd
14234           Store the usual "temporary" intermediate files permanently; place
14235           them in the current directory and name them based on the source
14236           file.  Thus, compiling foo.c with -c -save-temps produces files
14237           foo.i and foo.s, as well as foo.o.  This creates a preprocessed
14238           foo.i output file even though the compiler now normally uses an
14239           integrated preprocessor.
14240
14241           When used in combination with the -x command-line option,
14242           -save-temps is sensible enough to avoid over writing an input
14243           source file with the same extension as an intermediate file.  The
14244           corresponding intermediate file may be obtained by renaming the
14245           source file before using -save-temps.
14246
14247           If you invoke GCC in parallel, compiling several different source
14248           files that share a common base name in different subdirectories or
14249           the same source file compiled for multiple output destinations, it
14250           is likely that the different parallel compilers will interfere with
14251           each other, and overwrite the temporary files.  For instance:
14252
14253                   gcc -save-temps -o outdir1/foo.o indir1/foo.c&
14254                   gcc -save-temps -o outdir2/foo.o indir2/foo.c&
14255
14256           may result in foo.i and foo.o being written to simultaneously by
14257           both compilers.
14258
14259       -save-temps=obj
14260           Store the usual "temporary" intermediate files permanently.  If the
14261           -o option is used, the temporary files are based on the object
14262           file.  If the -o option is not used, the -save-temps=obj switch
14263           behaves like -save-temps.
14264
14265           For example:
14266
14267                   gcc -save-temps=obj -c foo.c
14268                   gcc -save-temps=obj -c bar.c -o dir/xbar.o
14269                   gcc -save-temps=obj foobar.c -o dir2/yfoobar
14270
14271           creates foo.i, foo.s, dir/xbar.i, dir/xbar.s, dir2/yfoobar.i,
14272           dir2/yfoobar.s, and dir2/yfoobar.o.
14273
14274       -time[=file]
14275           Report the CPU time taken by each subprocess in the compilation
14276           sequence.  For C source files, this is the compiler proper and
14277           assembler (plus the linker if linking is done).
14278
14279           Without the specification of an output file, the output looks like
14280           this:
14281
14282                   # cc1 0.12 0.01
14283                   # as 0.00 0.01
14284
14285           The first number on each line is the "user time", that is time
14286           spent executing the program itself.  The second number is "system
14287           time", time spent executing operating system routines on behalf of
14288           the program.  Both numbers are in seconds.
14289
14290           With the specification of an output file, the output is appended to
14291           the named file, and it looks like this:
14292
14293                   0.12 0.01 cc1 <options>
14294                   0.00 0.01 as <options>
14295
14296           The "user time" and the "system time" are moved before the program
14297           name, and the options passed to the program are displayed, so that
14298           one can later tell what file was being compiled, and with which
14299           options.
14300
14301       -fdump-final-insns[=file]
14302           Dump the final internal representation (RTL) to file.  If the
14303           optional argument is omitted (or if file is "."), the name of the
14304           dump file is determined by appending ".gkd" to the compilation
14305           output file name.
14306
14307       -fcompare-debug[=opts]
14308           If no error occurs during compilation, run the compiler a second
14309           time, adding opts and -fcompare-debug-second to the arguments
14310           passed to the second compilation.  Dump the final internal
14311           representation in both compilations, and print an error if they
14312           differ.
14313
14314           If the equal sign is omitted, the default -gtoggle is used.
14315
14316           The environment variable GCC_COMPARE_DEBUG, if defined, non-empty
14317           and nonzero, implicitly enables -fcompare-debug.  If
14318           GCC_COMPARE_DEBUG is defined to a string starting with a dash, then
14319           it is used for opts, otherwise the default -gtoggle is used.
14320
14321           -fcompare-debug=, with the equal sign but without opts, is
14322           equivalent to -fno-compare-debug, which disables the dumping of the
14323           final representation and the second compilation, preventing even
14324           GCC_COMPARE_DEBUG from taking effect.
14325
14326           To verify full coverage during -fcompare-debug testing, set
14327           GCC_COMPARE_DEBUG to say -fcompare-debug-not-overridden, which GCC
14328           rejects as an invalid option in any actual compilation (rather than
14329           preprocessing, assembly or linking).  To get just a warning,
14330           setting GCC_COMPARE_DEBUG to -w%n-fcompare-debug not overridden
14331           will do.
14332
14333       -fcompare-debug-second
14334           This option is implicitly passed to the compiler for the second
14335           compilation requested by -fcompare-debug, along with options to
14336           silence warnings, and omitting other options that would cause the
14337           compiler to produce output to files or to standard output as a side
14338           effect.  Dump files and preserved temporary files are renamed so as
14339           to contain the ".gk" additional extension during the second
14340           compilation, to avoid overwriting those generated by the first.
14341
14342           When this option is passed to the compiler driver, it causes the
14343           first compilation to be skipped, which makes it useful for little
14344           other than debugging the compiler proper.
14345
14346       -gtoggle
14347           Turn off generation of debug info, if leaving out this option
14348           generates it, or turn it on at level 2 otherwise.  The position of
14349           this argument in the command line does not matter; it takes effect
14350           after all other options are processed, and it does so only once, no
14351           matter how many times it is given.  This is mainly intended to be
14352           used with -fcompare-debug.
14353
14354       -fvar-tracking-assignments-toggle
14355           Toggle -fvar-tracking-assignments, in the same way that -gtoggle
14356           toggles -g.
14357
14358       -Q  Makes the compiler print out each function name as it is compiled,
14359           and print some statistics about each pass when it finishes.
14360
14361       -ftime-report
14362           Makes the compiler print some statistics about the time consumed by
14363           each pass when it finishes.
14364
14365       -ftime-report-details
14366           Record the time consumed by infrastructure parts separately for
14367           each pass.
14368
14369       -fira-verbose=n
14370           Control the verbosity of the dump file for the integrated register
14371           allocator.  The default value is 5.  If the value n is greater or
14372           equal to 10, the dump output is sent to stderr using the same
14373           format as n minus 10.
14374
14375       -flto-report
14376           Prints a report with internal details on the workings of the link-
14377           time optimizer.  The contents of this report vary from version to
14378           version.  It is meant to be useful to GCC developers when
14379           processing object files in LTO mode (via -flto).
14380
14381           Disabled by default.
14382
14383       -flto-report-wpa
14384           Like -flto-report, but only print for the WPA phase of link-time
14385           optimization.
14386
14387       -fmem-report
14388           Makes the compiler print some statistics about permanent memory
14389           allocation when it finishes.
14390
14391       -fmem-report-wpa
14392           Makes the compiler print some statistics about permanent memory
14393           allocation for the WPA phase only.
14394
14395       -fpre-ipa-mem-report
14396       -fpost-ipa-mem-report
14397           Makes the compiler print some statistics about permanent memory
14398           allocation before or after interprocedural optimization.
14399
14400       -fprofile-report
14401           Makes the compiler print some statistics about consistency of the
14402           (estimated) profile and effect of individual passes.
14403
14404       -fstack-usage
14405           Makes the compiler output stack usage information for the program,
14406           on a per-function basis.  The filename for the dump is made by
14407           appending .su to the auxname.  auxname is generated from the name
14408           of the output file, if explicitly specified and it is not an
14409           executable, otherwise it is the basename of the source file.  An
14410           entry is made up of three fields:
14411
14412           *   The name of the function.
14413
14414           *   A number of bytes.
14415
14416           *   One or more qualifiers: "static", "dynamic", "bounded".
14417
14418           The qualifier "static" means that the function manipulates the
14419           stack statically: a fixed number of bytes are allocated for the
14420           frame on function entry and released on function exit; no stack
14421           adjustments are otherwise made in the function.  The second field
14422           is this fixed number of bytes.
14423
14424           The qualifier "dynamic" means that the function manipulates the
14425           stack dynamically: in addition to the static allocation described
14426           above, stack adjustments are made in the body of the function, for
14427           example to push/pop arguments around function calls.  If the
14428           qualifier "bounded" is also present, the amount of these
14429           adjustments is bounded at compile time and the second field is an
14430           upper bound of the total amount of stack used by the function.  If
14431           it is not present, the amount of these adjustments is not bounded
14432           at compile time and the second field only represents the bounded
14433           part.
14434
14435       -fstats
14436           Emit statistics about front-end processing at the end of the
14437           compilation.  This option is supported only by the C++ front end,
14438           and the information is generally only useful to the G++ development
14439           team.
14440
14441       -fdbg-cnt-list
14442           Print the name and the counter upper bound for all debug counters.
14443
14444       -fdbg-cnt=counter-value-list
14445           Set the internal debug counter lower and upper bound.  counter-
14446           value-list is a comma-separated list of
14447           name:lower_bound1-upper_bound1 [:lower_bound2-upper_bound2...]
14448           tuples which sets the name of the counter and list of closed
14449           intervals.  The lower_bound is optional and is zero initialized if
14450           not set.  For example, with -fdbg-cnt=dce:2-4:10-11,tail_call:10,
14451           "dbg_cnt(dce)" returns true only for second, third, fourth, tenth
14452           and eleventh invocation.  For "dbg_cnt(tail_call)" true is returned
14453           for first 10 invocations.
14454
14455       -print-file-name=library
14456           Print the full absolute name of the library file library that would
14457           be used when linking---and don't do anything else.  With this
14458           option, GCC does not compile or link anything; it just prints the
14459           file name.
14460
14461       -print-multi-directory
14462           Print the directory name corresponding to the multilib selected by
14463           any other switches present in the command line.  This directory is
14464           supposed to exist in GCC_EXEC_PREFIX.
14465
14466       -print-multi-lib
14467           Print the mapping from multilib directory names to compiler
14468           switches that enable them.  The directory name is separated from
14469           the switches by ;, and each switch starts with an @ instead of the
14470           -, without spaces between multiple switches.  This is supposed to
14471           ease shell processing.
14472
14473       -print-multi-os-directory
14474           Print the path to OS libraries for the selected multilib, relative
14475           to some lib subdirectory.  If OS libraries are present in the lib
14476           subdirectory and no multilibs are used, this is usually just ., if
14477           OS libraries are present in libsuffix sibling directories this
14478           prints e.g. ../lib64, ../lib or ../lib32, or if OS libraries are
14479           present in lib/subdir subdirectories it prints e.g. amd64, sparcv9
14480           or ev6.
14481
14482       -print-multiarch
14483           Print the path to OS libraries for the selected multiarch, relative
14484           to some lib subdirectory.
14485
14486       -print-prog-name=program
14487           Like -print-file-name, but searches for a program such as cpp.
14488
14489       -print-libgcc-file-name
14490           Same as -print-file-name=libgcc.a.
14491
14492           This is useful when you use -nostdlib or -nodefaultlibs but you do
14493           want to link with libgcc.a.  You can do:
14494
14495                   gcc -nostdlib <files>... `gcc -print-libgcc-file-name`
14496
14497       -print-search-dirs
14498           Print the name of the configured installation directory and a list
14499           of program and library directories gcc searches---and don't do
14500           anything else.
14501
14502           This is useful when gcc prints the error message installation
14503           problem, cannot exec cpp0: No such file or directory.  To resolve
14504           this you either need to put cpp0 and the other compiler components
14505           where gcc expects to find them, or you can set the environment
14506           variable GCC_EXEC_PREFIX to the directory where you installed them.
14507           Don't forget the trailing /.
14508
14509       -print-sysroot
14510           Print the target sysroot directory that is used during compilation.
14511           This is the target sysroot specified either at configure time or
14512           using the --sysroot option, possibly with an extra suffix that
14513           depends on compilation options.  If no target sysroot is specified,
14514           the option prints nothing.
14515
14516       -print-sysroot-headers-suffix
14517           Print the suffix added to the target sysroot when searching for
14518           headers, or give an error if the compiler is not configured with
14519           such a suffix---and don't do anything else.
14520
14521       -dumpmachine
14522           Print the compiler's target machine (for example,
14523           i686-pc-linux-gnu)---and don't do anything else.
14524
14525       -dumpversion
14526           Print the compiler version (for example, 3.0, 6.3.0 or 7)---and
14527           don't do anything else.  This is the compiler version used in
14528           filesystem paths and specs. Depending on how the compiler has been
14529           configured it can be just a single number (major version), two
14530           numbers separated by a dot (major and minor version) or three
14531           numbers separated by dots (major, minor and patchlevel version).
14532
14533       -dumpfullversion
14534           Print the full compiler version---and don't do anything else. The
14535           output is always three numbers separated by dots, major, minor and
14536           patchlevel version.
14537
14538       -dumpspecs
14539           Print the compiler's built-in specs---and don't do anything else.
14540           (This is used when GCC itself is being built.)
14541
14542   Machine-Dependent Options
14543       Each target machine supported by GCC can have its own options---for
14544       example, to allow you to compile for a particular processor variant or
14545       ABI, or to control optimizations specific to that machine.  By
14546       convention, the names of machine-specific options start with -m.
14547
14548       Some configurations of the compiler also support additional target-
14549       specific options, usually for compatibility with other compilers on the
14550       same platform.
14551
14552   AArch64 Options
14553       These options are defined for AArch64 implementations:
14554
14555       -mabi=name
14556           Generate code for the specified data model.  Permissible values are
14557           ilp32 for SysV-like data model where int, long int and pointers are
14558           32 bits, and lp64 for SysV-like data model where int is 32 bits,
14559           but long int and pointers are 64 bits.
14560
14561           The default depends on the specific target configuration.  Note
14562           that the LP64 and ILP32 ABIs are not link-compatible; you must
14563           compile your entire program with the same ABI, and link with a
14564           compatible set of libraries.
14565
14566       -mbig-endian
14567           Generate big-endian code.  This is the default when GCC is
14568           configured for an aarch64_be-*-* target.
14569
14570       -mgeneral-regs-only
14571           Generate code which uses only the general-purpose registers.  This
14572           will prevent the compiler from using floating-point and Advanced
14573           SIMD registers but will not impose any restrictions on the
14574           assembler.
14575
14576       -mlittle-endian
14577           Generate little-endian code.  This is the default when GCC is
14578           configured for an aarch64-*-* but not an aarch64_be-*-* target.
14579
14580       -mcmodel=tiny
14581           Generate code for the tiny code model.  The program and its
14582           statically defined symbols must be within 1MB of each other.
14583           Programs can be statically or dynamically linked.
14584
14585       -mcmodel=small
14586           Generate code for the small code model.  The program and its
14587           statically defined symbols must be within 4GB of each other.
14588           Programs can be statically or dynamically linked.  This is the
14589           default code model.
14590
14591       -mcmodel=large
14592           Generate code for the large code model.  This makes no assumptions
14593           about addresses and sizes of sections.  Programs can be statically
14594           linked only.  The -mcmodel=large option is incompatible with
14595           -mabi=ilp32, -fpic and -fPIC.
14596
14597       -mstrict-align
14598       -mno-strict-align
14599           Avoid or allow generating memory accesses that may not be aligned
14600           on a natural object boundary as described in the architecture
14601           specification.
14602
14603       -momit-leaf-frame-pointer
14604       -mno-omit-leaf-frame-pointer
14605           Omit or keep the frame pointer in leaf functions.  The former
14606           behavior is the default.
14607
14608       -mstack-protector-guard=guard
14609       -mstack-protector-guard-reg=reg
14610       -mstack-protector-guard-offset=offset
14611           Generate stack protection code using canary at guard.  Supported
14612           locations are global for a global canary or sysreg for a canary in
14613           an appropriate system register.
14614
14615           With the latter choice the options -mstack-protector-guard-reg=reg
14616           and -mstack-protector-guard-offset=offset furthermore specify which
14617           system register to use as base register for reading the canary, and
14618           from what offset from that base register. There is no default
14619           register or offset as this is entirely for use within the Linux
14620           kernel.
14621
14622       -mstack-protector-guard=guard
14623       -mstack-protector-guard-reg=reg
14624       -mstack-protector-guard-offset=offset
14625           Generate stack protection code using canary at guard.  Supported
14626           locations are global for a global canary or sysreg for a canary in
14627           an appropriate system register.
14628
14629           With the latter choice the options -mstack-protector-guard-reg=reg
14630           and -mstack-protector-guard-offset=offset furthermore specify which
14631           system register to use as base register for reading the canary, and
14632           from what offset from that base register. There is no default
14633           register or offset as this is entirely for use within the Linux
14634           kernel.
14635
14636       -mtls-dialect=desc
14637           Use TLS descriptors as the thread-local storage mechanism for
14638           dynamic accesses of TLS variables.  This is the default.
14639
14640       -mtls-dialect=traditional
14641           Use traditional TLS as the thread-local storage mechanism for
14642           dynamic accesses of TLS variables.
14643
14644       -mtls-size=size
14645           Specify bit size of immediate TLS offsets.  Valid values are 12,
14646           24, 32, 48.  This option requires binutils 2.26 or newer.
14647
14648       -mfix-cortex-a53-835769
14649       -mno-fix-cortex-a53-835769
14650           Enable or disable the workaround for the ARM Cortex-A53 erratum
14651           number 835769.  This involves inserting a NOP instruction between
14652           memory instructions and 64-bit integer multiply-accumulate
14653           instructions.
14654
14655       -mfix-cortex-a53-843419
14656       -mno-fix-cortex-a53-843419
14657           Enable or disable the workaround for the ARM Cortex-A53 erratum
14658           number 843419.  This erratum workaround is made at link time and
14659           this will only pass the corresponding flag to the linker.
14660
14661       -mlow-precision-recip-sqrt
14662       -mno-low-precision-recip-sqrt
14663           Enable or disable the reciprocal square root approximation.  This
14664           option only has an effect if -ffast-math or
14665           -funsafe-math-optimizations is used as well.  Enabling this reduces
14666           precision of reciprocal square root results to about 16 bits for
14667           single precision and to 32 bits for double precision.
14668
14669       -mlow-precision-sqrt
14670       -mno-low-precision-sqrt
14671           Enable or disable the square root approximation.  This option only
14672           has an effect if -ffast-math or -funsafe-math-optimizations is used
14673           as well.  Enabling this reduces precision of square root results to
14674           about 16 bits for single precision and to 32 bits for double
14675           precision.  If enabled, it implies -mlow-precision-recip-sqrt.
14676
14677       -mlow-precision-div
14678       -mno-low-precision-div
14679           Enable or disable the division approximation.  This option only has
14680           an effect if -ffast-math or -funsafe-math-optimizations is used as
14681           well.  Enabling this reduces precision of division results to about
14682           16 bits for single precision and to 32 bits for double precision.
14683
14684       -mtrack-speculation
14685       -mno-track-speculation
14686           Enable or disable generation of additional code to track
14687           speculative execution through conditional branches.  The tracking
14688           state can then be used by the compiler when expanding calls to
14689           "__builtin_speculation_safe_copy" to permit a more efficient code
14690           sequence to be generated.
14691
14692       -moutline-atomics
14693       -mno-outline-atomics
14694           Enable or disable calls to out-of-line helpers to implement atomic
14695           operations.  These helpers will, at runtime, determine if the LSE
14696           instructions from ARMv8.1-A can be used; if not, they will use the
14697           load/store-exclusive instructions that are present in the base
14698           ARMv8.0 ISA.
14699
14700           This option is only applicable when compiling for the base ARMv8.0
14701           instruction set.  If using a later revision, e.g. -march=armv8.1-a
14702           or -march=armv8-a+lse, the ARMv8.1-Atomics instructions will be
14703           used directly.  The same applies when using -mcpu= when the
14704           selected cpu supports the lse feature.  This option is on by
14705           default.
14706
14707       -march=name
14708           Specify the name of the target architecture and, optionally, one or
14709           more feature modifiers.  This option has the form
14710           -march=arch{+[no]feature}*.
14711
14712           The table below summarizes the permissible values for arch and the
14713           features that they enable by default:
14714
14715           arch value : Architecture : Includes by default
14716           armv8-a : Armv8-A : +fp, +simd
14717           armv8.1-a : Armv8.1-A : armv8-a, +crc, +lse, +rdma
14718           armv8.2-a : Armv8.2-A : armv8.1-a
14719           armv8.3-a : Armv8.3-A : armv8.2-a
14720           armv8.4-a : Armv8.4-A : armv8.3-a, +fp16fml, +dotprod
14721           armv8.5-a : Armv8.5-A : armv8.4-a, +sb, +ssbs, +predres
14722           armv8.6-a : Armv8.6-A : armv8.5-a, +bf16, +i8mm
14723
14724           The value native is available on native AArch64 GNU/Linux and
14725           causes the compiler to pick the architecture of the host system.
14726           This option has no effect if the compiler is unable to recognize
14727           the architecture of the host system,
14728
14729           The permissible values for feature are listed in the sub-section on
14730           aarch64-feature-modifiers,,-march and -mcpu Feature Modifiers.
14731           Where conflicting feature modifiers are specified, the right-most
14732           feature is used.
14733
14734           GCC uses name to determine what kind of instructions it can emit
14735           when generating assembly code.  If -march is specified without
14736           either of -mtune or -mcpu also being specified, the code is tuned
14737           to perform well across a range of target processors implementing
14738           the target architecture.
14739
14740       -mtune=name
14741           Specify the name of the target processor for which GCC should tune
14742           the performance of the code.  Permissible values for this option
14743           are: generic, cortex-a35, cortex-a53, cortex-a55, cortex-a57,
14744           cortex-a72, cortex-a73, cortex-a75, cortex-a76, cortex-a76ae,
14745           cortex-a77, cortex-a65, cortex-a65ae, cortex-a34, ares, exynos-m1,
14746           emag, falkor, neoverse-e1, neoverse-n1, neoverse-n2, neoverse-v1,
14747           qdf24xx, saphira, phecda, xgene1, vulcan, octeontx, octeontx81,
14748           octeontx83, octeontx2, octeontx2t98, octeontx2t96 octeontx2t93,
14749           octeontx2f95, octeontx2f95n, octeontx2f95mm, a64fx, thunderx,
14750           thunderxt88, thunderxt88p1, thunderxt81, tsv110, thunderxt83,
14751           thunderx2t99, thunderx3t110, zeus, cortex-a57.cortex-a53,
14752           cortex-a72.cortex-a53, cortex-a73.cortex-a35,
14753           cortex-a73.cortex-a53, cortex-a75.cortex-a55, cortex-a76.cortex-a55
14754           native.
14755
14756           The values cortex-a57.cortex-a53, cortex-a72.cortex-a53,
14757           cortex-a73.cortex-a35, cortex-a73.cortex-a53,
14758           cortex-a75.cortex-a55, cortex-a76.cortex-a55 specify that GCC
14759           should tune for a big.LITTLE system.
14760
14761           Additionally on native AArch64 GNU/Linux systems the value native
14762           tunes performance to the host system.  This option has no effect if
14763           the compiler is unable to recognize the processor of the host
14764           system.
14765
14766           Where none of -mtune=, -mcpu= or -march= are specified, the code is
14767           tuned to perform well across a range of target processors.
14768
14769           This option cannot be suffixed by feature modifiers.
14770
14771       -mcpu=name
14772           Specify the name of the target processor, optionally suffixed by
14773           one or more feature modifiers.  This option has the form
14774           -mcpu=cpu{+[no]feature}*, where the permissible values for cpu are
14775           the same as those available for -mtune.  The permissible values for
14776           feature are documented in the sub-section on
14777           aarch64-feature-modifiers,,-march and -mcpu Feature Modifiers.
14778           Where conflicting feature modifiers are specified, the right-most
14779           feature is used.
14780
14781           GCC uses name to determine what kind of instructions it can emit
14782           when generating assembly code (as if by -march) and to determine
14783           the target processor for which to tune for performance (as if by
14784           -mtune).  Where this option is used in conjunction with -march or
14785           -mtune, those options take precedence over the appropriate part of
14786           this option.
14787
14788       -moverride=string
14789           Override tuning decisions made by the back-end in response to a
14790           -mtune= switch.  The syntax, semantics, and accepted values for
14791           string in this option are not guaranteed to be consistent across
14792           releases.
14793
14794           This option is only intended to be useful when developing GCC.
14795
14796       -mverbose-cost-dump
14797           Enable verbose cost model dumping in the debug dump files.  This
14798           option is provided for use in debugging the compiler.
14799
14800       -mpc-relative-literal-loads
14801       -mno-pc-relative-literal-loads
14802           Enable or disable PC-relative literal loads.  With this option
14803           literal pools are accessed using a single instruction and emitted
14804           after each function.  This limits the maximum size of functions to
14805           1MB.  This is enabled by default for -mcmodel=tiny.
14806
14807       -msign-return-address=scope
14808           Select the function scope on which return address signing will be
14809           applied.  Permissible values are none, which disables return
14810           address signing, non-leaf, which enables pointer signing for
14811           functions which are not leaf functions, and all, which enables
14812           pointer signing for all functions.  The default value is none. This
14813           option has been deprecated by -mbranch-protection.
14814
14815       -mbranch-protection=none|standard|pac-ret[+leaf+b-key]|bti
14816           Select the branch protection features to use.  none is the default
14817           and turns off all types of branch protection.  standard turns on
14818           all types of branch protection features.  If a feature has
14819           additional tuning options, then standard sets it to its standard
14820           level.  pac-ret[+leaf] turns on return address signing to its
14821           standard level: signing functions that save the return address to
14822           memory (non-leaf functions will practically always do this) using
14823           the a-key.  The optional argument leaf can be used to extend the
14824           signing to include leaf functions.  The optional argument b-key can
14825           be used to sign the functions with the B-key instead of the A-key.
14826           bti turns on branch target identification mechanism.
14827
14828       -mharden-sls=opts
14829           Enable compiler hardening against straight line speculation (SLS).
14830           opts is a comma-separated list of the following options:
14831
14832           retbr
14833           blr
14834
14835           In addition, -mharden-sls=all enables all SLS hardening while
14836           -mharden-sls=none disables all SLS hardening.
14837
14838       -msve-vector-bits=bits
14839           Specify the number of bits in an SVE vector register.  This option
14840           only has an effect when SVE is enabled.
14841
14842           GCC supports two forms of SVE code generation: "vector-length
14843           agnostic" output that works with any size of vector register and
14844           "vector-length specific" output that allows GCC to make assumptions
14845           about the vector length when it is useful for optimization reasons.
14846           The possible values of bits are: scalable, 128, 256, 512, 1024 and
14847           2048.  Specifying scalable selects vector-length agnostic output.
14848           At present -msve-vector-bits=128 also generates vector-length
14849           agnostic output for big-endian targets.  All other values generate
14850           vector-length specific code.  The behavior of these values may
14851           change in future releases and no value except scalable should be
14852           relied on for producing code that is portable across different
14853           hardware SVE vector lengths.
14854
14855           The default is -msve-vector-bits=scalable, which produces vector-
14856           length agnostic code.
14857
14858       -march and -mcpu Feature Modifiers
14859
14860       Feature modifiers used with -march and -mcpu can be any of the
14861       following and their inverses nofeature:
14862
14863       crc Enable CRC extension.  This is on by default for -march=armv8.1-a.
14864
14865       crypto
14866           Enable Crypto extension.  This also enables Advanced SIMD and
14867           floating-point instructions.
14868
14869       fp  Enable floating-point instructions.  This is on by default for all
14870           possible values for options -march and -mcpu.
14871
14872       simd
14873           Enable Advanced SIMD instructions.  This also enables floating-
14874           point instructions.  This is on by default for all possible values
14875           for options -march and -mcpu.
14876
14877       sve Enable Scalable Vector Extension instructions.  This also enables
14878           Advanced SIMD and floating-point instructions.
14879
14880       lse Enable Large System Extension instructions.  This is on by default
14881           for -march=armv8.1-a.
14882
14883       rdma
14884           Enable Round Double Multiply Accumulate instructions.  This is on
14885           by default for -march=armv8.1-a.
14886
14887       fp16
14888           Enable FP16 extension.  This also enables floating-point
14889           instructions.
14890
14891       fp16fml
14892           Enable FP16 fmla extension.  This also enables FP16 extensions and
14893           floating-point instructions. This option is enabled by default for
14894           -march=armv8.4-a. Use of this option with architectures prior to
14895           Armv8.2-A is not supported.
14896
14897       rcpc
14898           Enable the RcPc extension.  This does not change code generation
14899           from GCC, but is passed on to the assembler, enabling inline asm
14900           statements to use instructions from the RcPc extension.
14901
14902       dotprod
14903           Enable the Dot Product extension.  This also enables Advanced SIMD
14904           instructions.
14905
14906       aes Enable the Armv8-a aes and pmull crypto extension.  This also
14907           enables Advanced SIMD instructions.
14908
14909       sha2
14910           Enable the Armv8-a sha2 crypto extension.  This also enables
14911           Advanced SIMD instructions.
14912
14913       sha3
14914           Enable the sha512 and sha3 crypto extension.  This also enables
14915           Advanced SIMD instructions. Use of this option with architectures
14916           prior to Armv8.2-A is not supported.
14917
14918       sm4 Enable the sm3 and sm4 crypto extension.  This also enables
14919           Advanced SIMD instructions.  Use of this option with architectures
14920           prior to Armv8.2-A is not supported.
14921
14922       profile
14923           Enable the Statistical Profiling extension.  This option is only to
14924           enable the extension at the assembler level and does not affect
14925           code generation.
14926
14927       rng Enable the Armv8.5-a Random Number instructions.  This option is
14928           only to enable the extension at the assembler level and does not
14929           affect code generation.
14930
14931       memtag
14932           Enable the Armv8.5-a Memory Tagging Extensions.  Use of this option
14933           with architectures prior to Armv8.5-A is not supported.
14934
14935       sb  Enable the Armv8-a Speculation Barrier instruction.  This option is
14936           only to enable the extension at the assembler level and does not
14937           affect code generation.  This option is enabled by default for
14938           -march=armv8.5-a.
14939
14940       ssbs
14941           Enable the Armv8-a Speculative Store Bypass Safe instruction.  This
14942           option is only to enable the extension at the assembler level and
14943           does not affect code generation.  This option is enabled by default
14944           for -march=armv8.5-a.
14945
14946       predres
14947           Enable the Armv8-a Execution and Data Prediction Restriction
14948           instructions.  This option is only to enable the extension at the
14949           assembler level and does not affect code generation.  This option
14950           is enabled by default for -march=armv8.5-a.
14951
14952       sve2
14953           Enable the Armv8-a Scalable Vector Extension 2.  This also enables
14954           SVE instructions.
14955
14956       sve2-bitperm
14957           Enable SVE2 bitperm instructions.  This also enables SVE2
14958           instructions.
14959
14960       sve2-sm4
14961           Enable SVE2 sm4 instructions.  This also enables SVE2 instructions.
14962
14963       sve2-aes
14964           Enable SVE2 aes instructions.  This also enables SVE2 instructions.
14965
14966       sve2-sha3
14967           Enable SVE2 sha3 instructions.  This also enables SVE2
14968           instructions.
14969
14970       tme Enable the Transactional Memory Extension.
14971
14972       i8mm
14973           Enable 8-bit Integer Matrix Multiply instructions.  This also
14974           enables Advanced SIMD and floating-point instructions.  This option
14975           is enabled by default for -march=armv8.6-a.  Use of this option
14976           with architectures prior to Armv8.2-A is not supported.
14977
14978       f32mm
14979           Enable 32-bit Floating point Matrix Multiply instructions.  This
14980           also enables SVE instructions.  Use of this option with
14981           architectures prior to Armv8.2-A is not supported.
14982
14983       f64mm
14984           Enable 64-bit Floating point Matrix Multiply instructions.  This
14985           also enables SVE instructions.  Use of this option with
14986           architectures prior to Armv8.2-A is not supported.
14987
14988       bf16
14989           Enable brain half-precision floating-point instructions.  This also
14990           enables Advanced SIMD and floating-point instructions.  This option
14991           is enabled by default for -march=armv8.6-a.  Use of this option
14992           with architectures prior to Armv8.2-A is not supported.
14993
14994       Feature crypto implies aes, sha2, and simd, which implies fp.
14995       Conversely, nofp implies nosimd, which implies nocrypto, noaes and
14996       nosha2.
14997
14998   Adapteva Epiphany Options
14999       These -m options are defined for Adapteva Epiphany:
15000
15001       -mhalf-reg-file
15002           Don't allocate any register in the range "r32"..."r63".  That
15003           allows code to run on hardware variants that lack these registers.
15004
15005       -mprefer-short-insn-regs
15006           Preferentially allocate registers that allow short instruction
15007           generation.  This can result in increased instruction count, so
15008           this may either reduce or increase overall code size.
15009
15010       -mbranch-cost=num
15011           Set the cost of branches to roughly num "simple" instructions.
15012           This cost is only a heuristic and is not guaranteed to produce
15013           consistent results across releases.
15014
15015       -mcmove
15016           Enable the generation of conditional moves.
15017
15018       -mnops=num
15019           Emit num NOPs before every other generated instruction.
15020
15021       -mno-soft-cmpsf
15022           For single-precision floating-point comparisons, emit an "fsub"
15023           instruction and test the flags.  This is faster than a software
15024           comparison, but can get incorrect results in the presence of NaNs,
15025           or when two different small numbers are compared such that their
15026           difference is calculated as zero.  The default is -msoft-cmpsf,
15027           which uses slower, but IEEE-compliant, software comparisons.
15028
15029       -mstack-offset=num
15030           Set the offset between the top of the stack and the stack pointer.
15031           E.g., a value of 8 means that the eight bytes in the range
15032           "sp+0...sp+7" can be used by leaf functions without stack
15033           allocation.  Values other than 8 or 16 are untested and unlikely to
15034           work.  Note also that this option changes the ABI; compiling a
15035           program with a different stack offset than the libraries have been
15036           compiled with generally does not work.  This option can be useful
15037           if you want to evaluate if a different stack offset would give you
15038           better code, but to actually use a different stack offset to build
15039           working programs, it is recommended to configure the toolchain with
15040           the appropriate --with-stack-offset=num option.
15041
15042       -mno-round-nearest
15043           Make the scheduler assume that the rounding mode has been set to
15044           truncating.  The default is -mround-nearest.
15045
15046       -mlong-calls
15047           If not otherwise specified by an attribute, assume all calls might
15048           be beyond the offset range of the "b" / "bl" instructions, and
15049           therefore load the function address into a register before
15050           performing a (otherwise direct) call.  This is the default.
15051
15052       -mshort-calls
15053           If not otherwise specified by an attribute, assume all direct calls
15054           are in the range of the "b" / "bl" instructions, so use these
15055           instructions for direct calls.  The default is -mlong-calls.
15056
15057       -msmall16
15058           Assume addresses can be loaded as 16-bit unsigned values.  This
15059           does not apply to function addresses for which -mlong-calls
15060           semantics are in effect.
15061
15062       -mfp-mode=mode
15063           Set the prevailing mode of the floating-point unit.  This
15064           determines the floating-point mode that is provided and expected at
15065           function call and return time.  Making this mode match the mode you
15066           predominantly need at function start can make your programs smaller
15067           and faster by avoiding unnecessary mode switches.
15068
15069           mode can be set to one the following values:
15070
15071           caller
15072               Any mode at function entry is valid, and retained or restored
15073               when the function returns, and when it calls other functions.
15074               This mode is useful for compiling libraries or other
15075               compilation units you might want to incorporate into different
15076               programs with different prevailing FPU modes, and the
15077               convenience of being able to use a single object file outweighs
15078               the size and speed overhead for any extra mode switching that
15079               might be needed, compared with what would be needed with a more
15080               specific choice of prevailing FPU mode.
15081
15082           truncate
15083               This is the mode used for floating-point calculations with
15084               truncating (i.e. round towards zero) rounding mode.  That
15085               includes conversion from floating point to integer.
15086
15087           round-nearest
15088               This is the mode used for floating-point calculations with
15089               round-to-nearest-or-even rounding mode.
15090
15091           int This is the mode used to perform integer calculations in the
15092               FPU, e.g.  integer multiply, or integer multiply-and-
15093               accumulate.
15094
15095           The default is -mfp-mode=caller
15096
15097       -mno-split-lohi
15098       -mno-postinc
15099       -mno-postmodify
15100           Code generation tweaks that disable, respectively, splitting of
15101           32-bit loads, generation of post-increment addresses, and
15102           generation of post-modify addresses.  The defaults are msplit-lohi,
15103           -mpost-inc, and -mpost-modify.
15104
15105       -mnovect-double
15106           Change the preferred SIMD mode to SImode.  The default is
15107           -mvect-double, which uses DImode as preferred SIMD mode.
15108
15109       -max-vect-align=num
15110           The maximum alignment for SIMD vector mode types.  num may be 4 or
15111           8.  The default is 8.  Note that this is an ABI change, even though
15112           many library function interfaces are unaffected if they don't use
15113           SIMD vector modes in places that affect size and/or alignment of
15114           relevant types.
15115
15116       -msplit-vecmove-early
15117           Split vector moves into single word moves before reload.  In theory
15118           this can give better register allocation, but so far the reverse
15119           seems to be generally the case.
15120
15121       -m1reg-reg
15122           Specify a register to hold the constant -1, which makes loading
15123           small negative constants and certain bitmasks faster.  Allowable
15124           values for reg are r43 and r63, which specify use of that register
15125           as a fixed register, and none, which means that no register is used
15126           for this purpose.  The default is -m1reg-none.
15127
15128   AMD GCN Options
15129       These options are defined specifically for the AMD GCN port.
15130
15131       -march=gpu
15132       -mtune=gpu
15133           Set architecture type or tuning for gpu. Supported values for gpu
15134           are
15135
15136           fiji
15137               Compile for GCN3 Fiji devices (gfx803).
15138
15139           gfx900
15140               Compile for GCN5 Vega 10 devices (gfx900).
15141
15142           gfx906
15143               Compile for GCN5 Vega 20 devices (gfx906).
15144
15145       -mstack-size=bytes
15146           Specify how many bytes of stack space will be requested for each
15147           GPU thread (wave-front).  Beware that there may be many threads and
15148           limited memory available.  The size of the stack allocation may
15149           also have an impact on run-time performance.  The default is 32KB
15150           when using OpenACC or OpenMP, and 1MB otherwise.
15151
15152   ARC Options
15153       The following options control the architecture variant for which code
15154       is being compiled:
15155
15156       -mbarrel-shifter
15157           Generate instructions supported by barrel shifter.  This is the
15158           default unless -mcpu=ARC601 or -mcpu=ARCEM is in effect.
15159
15160       -mjli-always
15161           Force to call a function using jli_s instruction.  This option is
15162           valid only for ARCv2 architecture.
15163
15164       -mcpu=cpu
15165           Set architecture type, register usage, and instruction scheduling
15166           parameters for cpu.  There are also shortcut alias options
15167           available for backward compatibility and convenience.  Supported
15168           values for cpu are
15169
15170           arc600
15171               Compile for ARC600.  Aliases: -mA6, -mARC600.
15172
15173           arc601
15174               Compile for ARC601.  Alias: -mARC601.
15175
15176           arc700
15177               Compile for ARC700.  Aliases: -mA7, -mARC700.  This is the
15178               default when configured with --with-cpu=arc700.
15179
15180           arcem
15181               Compile for ARC EM.
15182
15183           archs
15184               Compile for ARC HS.
15185
15186           em  Compile for ARC EM CPU with no hardware extensions.
15187
15188           em4 Compile for ARC EM4 CPU.
15189
15190           em4_dmips
15191               Compile for ARC EM4 DMIPS CPU.
15192
15193           em4_fpus
15194               Compile for ARC EM4 DMIPS CPU with the single-precision
15195               floating-point extension.
15196
15197           em4_fpuda
15198               Compile for ARC EM4 DMIPS CPU with single-precision floating-
15199               point and double assist instructions.
15200
15201           hs  Compile for ARC HS CPU with no hardware extensions except the
15202               atomic instructions.
15203
15204           hs34
15205               Compile for ARC HS34 CPU.
15206
15207           hs38
15208               Compile for ARC HS38 CPU.
15209
15210           hs38_linux
15211               Compile for ARC HS38 CPU with all hardware extensions on.
15212
15213           arc600_norm
15214               Compile for ARC 600 CPU with "norm" instructions enabled.
15215
15216           arc600_mul32x16
15217               Compile for ARC 600 CPU with "norm" and 32x16-bit multiply
15218               instructions enabled.
15219
15220           arc600_mul64
15221               Compile for ARC 600 CPU with "norm" and "mul64"-family
15222               instructions enabled.
15223
15224           arc601_norm
15225               Compile for ARC 601 CPU with "norm" instructions enabled.
15226
15227           arc601_mul32x16
15228               Compile for ARC 601 CPU with "norm" and 32x16-bit multiply
15229               instructions enabled.
15230
15231           arc601_mul64
15232               Compile for ARC 601 CPU with "norm" and "mul64"-family
15233               instructions enabled.
15234
15235           nps400
15236               Compile for ARC 700 on NPS400 chip.
15237
15238           em_mini
15239               Compile for ARC EM minimalist configuration featuring reduced
15240               register set.
15241
15242       -mdpfp
15243       -mdpfp-compact
15244           Generate double-precision FPX instructions, tuned for the compact
15245           implementation.
15246
15247       -mdpfp-fast
15248           Generate double-precision FPX instructions, tuned for the fast
15249           implementation.
15250
15251       -mno-dpfp-lrsr
15252           Disable "lr" and "sr" instructions from using FPX extension aux
15253           registers.
15254
15255       -mea
15256           Generate extended arithmetic instructions.  Currently only "divaw",
15257           "adds", "subs", and "sat16" are supported.  Only valid for
15258           -mcpu=ARC700.
15259
15260       -mno-mpy
15261           Do not generate "mpy"-family instructions for ARC700.  This option
15262           is deprecated.
15263
15264       -mmul32x16
15265           Generate 32x16-bit multiply and multiply-accumulate instructions.
15266
15267       -mmul64
15268           Generate "mul64" and "mulu64" instructions.  Only valid for
15269           -mcpu=ARC600.
15270
15271       -mnorm
15272           Generate "norm" instructions.  This is the default if -mcpu=ARC700
15273           is in effect.
15274
15275       -mspfp
15276       -mspfp-compact
15277           Generate single-precision FPX instructions, tuned for the compact
15278           implementation.
15279
15280       -mspfp-fast
15281           Generate single-precision FPX instructions, tuned for the fast
15282           implementation.
15283
15284       -msimd
15285           Enable generation of ARC SIMD instructions via target-specific
15286           builtins.  Only valid for -mcpu=ARC700.
15287
15288       -msoft-float
15289           This option ignored; it is provided for compatibility purposes
15290           only.  Software floating-point code is emitted by default, and this
15291           default can overridden by FPX options; -mspfp, -mspfp-compact, or
15292           -mspfp-fast for single precision, and -mdpfp, -mdpfp-compact, or
15293           -mdpfp-fast for double precision.
15294
15295       -mswap
15296           Generate "swap" instructions.
15297
15298       -matomic
15299           This enables use of the locked load/store conditional extension to
15300           implement atomic memory built-in functions.  Not available for ARC
15301           6xx or ARC EM cores.
15302
15303       -mdiv-rem
15304           Enable "div" and "rem" instructions for ARCv2 cores.
15305
15306       -mcode-density
15307           Enable code density instructions for ARC EM.  This option is on by
15308           default for ARC HS.
15309
15310       -mll64
15311           Enable double load/store operations for ARC HS cores.
15312
15313       -mtp-regno=regno
15314           Specify thread pointer register number.
15315
15316       -mmpy-option=multo
15317           Compile ARCv2 code with a multiplier design option.  You can
15318           specify the option using either a string or numeric value for
15319           multo.  wlh1 is the default value.  The recognized values are:
15320
15321           0
15322           none
15323               No multiplier available.
15324
15325           1
15326           w   16x16 multiplier, fully pipelined.  The following instructions
15327               are enabled: "mpyw" and "mpyuw".
15328
15329           2
15330           wlh1
15331               32x32 multiplier, fully pipelined (1 stage).  The following
15332               instructions are additionally enabled: "mpy", "mpyu", "mpym",
15333               "mpymu", and "mpy_s".
15334
15335           3
15336           wlh2
15337               32x32 multiplier, fully pipelined (2 stages).  The following
15338               instructions are additionally enabled: "mpy", "mpyu", "mpym",
15339               "mpymu", and "mpy_s".
15340
15341           4
15342           wlh3
15343               Two 16x16 multipliers, blocking, sequential.  The following
15344               instructions are additionally enabled: "mpy", "mpyu", "mpym",
15345               "mpymu", and "mpy_s".
15346
15347           5
15348           wlh4
15349               One 16x16 multiplier, blocking, sequential.  The following
15350               instructions are additionally enabled: "mpy", "mpyu", "mpym",
15351               "mpymu", and "mpy_s".
15352
15353           6
15354           wlh5
15355               One 32x4 multiplier, blocking, sequential.  The following
15356               instructions are additionally enabled: "mpy", "mpyu", "mpym",
15357               "mpymu", and "mpy_s".
15358
15359           7
15360           plus_dmpy
15361               ARC HS SIMD support.
15362
15363           8
15364           plus_macd
15365               ARC HS SIMD support.
15366
15367           9
15368           plus_qmacw
15369               ARC HS SIMD support.
15370
15371           This option is only available for ARCv2 cores.
15372
15373       -mfpu=fpu
15374           Enables support for specific floating-point hardware extensions for
15375           ARCv2 cores.  Supported values for fpu are:
15376
15377           fpus
15378               Enables support for single-precision floating-point hardware
15379               extensions.
15380
15381           fpud
15382               Enables support for double-precision floating-point hardware
15383               extensions.  The single-precision floating-point extension is
15384               also enabled.  Not available for ARC EM.
15385
15386           fpuda
15387               Enables support for double-precision floating-point hardware
15388               extensions using double-precision assist instructions.  The
15389               single-precision floating-point extension is also enabled.
15390               This option is only available for ARC EM.
15391
15392           fpuda_div
15393               Enables support for double-precision floating-point hardware
15394               extensions using double-precision assist instructions.  The
15395               single-precision floating-point, square-root, and divide
15396               extensions are also enabled.  This option is only available for
15397               ARC EM.
15398
15399           fpuda_fma
15400               Enables support for double-precision floating-point hardware
15401               extensions using double-precision assist instructions.  The
15402               single-precision floating-point and fused multiply and add
15403               hardware extensions are also enabled.  This option is only
15404               available for ARC EM.
15405
15406           fpuda_all
15407               Enables support for double-precision floating-point hardware
15408               extensions using double-precision assist instructions.  All
15409               single-precision floating-point hardware extensions are also
15410               enabled.  This option is only available for ARC EM.
15411
15412           fpus_div
15413               Enables support for single-precision floating-point, square-
15414               root and divide hardware extensions.
15415
15416           fpud_div
15417               Enables support for double-precision floating-point, square-
15418               root and divide hardware extensions.  This option includes
15419               option fpus_div. Not available for ARC EM.
15420
15421           fpus_fma
15422               Enables support for single-precision floating-point and fused
15423               multiply and add hardware extensions.
15424
15425           fpud_fma
15426               Enables support for double-precision floating-point and fused
15427               multiply and add hardware extensions.  This option includes
15428               option fpus_fma.  Not available for ARC EM.
15429
15430           fpus_all
15431               Enables support for all single-precision floating-point
15432               hardware extensions.
15433
15434           fpud_all
15435               Enables support for all single- and double-precision floating-
15436               point hardware extensions.  Not available for ARC EM.
15437
15438       -mirq-ctrl-saved=register-range, blink, lp_count
15439           Specifies general-purposes registers that the processor
15440           automatically saves/restores on interrupt entry and exit.
15441           register-range is specified as two registers separated by a dash.
15442           The register range always starts with "r0", the upper limit is "fp"
15443           register.  blink and lp_count are optional.  This option is only
15444           valid for ARC EM and ARC HS cores.
15445
15446       -mrgf-banked-regs=number
15447           Specifies the number of registers replicated in second register
15448           bank on entry to fast interrupt.  Fast interrupts are interrupts
15449           with the highest priority level P0.  These interrupts save only PC
15450           and STATUS32 registers to avoid memory transactions during
15451           interrupt entry and exit sequences.  Use this option when you are
15452           using fast interrupts in an ARC V2 family processor.  Permitted
15453           values are 4, 8, 16, and 32.
15454
15455       -mlpc-width=width
15456           Specify the width of the "lp_count" register.  Valid values for
15457           width are 8, 16, 20, 24, 28 and 32 bits.  The default width is
15458           fixed to 32 bits.  If the width is less than 32, the compiler does
15459           not attempt to transform loops in your program to use the zero-
15460           delay loop mechanism unless it is known that the "lp_count"
15461           register can hold the required loop-counter value.  Depending on
15462           the width specified, the compiler and run-time library might
15463           continue to use the loop mechanism for various needs.  This option
15464           defines macro "__ARC_LPC_WIDTH__" with the value of width.
15465
15466       -mrf16
15467           This option instructs the compiler to generate code for a 16-entry
15468           register file.  This option defines the "__ARC_RF16__" preprocessor
15469           macro.
15470
15471       -mbranch-index
15472           Enable use of "bi" or "bih" instructions to implement jump tables.
15473
15474       The following options are passed through to the assembler, and also
15475       define preprocessor macro symbols.
15476
15477       -mdsp-packa
15478           Passed down to the assembler to enable the DSP Pack A extensions.
15479           Also sets the preprocessor symbol "__Xdsp_packa".  This option is
15480           deprecated.
15481
15482       -mdvbf
15483           Passed down to the assembler to enable the dual Viterbi butterfly
15484           extension.  Also sets the preprocessor symbol "__Xdvbf".  This
15485           option is deprecated.
15486
15487       -mlock
15488           Passed down to the assembler to enable the locked load/store
15489           conditional extension.  Also sets the preprocessor symbol
15490           "__Xlock".
15491
15492       -mmac-d16
15493           Passed down to the assembler.  Also sets the preprocessor symbol
15494           "__Xxmac_d16".  This option is deprecated.
15495
15496       -mmac-24
15497           Passed down to the assembler.  Also sets the preprocessor symbol
15498           "__Xxmac_24".  This option is deprecated.
15499
15500       -mrtsc
15501           Passed down to the assembler to enable the 64-bit time-stamp
15502           counter extension instruction.  Also sets the preprocessor symbol
15503           "__Xrtsc".  This option is deprecated.
15504
15505       -mswape
15506           Passed down to the assembler to enable the swap byte ordering
15507           extension instruction.  Also sets the preprocessor symbol
15508           "__Xswape".
15509
15510       -mtelephony
15511           Passed down to the assembler to enable dual- and single-operand
15512           instructions for telephony.  Also sets the preprocessor symbol
15513           "__Xtelephony".  This option is deprecated.
15514
15515       -mxy
15516           Passed down to the assembler to enable the XY memory extension.
15517           Also sets the preprocessor symbol "__Xxy".
15518
15519       The following options control how the assembly code is annotated:
15520
15521       -misize
15522           Annotate assembler instructions with estimated addresses.
15523
15524       -mannotate-align
15525           Explain what alignment considerations lead to the decision to make
15526           an instruction short or long.
15527
15528       The following options are passed through to the linker:
15529
15530       -marclinux
15531           Passed through to the linker, to specify use of the "arclinux"
15532           emulation.  This option is enabled by default in tool chains built
15533           for "arc-linux-uclibc" and "arceb-linux-uclibc" targets when
15534           profiling is not requested.
15535
15536       -marclinux_prof
15537           Passed through to the linker, to specify use of the "arclinux_prof"
15538           emulation.  This option is enabled by default in tool chains built
15539           for "arc-linux-uclibc" and "arceb-linux-uclibc" targets when
15540           profiling is requested.
15541
15542       The following options control the semantics of generated code:
15543
15544       -mlong-calls
15545           Generate calls as register indirect calls, thus providing access to
15546           the full 32-bit address range.
15547
15548       -mmedium-calls
15549           Don't use less than 25-bit addressing range for calls, which is the
15550           offset available for an unconditional branch-and-link instruction.
15551           Conditional execution of function calls is suppressed, to allow use
15552           of the 25-bit range, rather than the 21-bit range with conditional
15553           branch-and-link.  This is the default for tool chains built for
15554           "arc-linux-uclibc" and "arceb-linux-uclibc" targets.
15555
15556       -G num
15557           Put definitions of externally-visible data in a small data section
15558           if that data is no bigger than num bytes.  The default value of num
15559           is 4 for any ARC configuration, or 8 when we have double load/store
15560           operations.
15561
15562       -mno-sdata
15563           Do not generate sdata references.  This is the default for tool
15564           chains built for "arc-linux-uclibc" and "arceb-linux-uclibc"
15565           targets.
15566
15567       -mvolatile-cache
15568           Use ordinarily cached memory accesses for volatile references.
15569           This is the default.
15570
15571       -mno-volatile-cache
15572           Enable cache bypass for volatile references.
15573
15574       The following options fine tune code generation:
15575
15576       -malign-call
15577           Do alignment optimizations for call instructions.
15578
15579       -mauto-modify-reg
15580           Enable the use of pre/post modify with register displacement.
15581
15582       -mbbit-peephole
15583           Enable bbit peephole2.
15584
15585       -mno-brcc
15586           This option disables a target-specific pass in arc_reorg to
15587           generate compare-and-branch ("brcc") instructions.  It has no
15588           effect on generation of these instructions driven by the combiner
15589           pass.
15590
15591       -mcase-vector-pcrel
15592           Use PC-relative switch case tables to enable case table shortening.
15593           This is the default for -Os.
15594
15595       -mcompact-casesi
15596           Enable compact "casesi" pattern.  This is the default for -Os, and
15597           only available for ARCv1 cores.  This option is deprecated.
15598
15599       -mno-cond-exec
15600           Disable the ARCompact-specific pass to generate conditional
15601           execution instructions.
15602
15603           Due to delay slot scheduling and interactions between operand
15604           numbers, literal sizes, instruction lengths, and the support for
15605           conditional execution, the target-independent pass to generate
15606           conditional execution is often lacking, so the ARC port has kept a
15607           special pass around that tries to find more conditional execution
15608           generation opportunities after register allocation, branch
15609           shortening, and delay slot scheduling have been done.  This pass
15610           generally, but not always, improves performance and code size, at
15611           the cost of extra compilation time, which is why there is an option
15612           to switch it off.  If you have a problem with call instructions
15613           exceeding their allowable offset range because they are
15614           conditionalized, you should consider using -mmedium-calls instead.
15615
15616       -mearly-cbranchsi
15617           Enable pre-reload use of the "cbranchsi" pattern.
15618
15619       -mexpand-adddi
15620           Expand "adddi3" and "subdi3" at RTL generation time into "add.f",
15621           "adc" etc.  This option is deprecated.
15622
15623       -mindexed-loads
15624           Enable the use of indexed loads.  This can be problematic because
15625           some optimizers then assume that indexed stores exist, which is not
15626           the case.
15627
15628       -mlra
15629           Enable Local Register Allocation.  This is still experimental for
15630           ARC, so by default the compiler uses standard reload (i.e.
15631           -mno-lra).
15632
15633       -mlra-priority-none
15634           Don't indicate any priority for target registers.
15635
15636       -mlra-priority-compact
15637           Indicate target register priority for r0..r3 / r12..r15.
15638
15639       -mlra-priority-noncompact
15640           Reduce target register priority for r0..r3 / r12..r15.
15641
15642       -mmillicode
15643           When optimizing for size (using -Os), prologues and epilogues that
15644           have to save or restore a large number of registers are often
15645           shortened by using call to a special function in libgcc; this is
15646           referred to as a millicode call.  As these calls can pose
15647           performance issues, and/or cause linking issues when linking in a
15648           nonstandard way, this option is provided to turn on or off
15649           millicode call generation.
15650
15651       -mcode-density-frame
15652           This option enable the compiler to emit "enter" and "leave"
15653           instructions.  These instructions are only valid for CPUs with
15654           code-density feature.
15655
15656       -mmixed-code
15657           Tweak register allocation to help 16-bit instruction generation.
15658           This generally has the effect of decreasing the average instruction
15659           size while increasing the instruction count.
15660
15661       -mq-class
15662           Ths option is deprecated.  Enable q instruction alternatives.  This
15663           is the default for -Os.
15664
15665       -mRcq
15666           Enable Rcq constraint handling.  Most short code generation depends
15667           on this.  This is the default.
15668
15669       -mRcw
15670           Enable Rcw constraint handling.  Most ccfsm condexec mostly depends
15671           on this.  This is the default.
15672
15673       -msize-level=level
15674           Fine-tune size optimization with regards to instruction lengths and
15675           alignment.  The recognized values for level are:
15676
15677           0   No size optimization.  This level is deprecated and treated
15678               like 1.
15679
15680           1   Short instructions are used opportunistically.
15681
15682           2   In addition, alignment of loops and of code after barriers are
15683               dropped.
15684
15685           3   In addition, optional data alignment is dropped, and the option
15686               Os is enabled.
15687
15688           This defaults to 3 when -Os is in effect.  Otherwise, the behavior
15689           when this is not set is equivalent to level 1.
15690
15691       -mtune=cpu
15692           Set instruction scheduling parameters for cpu, overriding any
15693           implied by -mcpu=.
15694
15695           Supported values for cpu are
15696
15697           ARC600
15698               Tune for ARC600 CPU.
15699
15700           ARC601
15701               Tune for ARC601 CPU.
15702
15703           ARC700
15704               Tune for ARC700 CPU with standard multiplier block.
15705
15706           ARC700-xmac
15707               Tune for ARC700 CPU with XMAC block.
15708
15709           ARC725D
15710               Tune for ARC725D CPU.
15711
15712           ARC750D
15713               Tune for ARC750D CPU.
15714
15715       -mmultcost=num
15716           Cost to assume for a multiply instruction, with 4 being equal to a
15717           normal instruction.
15718
15719       -munalign-prob-threshold=probability
15720           Set probability threshold for unaligning branches.  When tuning for
15721           ARC700 and optimizing for speed, branches without filled delay slot
15722           are preferably emitted unaligned and long, unless profiling
15723           indicates that the probability for the branch to be taken is below
15724           probability.  The default is (REG_BR_PROB_BASE/2), i.e. 5000.
15725
15726       The following options are maintained for backward compatibility, but
15727       are now deprecated and will be removed in a future release:
15728
15729       -margonaut
15730           Obsolete FPX.
15731
15732       -mbig-endian
15733       -EB Compile code for big-endian targets.  Use of these options is now
15734           deprecated.  Big-endian code is supported by configuring GCC to
15735           build "arceb-elf32" and "arceb-linux-uclibc" targets, for which big
15736           endian is the default.
15737
15738       -mlittle-endian
15739       -EL Compile code for little-endian targets.  Use of these options is
15740           now deprecated.  Little-endian code is supported by configuring GCC
15741           to build "arc-elf32" and "arc-linux-uclibc" targets, for which
15742           little endian is the default.
15743
15744       -mbarrel_shifter
15745           Replaced by -mbarrel-shifter.
15746
15747       -mdpfp_compact
15748           Replaced by -mdpfp-compact.
15749
15750       -mdpfp_fast
15751           Replaced by -mdpfp-fast.
15752
15753       -mdsp_packa
15754           Replaced by -mdsp-packa.
15755
15756       -mEA
15757           Replaced by -mea.
15758
15759       -mmac_24
15760           Replaced by -mmac-24.
15761
15762       -mmac_d16
15763           Replaced by -mmac-d16.
15764
15765       -mspfp_compact
15766           Replaced by -mspfp-compact.
15767
15768       -mspfp_fast
15769           Replaced by -mspfp-fast.
15770
15771       -mtune=cpu
15772           Values arc600, arc601, arc700 and arc700-xmac for cpu are replaced
15773           by ARC600, ARC601, ARC700 and ARC700-xmac respectively.
15774
15775       -multcost=num
15776           Replaced by -mmultcost.
15777
15778   ARM Options
15779       These -m options are defined for the ARM port:
15780
15781       -mabi=name
15782           Generate code for the specified ABI.  Permissible values are: apcs-
15783           gnu, atpcs, aapcs, aapcs-linux and iwmmxt.
15784
15785       -mapcs-frame
15786           Generate a stack frame that is compliant with the ARM Procedure
15787           Call Standard for all functions, even if this is not strictly
15788           necessary for correct execution of the code.  Specifying
15789           -fomit-frame-pointer with this option causes the stack frames not
15790           to be generated for leaf functions.  The default is
15791           -mno-apcs-frame.  This option is deprecated.
15792
15793       -mapcs
15794           This is a synonym for -mapcs-frame and is deprecated.
15795
15796       -mthumb-interwork
15797           Generate code that supports calling between the ARM and Thumb
15798           instruction sets.  Without this option, on pre-v5 architectures,
15799           the two instruction sets cannot be reliably used inside one
15800           program.  The default is -mno-thumb-interwork, since slightly
15801           larger code is generated when -mthumb-interwork is specified.  In
15802           AAPCS configurations this option is meaningless.
15803
15804       -mno-sched-prolog
15805           Prevent the reordering of instructions in the function prologue, or
15806           the merging of those instruction with the instructions in the
15807           function's body.  This means that all functions start with a
15808           recognizable set of instructions (or in fact one of a choice from a
15809           small set of different function prologues), and this information
15810           can be used to locate the start of functions inside an executable
15811           piece of code.  The default is -msched-prolog.
15812
15813       -mfloat-abi=name
15814           Specifies which floating-point ABI to use.  Permissible values are:
15815           soft, softfp and hard.
15816
15817           Specifying soft causes GCC to generate output containing library
15818           calls for floating-point operations.  softfp allows the generation
15819           of code using hardware floating-point instructions, but still uses
15820           the soft-float calling conventions.  hard allows generation of
15821           floating-point instructions and uses FPU-specific calling
15822           conventions.
15823
15824           The default depends on the specific target configuration.  Note
15825           that the hard-float and soft-float ABIs are not link-compatible;
15826           you must compile your entire program with the same ABI, and link
15827           with a compatible set of libraries.
15828
15829       -mgeneral-regs-only
15830           Generate code which uses only the general-purpose registers.  This
15831           will prevent the compiler from using floating-point and Advanced
15832           SIMD registers but will not impose any restrictions on the
15833           assembler.
15834
15835       -mlittle-endian
15836           Generate code for a processor running in little-endian mode.  This
15837           is the default for all standard configurations.
15838
15839       -mbig-endian
15840           Generate code for a processor running in big-endian mode; the
15841           default is to compile code for a little-endian processor.
15842
15843       -mbe8
15844       -mbe32
15845           When linking a big-endian image select between BE8 and BE32
15846           formats.  The option has no effect for little-endian images and is
15847           ignored.  The default is dependent on the selected target
15848           architecture.  For ARMv6 and later architectures the default is
15849           BE8, for older architectures the default is BE32.  BE32 format has
15850           been deprecated by ARM.
15851
15852       -march=name[+extension...]
15853           This specifies the name of the target ARM architecture.  GCC uses
15854           this name to determine what kind of instructions it can emit when
15855           generating assembly code.  This option can be used in conjunction
15856           with or instead of the -mcpu= option.
15857
15858           Permissible names are: armv4t, armv5t, armv5te, armv6, armv6j,
15859           armv6k, armv6kz, armv6t2, armv6z, armv6zk, armv7, armv7-a, armv7ve,
15860           armv8-a, armv8.1-a, armv8.2-a, armv8.3-a, armv8.4-a, armv8.5-a,
15861           armv8.6-a, armv7-r, armv8-r, armv6-m, armv6s-m, armv7-m, armv7e-m,
15862           armv8-m.base, armv8-m.main, armv8.1-m.main, iwmmxt and iwmmxt2.
15863
15864           Additionally, the following architectures, which lack support for
15865           the Thumb execution state, are recognized but support is
15866           deprecated: armv4.
15867
15868           Many of the architectures support extensions.  These can be added
15869           by appending +extension to the architecture name.  Extension
15870           options are processed in order and capabilities accumulate.  An
15871           extension will also enable any necessary base extensions upon which
15872           it depends.  For example, the +crypto extension will always enable
15873           the +simd extension.  The exception to the additive construction is
15874           for extensions that are prefixed with +no...: these extensions
15875           disable the specified option and any other extensions that may
15876           depend on the presence of that extension.
15877
15878           For example, -march=armv7-a+simd+nofp+vfpv4 is equivalent to
15879           writing -march=armv7-a+vfpv4 since the +simd option is entirely
15880           disabled by the +nofp option that follows it.
15881
15882           Most extension names are generically named, but have an effect that
15883           is dependent upon the architecture to which it is applied.  For
15884           example, the +simd option can be applied to both armv7-a and
15885           armv8-a architectures, but will enable the original ARMv7-A
15886           Advanced SIMD (Neon) extensions for armv7-a and the ARMv8-A variant
15887           for armv8-a.
15888
15889           The table below lists the supported extensions for each
15890           architecture.  Architectures not mentioned do not support any
15891           extensions.
15892
15893           armv5te
15894           armv6
15895           armv6j
15896           armv6k
15897           armv6kz
15898           armv6t2
15899           armv6z
15900           armv6zk
15901               +fp The VFPv2 floating-point instructions.  The extension
15902                   +vfpv2 can be used as an alias for this extension.
15903
15904               +nofp
15905                   Disable the floating-point instructions.
15906
15907           armv7
15908               The common subset of the ARMv7-A, ARMv7-R and ARMv7-M
15909               architectures.
15910
15911               +fp The VFPv3 floating-point instructions, with 16 double-
15912                   precision registers.  The extension +vfpv3-d16 can be used
15913                   as an alias for this extension.  Note that floating-point
15914                   is not supported by the base ARMv7-M architecture, but is
15915                   compatible with both the ARMv7-A and ARMv7-R architectures.
15916
15917               +nofp
15918                   Disable the floating-point instructions.
15919
15920           armv7-a
15921               +mp The multiprocessing extension.
15922
15923               +sec
15924                   The security extension.
15925
15926               +fp The VFPv3 floating-point instructions, with 16 double-
15927                   precision registers.  The extension +vfpv3-d16 can be used
15928                   as an alias for this extension.
15929
15930               +simd
15931                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
15932                   instructions.  The extensions +neon and +neon-vfpv3 can be
15933                   used as aliases for this extension.
15934
15935               +vfpv3
15936                   The VFPv3 floating-point instructions, with 32 double-
15937                   precision registers.
15938
15939               +vfpv3-d16-fp16
15940                   The VFPv3 floating-point instructions, with 16 double-
15941                   precision registers and the half-precision floating-point
15942                   conversion operations.
15943
15944               +vfpv3-fp16
15945                   The VFPv3 floating-point instructions, with 32 double-
15946                   precision registers and the half-precision floating-point
15947                   conversion operations.
15948
15949               +vfpv4-d16
15950                   The VFPv4 floating-point instructions, with 16 double-
15951                   precision registers.
15952
15953               +vfpv4
15954                   The VFPv4 floating-point instructions, with 32 double-
15955                   precision registers.
15956
15957               +neon-fp16
15958                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
15959                   instructions, with the half-precision floating-point
15960                   conversion operations.
15961
15962               +neon-vfpv4
15963                   The Advanced SIMD (Neon) v2 and the VFPv4 floating-point
15964                   instructions.
15965
15966               +nosimd
15967                   Disable the Advanced SIMD instructions (does not disable
15968                   floating point).
15969
15970               +nofp
15971                   Disable the floating-point and Advanced SIMD instructions.
15972
15973           armv7ve
15974               The extended version of the ARMv7-A architecture with support
15975               for virtualization.
15976
15977               +fp The VFPv4 floating-point instructions, with 16 double-
15978                   precision registers.  The extension +vfpv4-d16 can be used
15979                   as an alias for this extension.
15980
15981               +simd
15982                   The Advanced SIMD (Neon) v2 and the VFPv4 floating-point
15983                   instructions.  The extension +neon-vfpv4 can be used as an
15984                   alias for this extension.
15985
15986               +vfpv3-d16
15987                   The VFPv3 floating-point instructions, with 16 double-
15988                   precision registers.
15989
15990               +vfpv3
15991                   The VFPv3 floating-point instructions, with 32 double-
15992                   precision registers.
15993
15994               +vfpv3-d16-fp16
15995                   The VFPv3 floating-point instructions, with 16 double-
15996                   precision registers and the half-precision floating-point
15997                   conversion operations.
15998
15999               +vfpv3-fp16
16000                   The VFPv3 floating-point instructions, with 32 double-
16001                   precision registers and the half-precision floating-point
16002                   conversion operations.
16003
16004               +vfpv4-d16
16005                   The VFPv4 floating-point instructions, with 16 double-
16006                   precision registers.
16007
16008               +vfpv4
16009                   The VFPv4 floating-point instructions, with 32 double-
16010                   precision registers.
16011
16012               +neon
16013                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
16014                   instructions.  The extension +neon-vfpv3 can be used as an
16015                   alias for this extension.
16016
16017               +neon-fp16
16018                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
16019                   instructions, with the half-precision floating-point
16020                   conversion operations.
16021
16022               +nosimd
16023                   Disable the Advanced SIMD instructions (does not disable
16024                   floating point).
16025
16026               +nofp
16027                   Disable the floating-point and Advanced SIMD instructions.
16028
16029           armv8-a
16030               +crc
16031                   The Cyclic Redundancy Check (CRC) instructions.
16032
16033               +simd
16034                   The ARMv8-A Advanced SIMD and floating-point instructions.
16035
16036               +crypto
16037                   The cryptographic instructions.
16038
16039               +nocrypto
16040                   Disable the cryptographic instructions.
16041
16042               +nofp
16043                   Disable the floating-point, Advanced SIMD and cryptographic
16044                   instructions.
16045
16046               +sb Speculation Barrier Instruction.
16047
16048               +predres
16049                   Execution and Data Prediction Restriction Instructions.
16050
16051           armv8.1-a
16052               +simd
16053                   The ARMv8.1-A Advanced SIMD and floating-point
16054                   instructions.
16055
16056               +crypto
16057                   The cryptographic instructions.  This also enables the
16058                   Advanced SIMD and floating-point instructions.
16059
16060               +nocrypto
16061                   Disable the cryptographic instructions.
16062
16063               +nofp
16064                   Disable the floating-point, Advanced SIMD and cryptographic
16065                   instructions.
16066
16067               +sb Speculation Barrier Instruction.
16068
16069               +predres
16070                   Execution and Data Prediction Restriction Instructions.
16071
16072           armv8.2-a
16073           armv8.3-a
16074               +fp16
16075                   The half-precision floating-point data processing
16076                   instructions.  This also enables the Advanced SIMD and
16077                   floating-point instructions.
16078
16079               +fp16fml
16080                   The half-precision floating-point fmla extension.  This
16081                   also enables the half-precision floating-point extension
16082                   and Advanced SIMD and floating-point instructions.
16083
16084               +simd
16085                   The ARMv8.1-A Advanced SIMD and floating-point
16086                   instructions.
16087
16088               +crypto
16089                   The cryptographic instructions.  This also enables the
16090                   Advanced SIMD and floating-point instructions.
16091
16092               +dotprod
16093                   Enable the Dot Product extension.  This also enables
16094                   Advanced SIMD instructions.
16095
16096               +nocrypto
16097                   Disable the cryptographic extension.
16098
16099               +nofp
16100                   Disable the floating-point, Advanced SIMD and cryptographic
16101                   instructions.
16102
16103               +sb Speculation Barrier Instruction.
16104
16105               +predres
16106                   Execution and Data Prediction Restriction Instructions.
16107
16108               +i8mm
16109                   8-bit Integer Matrix Multiply instructions.  This also
16110                   enables Advanced SIMD and floating-point instructions.
16111
16112               +bf16
16113                   Brain half-precision floating-point instructions.  This
16114                   also enables Advanced SIMD and floating-point instructions.
16115
16116           armv8.4-a
16117               +fp16
16118                   The half-precision floating-point data processing
16119                   instructions.  This also enables the Advanced SIMD and
16120                   floating-point instructions as well as the Dot Product
16121                   extension and the half-precision floating-point fmla
16122                   extension.
16123
16124               +simd
16125                   The ARMv8.3-A Advanced SIMD and floating-point instructions
16126                   as well as the Dot Product extension.
16127
16128               +crypto
16129                   The cryptographic instructions.  This also enables the
16130                   Advanced SIMD and floating-point instructions as well as
16131                   the Dot Product extension.
16132
16133               +nocrypto
16134                   Disable the cryptographic extension.
16135
16136               +nofp
16137                   Disable the floating-point, Advanced SIMD and cryptographic
16138                   instructions.
16139
16140               +sb Speculation Barrier Instruction.
16141
16142               +predres
16143                   Execution and Data Prediction Restriction Instructions.
16144
16145               +i8mm
16146                   8-bit Integer Matrix Multiply instructions.  This also
16147                   enables Advanced SIMD and floating-point instructions.
16148
16149               +bf16
16150                   Brain half-precision floating-point instructions.  This
16151                   also enables Advanced SIMD and floating-point instructions.
16152
16153           armv8.5-a
16154               +fp16
16155                   The half-precision floating-point data processing
16156                   instructions.  This also enables the Advanced SIMD and
16157                   floating-point instructions as well as the Dot Product
16158                   extension and the half-precision floating-point fmla
16159                   extension.
16160
16161               +simd
16162                   The ARMv8.3-A Advanced SIMD and floating-point instructions
16163                   as well as the Dot Product extension.
16164
16165               +crypto
16166                   The cryptographic instructions.  This also enables the
16167                   Advanced SIMD and floating-point instructions as well as
16168                   the Dot Product extension.
16169
16170               +nocrypto
16171                   Disable the cryptographic extension.
16172
16173               +nofp
16174                   Disable the floating-point, Advanced SIMD and cryptographic
16175                   instructions.
16176
16177               +i8mm
16178                   8-bit Integer Matrix Multiply instructions.  This also
16179                   enables Advanced SIMD and floating-point instructions.
16180
16181               +bf16
16182                   Brain half-precision floating-point instructions.  This
16183                   also enables Advanced SIMD and floating-point instructions.
16184
16185           armv8.6-a
16186               +fp16
16187                   The half-precision floating-point data processing
16188                   instructions.  This also enables the Advanced SIMD and
16189                   floating-point instructions as well as the Dot Product
16190                   extension and the half-precision floating-point fmla
16191                   extension.
16192
16193               +simd
16194                   The ARMv8.3-A Advanced SIMD and floating-point instructions
16195                   as well as the Dot Product extension.
16196
16197               +crypto
16198                   The cryptographic instructions.  This also enables the
16199                   Advanced SIMD and floating-point instructions as well as
16200                   the Dot Product extension.
16201
16202               +nocrypto
16203                   Disable the cryptographic extension.
16204
16205               +nofp
16206                   Disable the floating-point, Advanced SIMD and cryptographic
16207                   instructions.
16208
16209               +i8mm
16210                   8-bit Integer Matrix Multiply instructions.  This also
16211                   enables Advanced SIMD and floating-point instructions.
16212
16213               +bf16
16214                   Brain half-precision floating-point instructions.  This
16215                   also enables Advanced SIMD and floating-point instructions.
16216
16217           armv7-r
16218               +fp.sp
16219                   The single-precision VFPv3 floating-point instructions.
16220                   The extension +vfpv3xd can be used as an alias for this
16221                   extension.
16222
16223               +fp The VFPv3 floating-point instructions with 16 double-
16224                   precision registers.  The extension +vfpv3-d16 can be used
16225                   as an alias for this extension.
16226
16227               +vfpv3xd-d16-fp16
16228                   The single-precision VFPv3 floating-point instructions with
16229                   16 double-precision registers and the half-precision
16230                   floating-point conversion operations.
16231
16232               +vfpv3-d16-fp16
16233                   The VFPv3 floating-point instructions with 16 double-
16234                   precision registers and the half-precision floating-point
16235                   conversion operations.
16236
16237               +nofp
16238                   Disable the floating-point extension.
16239
16240               +idiv
16241                   The ARM-state integer division instructions.
16242
16243               +noidiv
16244                   Disable the ARM-state integer division extension.
16245
16246           armv7e-m
16247               +fp The single-precision VFPv4 floating-point instructions.
16248
16249               +fpv5
16250                   The single-precision FPv5 floating-point instructions.
16251
16252               +fp.dp
16253                   The single- and double-precision FPv5 floating-point
16254                   instructions.
16255
16256               +nofp
16257                   Disable the floating-point extensions.
16258
16259           armv8.1-m.main
16260               +dsp
16261                   The DSP instructions.
16262
16263               +mve
16264                   The M-Profile Vector Extension (MVE) integer instructions.
16265
16266               +mve.fp
16267                   The M-Profile Vector Extension (MVE) integer and single
16268                   precision floating-point instructions.
16269
16270               +fp The single-precision floating-point instructions.
16271
16272               +fp.dp
16273                   The single- and double-precision floating-point
16274                   instructions.
16275
16276               +nofp
16277                   Disable the floating-point extension.
16278
16279               +cdecp0, +cdecp1, ... , +cdecp7
16280                   Enable the Custom Datapath Extension (CDE) on selected
16281                   coprocessors according to the numbers given in the options
16282                   in the range 0 to 7.
16283
16284           armv8-m.main
16285               +dsp
16286                   The DSP instructions.
16287
16288               +nodsp
16289                   Disable the DSP extension.
16290
16291               +fp The single-precision floating-point instructions.
16292
16293               +fp.dp
16294                   The single- and double-precision floating-point
16295                   instructions.
16296
16297               +nofp
16298                   Disable the floating-point extension.
16299
16300               +cdecp0, +cdecp1, ... , +cdecp7
16301                   Enable the Custom Datapath Extension (CDE) on selected
16302                   coprocessors according to the numbers given in the options
16303                   in the range 0 to 7.
16304
16305           armv8-r
16306               +crc
16307                   The Cyclic Redundancy Check (CRC) instructions.
16308
16309               +fp.sp
16310                   The single-precision FPv5 floating-point instructions.
16311
16312               +simd
16313                   The ARMv8-A Advanced SIMD and floating-point instructions.
16314
16315               +crypto
16316                   The cryptographic instructions.
16317
16318               +nocrypto
16319                   Disable the cryptographic instructions.
16320
16321               +nofp
16322                   Disable the floating-point, Advanced SIMD and cryptographic
16323                   instructions.
16324
16325           -march=native causes the compiler to auto-detect the architecture
16326           of the build computer.  At present, this feature is only supported
16327           on GNU/Linux, and not all architectures are recognized.  If the
16328           auto-detect is unsuccessful the option has no effect.
16329
16330       -mtune=name
16331           This option specifies the name of the target ARM processor for
16332           which GCC should tune the performance of the code.  For some ARM
16333           implementations better performance can be obtained by using this
16334           option.  Permissible names are: arm7tdmi, arm7tdmi-s, arm710t,
16335           arm720t, arm740t, strongarm, strongarm110, strongarm1100,
16336           0strongarm1110, arm8, arm810, arm9, arm9e, arm920, arm920t,
16337           arm922t, arm946e-s, arm966e-s, arm968e-s, arm926ej-s, arm940t,
16338           arm9tdmi, arm10tdmi, arm1020t, arm1026ej-s, arm10e, arm1020e,
16339           arm1022e, arm1136j-s, arm1136jf-s, mpcore, mpcorenovfp,
16340           arm1156t2-s, arm1156t2f-s, arm1176jz-s, arm1176jzf-s,
16341           generic-armv7-a, cortex-a5, cortex-a7, cortex-a8, cortex-a9,
16342           cortex-a12, cortex-a15, cortex-a17, cortex-a32, cortex-a35,
16343           cortex-a53, cortex-a55, cortex-a57, cortex-a72, cortex-a73,
16344           cortex-a75, cortex-a76, cortex-a76ae, cortex-a77, ares, cortex-r4,
16345           cortex-r4f, cortex-r5, cortex-r7, cortex-r8, cortex-r52, cortex-m0,
16346           cortex-m0plus, cortex-m1, cortex-m3, cortex-m4, cortex-m7,
16347           cortex-m23, cortex-m33, cortex-m35p, cortex-m55,
16348           cortex-m1.small-multiply, cortex-m0.small-multiply,
16349           cortex-m0plus.small-multiply, exynos-m1, marvell-pj4, neoverse-n1,
16350           neoverse-n2, neoverse-v1, xscale, iwmmxt, iwmmxt2, ep9312, fa526,
16351           fa626, fa606te, fa626te, fmp626, fa726te, xgene1.
16352
16353           Additionally, this option can specify that GCC should tune the
16354           performance of the code for a big.LITTLE system.  Permissible names
16355           are: cortex-a15.cortex-a7, cortex-a17.cortex-a7,
16356           cortex-a57.cortex-a53, cortex-a72.cortex-a53,
16357           cortex-a72.cortex-a35, cortex-a73.cortex-a53,
16358           cortex-a75.cortex-a55, cortex-a76.cortex-a55.
16359
16360           -mtune=generic-arch specifies that GCC should tune the performance
16361           for a blend of processors within architecture arch.  The aim is to
16362           generate code that run well on the current most popular processors,
16363           balancing between optimizations that benefit some CPUs in the
16364           range, and avoiding performance pitfalls of other CPUs.  The
16365           effects of this option may change in future GCC versions as CPU
16366           models come and go.
16367
16368           -mtune permits the same extension options as -mcpu, but the
16369           extension options do not affect the tuning of the generated code.
16370
16371           -mtune=native causes the compiler to auto-detect the CPU of the
16372           build computer.  At present, this feature is only supported on
16373           GNU/Linux, and not all architectures are recognized.  If the auto-
16374           detect is unsuccessful the option has no effect.
16375
16376       -mcpu=name[+extension...]
16377           This specifies the name of the target ARM processor.  GCC uses this
16378           name to derive the name of the target ARM architecture (as if
16379           specified by -march) and the ARM processor type for which to tune
16380           for performance (as if specified by -mtune).  Where this option is
16381           used in conjunction with -march or -mtune, those options take
16382           precedence over the appropriate part of this option.
16383
16384           Many of the supported CPUs implement optional architectural
16385           extensions.  Where this is so the architectural extensions are
16386           normally enabled by default.  If implementations that lack the
16387           extension exist, then the extension syntax can be used to disable
16388           those extensions that have been omitted.  For floating-point and
16389           Advanced SIMD (Neon) instructions, the settings of the options
16390           -mfloat-abi and -mfpu must also be considered: floating-point and
16391           Advanced SIMD instructions will only be used if -mfloat-abi is not
16392           set to soft; and any setting of -mfpu other than auto will override
16393           the available floating-point and SIMD extension instructions.
16394
16395           For example, cortex-a9 can be found in three major configurations:
16396           integer only, with just a floating-point unit or with floating-
16397           point and Advanced SIMD.  The default is to enable all the
16398           instructions, but the extensions +nosimd and +nofp can be used to
16399           disable just the SIMD or both the SIMD and floating-point
16400           instructions respectively.
16401
16402           Permissible names for this option are the same as those for -mtune.
16403
16404           The following extension options are common to the listed CPUs:
16405
16406           +nodsp
16407               Disable the DSP instructions on cortex-m33, cortex-m35p.
16408
16409           +nofp
16410               Disables the floating-point instructions on arm9e, arm946e-s,
16411               arm966e-s, arm968e-s, arm10e, arm1020e, arm1022e, arm926ej-s,
16412               arm1026ej-s, cortex-r5, cortex-r7, cortex-r8, cortex-m4,
16413               cortex-m7, cortex-m33 and cortex-m35p.  Disables the floating-
16414               point and SIMD instructions on generic-armv7-a, cortex-a5,
16415               cortex-a7, cortex-a8, cortex-a9, cortex-a12, cortex-a15,
16416               cortex-a17, cortex-a15.cortex-a7, cortex-a17.cortex-a7,
16417               cortex-a32, cortex-a35, cortex-a53 and cortex-a55.
16418
16419           +nofp.dp
16420               Disables the double-precision component of the floating-point
16421               instructions on cortex-r5, cortex-r7, cortex-r8, cortex-r52 and
16422               cortex-m7.
16423
16424           +nosimd
16425               Disables the SIMD (but not floating-point) instructions on
16426               generic-armv7-a, cortex-a5, cortex-a7 and cortex-a9.
16427
16428           +crypto
16429               Enables the cryptographic instructions on cortex-a32,
16430               cortex-a35, cortex-a53, cortex-a55, cortex-a57, cortex-a72,
16431               cortex-a73, cortex-a75, exynos-m1, xgene1,
16432               cortex-a57.cortex-a53, cortex-a72.cortex-a53,
16433               cortex-a73.cortex-a35, cortex-a73.cortex-a53 and
16434               cortex-a75.cortex-a55.
16435
16436           Additionally the generic-armv7-a pseudo target defaults to VFPv3
16437           with 16 double-precision registers.  It supports the following
16438           extension options: mp, sec, vfpv3-d16, vfpv3, vfpv3-d16-fp16,
16439           vfpv3-fp16, vfpv4-d16, vfpv4, neon, neon-vfpv3, neon-fp16,
16440           neon-vfpv4.  The meanings are the same as for the extensions to
16441           -march=armv7-a.
16442
16443           -mcpu=generic-arch is also permissible, and is equivalent to
16444           -march=arch -mtune=generic-arch.  See -mtune for more information.
16445
16446           -mcpu=native causes the compiler to auto-detect the CPU of the
16447           build computer.  At present, this feature is only supported on
16448           GNU/Linux, and not all architectures are recognized.  If the auto-
16449           detect is unsuccessful the option has no effect.
16450
16451       -mfpu=name
16452           This specifies what floating-point hardware (or hardware emulation)
16453           is available on the target.  Permissible names are: auto, vfpv2,
16454           vfpv3, vfpv3-fp16, vfpv3-d16, vfpv3-d16-fp16, vfpv3xd,
16455           vfpv3xd-fp16, neon-vfpv3, neon-fp16, vfpv4, vfpv4-d16, fpv4-sp-d16,
16456           neon-vfpv4, fpv5-d16, fpv5-sp-d16, fp-armv8, neon-fp-armv8 and
16457           crypto-neon-fp-armv8.  Note that neon is an alias for neon-vfpv3
16458           and vfp is an alias for vfpv2.
16459
16460           The setting auto is the default and is special.  It causes the
16461           compiler to select the floating-point and Advanced SIMD
16462           instructions based on the settings of -mcpu and -march.
16463
16464           If the selected floating-point hardware includes the NEON extension
16465           (e.g. -mfpu=neon), note that floating-point operations are not
16466           generated by GCC's auto-vectorization pass unless
16467           -funsafe-math-optimizations is also specified.  This is because
16468           NEON hardware does not fully implement the IEEE 754 standard for
16469           floating-point arithmetic (in particular denormal values are
16470           treated as zero), so the use of NEON instructions may lead to a
16471           loss of precision.
16472
16473           You can also set the fpu name at function level by using the
16474           "target("fpu=")" function attributes or pragmas.
16475
16476       -mfp16-format=name
16477           Specify the format of the "__fp16" half-precision floating-point
16478           type.  Permissible names are none, ieee, and alternative; the
16479           default is none, in which case the "__fp16" type is not defined.
16480
16481       -mstructure-size-boundary=n
16482           The sizes of all structures and unions are rounded up to a multiple
16483           of the number of bits set by this option.  Permissible values are
16484           8, 32 and 64.  The default value varies for different toolchains.
16485           For the COFF targeted toolchain the default value is 8.  A value of
16486           64 is only allowed if the underlying ABI supports it.
16487
16488           Specifying a larger number can produce faster, more efficient code,
16489           but can also increase the size of the program.  Different values
16490           are potentially incompatible.  Code compiled with one value cannot
16491           necessarily expect to work with code or libraries compiled with
16492           another value, if they exchange information using structures or
16493           unions.
16494
16495           This option is deprecated.
16496
16497       -mabort-on-noreturn
16498           Generate a call to the function "abort" at the end of a "noreturn"
16499           function.  It is executed if the function tries to return.
16500
16501       -mlong-calls
16502       -mno-long-calls
16503           Tells the compiler to perform function calls by first loading the
16504           address of the function into a register and then performing a
16505           subroutine call on this register.  This switch is needed if the
16506           target function lies outside of the 64-megabyte addressing range of
16507           the offset-based version of subroutine call instruction.
16508
16509           Even if this switch is enabled, not all function calls are turned
16510           into long calls.  The heuristic is that static functions, functions
16511           that have the "short_call" attribute, functions that are inside the
16512           scope of a "#pragma no_long_calls" directive, and functions whose
16513           definitions have already been compiled within the current
16514           compilation unit are not turned into long calls.  The exceptions to
16515           this rule are that weak function definitions, functions with the
16516           "long_call" attribute or the "section" attribute, and functions
16517           that are within the scope of a "#pragma long_calls" directive are
16518           always turned into long calls.
16519
16520           This feature is not enabled by default.  Specifying -mno-long-calls
16521           restores the default behavior, as does placing the function calls
16522           within the scope of a "#pragma long_calls_off" directive.  Note
16523           these switches have no effect on how the compiler generates code to
16524           handle function calls via function pointers.
16525
16526       -msingle-pic-base
16527           Treat the register used for PIC addressing as read-only, rather
16528           than loading it in the prologue for each function.  The runtime
16529           system is responsible for initializing this register with an
16530           appropriate value before execution begins.
16531
16532       -mpic-register=reg
16533           Specify the register to be used for PIC addressing.  For standard
16534           PIC base case, the default is any suitable register determined by
16535           compiler.  For single PIC base case, the default is R9 if target is
16536           EABI based or stack-checking is enabled, otherwise the default is
16537           R10.
16538
16539       -mpic-data-is-text-relative
16540           Assume that the displacement between the text and data segments is
16541           fixed at static link time.  This permits using PC-relative
16542           addressing operations to access data known to be in the data
16543           segment.  For non-VxWorks RTP targets, this option is enabled by
16544           default.  When disabled on such targets, it will enable
16545           -msingle-pic-base by default.
16546
16547       -mpoke-function-name
16548           Write the name of each function into the text section, directly
16549           preceding the function prologue.  The generated code is similar to
16550           this:
16551
16552                        t0
16553                            .ascii "arm_poke_function_name", 0
16554                            .align
16555                        t1
16556                            .word 0xff000000 + (t1 - t0)
16557                        arm_poke_function_name
16558                            mov     ip, sp
16559                            stmfd   sp!, {fp, ip, lr, pc}
16560                            sub     fp, ip, #4
16561
16562           When performing a stack backtrace, code can inspect the value of
16563           "pc" stored at "fp + 0".  If the trace function then looks at
16564           location "pc - 12" and the top 8 bits are set, then we know that
16565           there is a function name embedded immediately preceding this
16566           location and has length "((pc[-3]) & 0xff000000)".
16567
16568       -mthumb
16569       -marm
16570           Select between generating code that executes in ARM and Thumb
16571           states.  The default for most configurations is to generate code
16572           that executes in ARM state, but the default can be changed by
16573           configuring GCC with the --with-mode=state configure option.
16574
16575           You can also override the ARM and Thumb mode for each function by
16576           using the "target("thumb")" and "target("arm")" function attributes
16577           or pragmas.
16578
16579       -mflip-thumb
16580           Switch ARM/Thumb modes on alternating functions.  This option is
16581           provided for regression testing of mixed Thumb/ARM code generation,
16582           and is not intended for ordinary use in compiling code.
16583
16584       -mtpcs-frame
16585           Generate a stack frame that is compliant with the Thumb Procedure
16586           Call Standard for all non-leaf functions.  (A leaf function is one
16587           that does not call any other functions.)  The default is
16588           -mno-tpcs-frame.
16589
16590       -mtpcs-leaf-frame
16591           Generate a stack frame that is compliant with the Thumb Procedure
16592           Call Standard for all leaf functions.  (A leaf function is one that
16593           does not call any other functions.)  The default is
16594           -mno-apcs-leaf-frame.
16595
16596       -mcallee-super-interworking
16597           Gives all externally visible functions in the file being compiled
16598           an ARM instruction set header which switches to Thumb mode before
16599           executing the rest of the function.  This allows these functions to
16600           be called from non-interworking code.  This option is not valid in
16601           AAPCS configurations because interworking is enabled by default.
16602
16603       -mcaller-super-interworking
16604           Allows calls via function pointers (including virtual functions) to
16605           execute correctly regardless of whether the target code has been
16606           compiled for interworking or not.  There is a small overhead in the
16607           cost of executing a function pointer if this option is enabled.
16608           This option is not valid in AAPCS configurations because
16609           interworking is enabled by default.
16610
16611       -mtp=name
16612           Specify the access model for the thread local storage pointer.  The
16613           valid models are soft, which generates calls to "__aeabi_read_tp",
16614           cp15, which fetches the thread pointer from "cp15" directly
16615           (supported in the arm6k architecture), and auto, which uses the
16616           best available method for the selected processor.  The default
16617           setting is auto.
16618
16619       -mtls-dialect=dialect
16620           Specify the dialect to use for accessing thread local storage.  Two
16621           dialects are supported---gnu and gnu2.  The gnu dialect selects the
16622           original GNU scheme for supporting local and global dynamic TLS
16623           models.  The gnu2 dialect selects the GNU descriptor scheme, which
16624           provides better performance for shared libraries.  The GNU
16625           descriptor scheme is compatible with the original scheme, but does
16626           require new assembler, linker and library support.  Initial and
16627           local exec TLS models are unaffected by this option and always use
16628           the original scheme.
16629
16630       -mword-relocations
16631           Only generate absolute relocations on word-sized values (i.e.
16632           R_ARM_ABS32).  This is enabled by default on targets (uClinux,
16633           SymbianOS) where the runtime loader imposes this restriction, and
16634           when -fpic or -fPIC is specified. This option conflicts with
16635           -mslow-flash-data.
16636
16637       -mfix-cortex-m3-ldrd
16638           Some Cortex-M3 cores can cause data corruption when "ldrd"
16639           instructions with overlapping destination and base registers are
16640           used.  This option avoids generating these instructions.  This
16641           option is enabled by default when -mcpu=cortex-m3 is specified.
16642
16643       -munaligned-access
16644       -mno-unaligned-access
16645           Enables (or disables) reading and writing of 16- and 32- bit values
16646           from addresses that are not 16- or 32- bit aligned.  By default
16647           unaligned access is disabled for all pre-ARMv6, all ARMv6-M and for
16648           ARMv8-M Baseline architectures, and enabled for all other
16649           architectures.  If unaligned access is not enabled then words in
16650           packed data structures are accessed a byte at a time.
16651
16652           The ARM attribute "Tag_CPU_unaligned_access" is set in the
16653           generated object file to either true or false, depending upon the
16654           setting of this option.  If unaligned access is enabled then the
16655           preprocessor symbol "__ARM_FEATURE_UNALIGNED" is also defined.
16656
16657       -mneon-for-64bits
16658           This option is deprecated and has no effect.
16659
16660       -mslow-flash-data
16661           Assume loading data from flash is slower than fetching instruction.
16662           Therefore literal load is minimized for better performance.  This
16663           option is only supported when compiling for ARMv7 M-profile and off
16664           by default. It conflicts with -mword-relocations.
16665
16666       -masm-syntax-unified
16667           Assume inline assembler is using unified asm syntax.  The default
16668           is currently off which implies divided syntax.  This option has no
16669           impact on Thumb2. However, this may change in future releases of
16670           GCC.  Divided syntax should be considered deprecated.
16671
16672       -mrestrict-it
16673           Restricts generation of IT blocks to conform to the rules of
16674           ARMv8-A.  IT blocks can only contain a single 16-bit instruction
16675           from a select set of instructions. This option is on by default for
16676           ARMv8-A Thumb mode.
16677
16678       -mprint-tune-info
16679           Print CPU tuning information as comment in assembler file.  This is
16680           an option used only for regression testing of the compiler and not
16681           intended for ordinary use in compiling code.  This option is
16682           disabled by default.
16683
16684       -mverbose-cost-dump
16685           Enable verbose cost model dumping in the debug dump files.  This
16686           option is provided for use in debugging the compiler.
16687
16688       -mpure-code
16689           Do not allow constant data to be placed in code sections.
16690           Additionally, when compiling for ELF object format give all text
16691           sections the ELF processor-specific section attribute
16692           "SHF_ARM_PURECODE".  This option is only available when generating
16693           non-pic code for M-profile targets.
16694
16695       -mcmse
16696           Generate secure code as per the "ARMv8-M Security Extensions:
16697           Requirements on Development Tools Engineering Specification", which
16698           can be found on
16699           <https://developer.arm.com/documentation/ecm0359818/latest/>.
16700
16701       -mfdpic
16702       -mno-fdpic
16703           Select the FDPIC ABI, which uses 64-bit function descriptors to
16704           represent pointers to functions.  When the compiler is configured
16705           for "arm-*-uclinuxfdpiceabi" targets, this option is on by default
16706           and implies -fPIE if none of the PIC/PIE-related options is
16707           provided.  On other targets, it only enables the FDPIC-specific
16708           code generation features, and the user should explicitly provide
16709           the PIC/PIE-related options as needed.
16710
16711           Note that static linking is not supported because it would still
16712           involve the dynamic linker when the program self-relocates.  If
16713           such behavior is acceptable, use -static and -Wl,-dynamic-linker
16714           options.
16715
16716           The opposite -mno-fdpic option is useful (and required) to build
16717           the Linux kernel using the same ("arm-*-uclinuxfdpiceabi")
16718           toolchain as the one used to build the userland programs.
16719
16720   AVR Options
16721       These options are defined for AVR implementations:
16722
16723       -mmcu=mcu
16724           Specify Atmel AVR instruction set architectures (ISA) or MCU type.
16725
16726           The default for this option is@tie{}avr2.
16727
16728           GCC supports the following AVR devices and ISAs:
16729
16730           "avr2"
16731               "Classic" devices with up to 8@tie{}KiB of program memory.
16732               mcu@tie{}= "attiny22", "attiny26", "at90s2313", "at90s2323",
16733               "at90s2333", "at90s2343", "at90s4414", "at90s4433",
16734               "at90s4434", "at90c8534", "at90s8515", "at90s8535".
16735
16736           "avr25"
16737               "Classic" devices with up to 8@tie{}KiB of program memory and
16738               with the "MOVW" instruction.  mcu@tie{}= "attiny13",
16739               "attiny13a", "attiny24", "attiny24a", "attiny25", "attiny261",
16740               "attiny261a", "attiny2313", "attiny2313a", "attiny43u",
16741               "attiny44", "attiny44a", "attiny45", "attiny48", "attiny441",
16742               "attiny461", "attiny461a", "attiny4313", "attiny84",
16743               "attiny84a", "attiny85", "attiny87", "attiny88", "attiny828",
16744               "attiny841", "attiny861", "attiny861a", "ata5272", "ata6616c",
16745               "at86rf401".
16746
16747           "avr3"
16748               "Classic" devices with 16@tie{}KiB up to 64@tie{}KiB of program
16749               memory.  mcu@tie{}= "at76c711", "at43usb355".
16750
16751           "avr31"
16752               "Classic" devices with 128@tie{}KiB of program memory.
16753               mcu@tie{}= "atmega103", "at43usb320".
16754
16755           "avr35"
16756               "Classic" devices with 16@tie{}KiB up to 64@tie{}KiB of program
16757               memory and with the "MOVW" instruction.  mcu@tie{}=
16758               "attiny167", "attiny1634", "atmega8u2", "atmega16u2",
16759               "atmega32u2", "ata5505", "ata6617c", "ata664251", "at90usb82",
16760               "at90usb162".
16761
16762           "avr4"
16763               "Enhanced" devices with up to 8@tie{}KiB of program memory.
16764               mcu@tie{}= "atmega48", "atmega48a", "atmega48p", "atmega48pa",
16765               "atmega48pb", "atmega8", "atmega8a", "atmega8hva", "atmega88",
16766               "atmega88a", "atmega88p", "atmega88pa", "atmega88pb",
16767               "atmega8515", "atmega8535", "ata6285", "ata6286", "ata6289",
16768               "ata6612c", "at90pwm1", "at90pwm2", "at90pwm2b", "at90pwm3",
16769               "at90pwm3b", "at90pwm81".
16770
16771           "avr5"
16772               "Enhanced" devices with 16@tie{}KiB up to 64@tie{}KiB of
16773               program memory.  mcu@tie{}= "atmega16", "atmega16a",
16774               "atmega16hva", "atmega16hva2", "atmega16hvb",
16775               "atmega16hvbrevb", "atmega16m1", "atmega16u4", "atmega161",
16776               "atmega162", "atmega163", "atmega164a", "atmega164p",
16777               "atmega164pa", "atmega165", "atmega165a", "atmega165p",
16778               "atmega165pa", "atmega168", "atmega168a", "atmega168p",
16779               "atmega168pa", "atmega168pb", "atmega169", "atmega169a",
16780               "atmega169p", "atmega169pa", "atmega32", "atmega32a",
16781               "atmega32c1", "atmega32hvb", "atmega32hvbrevb", "atmega32m1",
16782               "atmega32u4", "atmega32u6", "atmega323", "atmega324a",
16783               "atmega324p", "atmega324pa", "atmega325", "atmega325a",
16784               "atmega325p", "atmega325pa", "atmega328", "atmega328p",
16785               "atmega328pb", "atmega329", "atmega329a", "atmega329p",
16786               "atmega329pa", "atmega3250", "atmega3250a", "atmega3250p",
16787               "atmega3250pa", "atmega3290", "atmega3290a", "atmega3290p",
16788               "atmega3290pa", "atmega406", "atmega64", "atmega64a",
16789               "atmega64c1", "atmega64hve", "atmega64hve2", "atmega64m1",
16790               "atmega64rfr2", "atmega640", "atmega644", "atmega644a",
16791               "atmega644p", "atmega644pa", "atmega644rfr2", "atmega645",
16792               "atmega645a", "atmega645p", "atmega649", "atmega649a",
16793               "atmega649p", "atmega6450", "atmega6450a", "atmega6450p",
16794               "atmega6490", "atmega6490a", "atmega6490p", "ata5795",
16795               "ata5790", "ata5790n", "ata5791", "ata6613c", "ata6614q",
16796               "ata5782", "ata5831", "ata8210", "ata8510", "ata5702m322",
16797               "at90pwm161", "at90pwm216", "at90pwm316", "at90can32",
16798               "at90can64", "at90scr100", "at90usb646", "at90usb647", "at94k",
16799               "m3000".
16800
16801           "avr51"
16802               "Enhanced" devices with 128@tie{}KiB of program memory.
16803               mcu@tie{}= "atmega128", "atmega128a", "atmega128rfa1",
16804               "atmega128rfr2", "atmega1280", "atmega1281", "atmega1284",
16805               "atmega1284p", "atmega1284rfr2", "at90can128", "at90usb1286",
16806               "at90usb1287".
16807
16808           "avr6"
16809               "Enhanced" devices with 3-byte PC, i.e. with more than
16810               128@tie{}KiB of program memory.  mcu@tie{}= "atmega256rfr2",
16811               "atmega2560", "atmega2561", "atmega2564rfr2".
16812
16813           "avrxmega2"
16814               "XMEGA" devices with more than 8@tie{}KiB and up to 64@tie{}KiB
16815               of program memory.  mcu@tie{}= "atxmega8e5", "atxmega16a4",
16816               "atxmega16a4u", "atxmega16c4", "atxmega16d4", "atxmega16e5",
16817               "atxmega32a4", "atxmega32a4u", "atxmega32c3", "atxmega32c4",
16818               "atxmega32d3", "atxmega32d4", "atxmega32e5".
16819
16820           "avrxmega3"
16821               "XMEGA" devices with up to 64@tie{}KiB of combined program
16822               memory and RAM, and with program memory visible in the RAM
16823               address space.  mcu@tie{}= "attiny202", "attiny204",
16824               "attiny212", "attiny214", "attiny402", "attiny404",
16825               "attiny406", "attiny412", "attiny414", "attiny416",
16826               "attiny417", "attiny804", "attiny806", "attiny807",
16827               "attiny814", "attiny816", "attiny817", "attiny1604",
16828               "attiny1606", "attiny1607", "attiny1614", "attiny1616",
16829               "attiny1617", "attiny3214", "attiny3216", "attiny3217",
16830               "atmega808", "atmega809", "atmega1608", "atmega1609",
16831               "atmega3208", "atmega3209", "atmega4808", "atmega4809".
16832
16833           "avrxmega4"
16834               "XMEGA" devices with more than 64@tie{}KiB and up to
16835               128@tie{}KiB of program memory.  mcu@tie{}= "atxmega64a3",
16836               "atxmega64a3u", "atxmega64a4u", "atxmega64b1", "atxmega64b3",
16837               "atxmega64c3", "atxmega64d3", "atxmega64d4".
16838
16839           "avrxmega5"
16840               "XMEGA" devices with more than 64@tie{}KiB and up to
16841               128@tie{}KiB of program memory and more than 64@tie{}KiB of
16842               RAM.  mcu@tie{}= "atxmega64a1", "atxmega64a1u".
16843
16844           "avrxmega6"
16845               "XMEGA" devices with more than 128@tie{}KiB of program memory.
16846               mcu@tie{}= "atxmega128a3", "atxmega128a3u", "atxmega128b1",
16847               "atxmega128b3", "atxmega128c3", "atxmega128d3", "atxmega128d4",
16848               "atxmega192a3", "atxmega192a3u", "atxmega192c3",
16849               "atxmega192d3", "atxmega256a3", "atxmega256a3b",
16850               "atxmega256a3bu", "atxmega256a3u", "atxmega256c3",
16851               "atxmega256d3", "atxmega384c3", "atxmega384d3".
16852
16853           "avrxmega7"
16854               "XMEGA" devices with more than 128@tie{}KiB of program memory
16855               and more than 64@tie{}KiB of RAM.  mcu@tie{}= "atxmega128a1",
16856               "atxmega128a1u", "atxmega128a4u".
16857
16858           "avrtiny"
16859               "TINY" Tiny core devices with 512@tie{}B up to 4@tie{}KiB of
16860               program memory.  mcu@tie{}= "attiny4", "attiny5", "attiny9",
16861               "attiny10", "attiny20", "attiny40".
16862
16863           "avr1"
16864               This ISA is implemented by the minimal AVR core and supported
16865               for assembler only.  mcu@tie{}= "attiny11", "attiny12",
16866               "attiny15", "attiny28", "at90s1200".
16867
16868       -mabsdata
16869           Assume that all data in static storage can be accessed by LDS / STS
16870           instructions.  This option has only an effect on reduced Tiny
16871           devices like ATtiny40.  See also the "absdata" AVR Variable
16872           Attributes,variable attribute.
16873
16874       -maccumulate-args
16875           Accumulate outgoing function arguments and acquire/release the
16876           needed stack space for outgoing function arguments once in function
16877           prologue/epilogue.  Without this option, outgoing arguments are
16878           pushed before calling a function and popped afterwards.
16879
16880           Popping the arguments after the function call can be expensive on
16881           AVR so that accumulating the stack space might lead to smaller
16882           executables because arguments need not be removed from the stack
16883           after such a function call.
16884
16885           This option can lead to reduced code size for functions that
16886           perform several calls to functions that get their arguments on the
16887           stack like calls to printf-like functions.
16888
16889       -mbranch-cost=cost
16890           Set the branch costs for conditional branch instructions to cost.
16891           Reasonable values for cost are small, non-negative integers. The
16892           default branch cost is 0.
16893
16894       -mcall-prologues
16895           Functions prologues/epilogues are expanded as calls to appropriate
16896           subroutines.  Code size is smaller.
16897
16898       -mdouble=bits
16899       -mlong-double=bits
16900           Set the size (in bits) of the "double" or "long double" type,
16901           respectively.  Possible values for bits are 32 and 64.  Whether or
16902           not a specific value for bits is allowed depends on the
16903           "--with-double=" and "--with-long-double=" configure options
16904           ("https://gcc.gnu.org/install/configure.html#avr"), and the same
16905           applies for the default values of the options.
16906
16907       -mgas-isr-prologues
16908           Interrupt service routines (ISRs) may use the "__gcc_isr" pseudo
16909           instruction supported by GNU Binutils.  If this option is on, the
16910           feature can still be disabled for individual ISRs by means of the
16911           AVR Function Attributes,,"no_gccisr" function attribute.  This
16912           feature is activated per default if optimization is on (but not
16913           with -Og, @pxref{Optimize Options}), and if GNU Binutils support
16914           PR21683 ("https://sourceware.org/PR21683").
16915
16916       -mint8
16917           Assume "int" to be 8-bit integer.  This affects the sizes of all
16918           types: a "char" is 1 byte, an "int" is 1 byte, a "long" is 2 bytes,
16919           and "long long" is 4 bytes.  Please note that this option does not
16920           conform to the C standards, but it results in smaller code size.
16921
16922       -mmain-is-OS_task
16923           Do not save registers in "main".  The effect is the same like
16924           attaching attribute AVR Function Attributes,,"OS_task" to "main".
16925           It is activated per default if optimization is on.
16926
16927       -mn-flash=num
16928           Assume that the flash memory has a size of num times 64@tie{}KiB.
16929
16930       -mno-interrupts
16931           Generated code is not compatible with hardware interrupts.  Code
16932           size is smaller.
16933
16934       -mrelax
16935           Try to replace "CALL" resp. "JMP" instruction by the shorter
16936           "RCALL" resp. "RJMP" instruction if applicable.  Setting -mrelax
16937           just adds the --mlink-relax option to the assembler's command line
16938           and the --relax option to the linker's command line.
16939
16940           Jump relaxing is performed by the linker because jump offsets are
16941           not known before code is located. Therefore, the assembler code
16942           generated by the compiler is the same, but the instructions in the
16943           executable may differ from instructions in the assembler code.
16944
16945           Relaxing must be turned on if linker stubs are needed, see the
16946           section on "EIND" and linker stubs below.
16947
16948       -mrmw
16949           Assume that the device supports the Read-Modify-Write instructions
16950           "XCH", "LAC", "LAS" and "LAT".
16951
16952       -mshort-calls
16953           Assume that "RJMP" and "RCALL" can target the whole program memory.
16954
16955           This option is used internally for multilib selection.  It is not
16956           an optimization option, and you don't need to set it by hand.
16957
16958       -msp8
16959           Treat the stack pointer register as an 8-bit register, i.e. assume
16960           the high byte of the stack pointer is zero.  In general, you don't
16961           need to set this option by hand.
16962
16963           This option is used internally by the compiler to select and build
16964           multilibs for architectures "avr2" and "avr25".  These
16965           architectures mix devices with and without "SPH".  For any setting
16966           other than -mmcu=avr2 or -mmcu=avr25 the compiler driver adds or
16967           removes this option from the compiler proper's command line,
16968           because the compiler then knows if the device or architecture has
16969           an 8-bit stack pointer and thus no "SPH" register or not.
16970
16971       -mstrict-X
16972           Use address register "X" in a way proposed by the hardware.  This
16973           means that "X" is only used in indirect, post-increment or pre-
16974           decrement addressing.
16975
16976           Without this option, the "X" register may be used in the same way
16977           as "Y" or "Z" which then is emulated by additional instructions.
16978           For example, loading a value with "X+const" addressing with a small
16979           non-negative "const < 64" to a register Rn is performed as
16980
16981                   adiw r26, const   ; X += const
16982                   ld   <Rn>, X        ; <Rn> = *X
16983                   sbiw r26, const   ; X -= const
16984
16985       -mtiny-stack
16986           Only change the lower 8@tie{}bits of the stack pointer.
16987
16988       -mfract-convert-truncate
16989           Allow to use truncation instead of rounding towards zero for
16990           fractional fixed-point types.
16991
16992       -nodevicelib
16993           Don't link against AVR-LibC's device specific library "lib<mcu>.a".
16994
16995       -nodevicespecs
16996           Don't add -specs=device-specs/specs-mcu to the compiler driver's
16997           command line.  The user takes responsibility for supplying the sub-
16998           processes like compiler proper, assembler and linker with
16999           appropriate command line options.  This means that the user has to
17000           supply her private device specs file by means of -specs=path-to-
17001           specs-file.  There is no more need for option -mmcu=mcu.
17002
17003           This option can also serve as a replacement for the older way of
17004           specifying custom device-specs files that needed -B some-path to
17005           point to a directory which contains a folder named "device-specs"
17006           which contains a specs file named "specs-mcu", where mcu was
17007           specified by -mmcu=mcu.
17008
17009       -Waddr-space-convert
17010           Warn about conversions between address spaces in the case where the
17011           resulting address space is not contained in the incoming address
17012           space.
17013
17014       -Wmisspelled-isr
17015           Warn if the ISR is misspelled, i.e. without __vector prefix.
17016           Enabled by default.
17017
17018       "EIND" and Devices with More Than 128 Ki Bytes of Flash
17019
17020       Pointers in the implementation are 16@tie{}bits wide.  The address of a
17021       function or label is represented as word address so that indirect jumps
17022       and calls can target any code address in the range of 64@tie{}Ki words.
17023
17024       In order to facilitate indirect jump on devices with more than
17025       128@tie{}Ki bytes of program memory space, there is a special function
17026       register called "EIND" that serves as most significant part of the
17027       target address when "EICALL" or "EIJMP" instructions are used.
17028
17029       Indirect jumps and calls on these devices are handled as follows by the
17030       compiler and are subject to some limitations:
17031
17032       *   The compiler never sets "EIND".
17033
17034       *   The compiler uses "EIND" implicitly in "EICALL"/"EIJMP"
17035           instructions or might read "EIND" directly in order to emulate an
17036           indirect call/jump by means of a "RET" instruction.
17037
17038       *   The compiler assumes that "EIND" never changes during the startup
17039           code or during the application. In particular, "EIND" is not
17040           saved/restored in function or interrupt service routine
17041           prologue/epilogue.
17042
17043       *   For indirect calls to functions and computed goto, the linker
17044           generates stubs. Stubs are jump pads sometimes also called
17045           trampolines. Thus, the indirect call/jump jumps to such a stub.
17046           The stub contains a direct jump to the desired address.
17047
17048       *   Linker relaxation must be turned on so that the linker generates
17049           the stubs correctly in all situations. See the compiler option
17050           -mrelax and the linker option --relax.  There are corner cases
17051           where the linker is supposed to generate stubs but aborts without
17052           relaxation and without a helpful error message.
17053
17054       *   The default linker script is arranged for code with "EIND = 0".  If
17055           code is supposed to work for a setup with "EIND != 0", a custom
17056           linker script has to be used in order to place the sections whose
17057           name start with ".trampolines" into the segment where "EIND" points
17058           to.
17059
17060       *   The startup code from libgcc never sets "EIND".  Notice that
17061           startup code is a blend of code from libgcc and AVR-LibC.  For the
17062           impact of AVR-LibC on "EIND", see the AVR-LibC user manual
17063           ("http://nongnu.org/avr-libc/user-manual/").
17064
17065       *   It is legitimate for user-specific startup code to set up "EIND"
17066           early, for example by means of initialization code located in
17067           section ".init3". Such code runs prior to general startup code that
17068           initializes RAM and calls constructors, but after the bit of
17069           startup code from AVR-LibC that sets "EIND" to the segment where
17070           the vector table is located.
17071
17072                   #include <avr/io.h>
17073
17074                   static void
17075                   __attribute__((section(".init3"),naked,used,no_instrument_function))
17076                   init3_set_eind (void)
17077                   {
17078                     __asm volatile ("ldi r24,pm_hh8(__trampolines_start)\n\t"
17079                                     "out %i0,r24" :: "n" (&EIND) : "r24","memory");
17080                   }
17081
17082           The "__trampolines_start" symbol is defined in the linker script.
17083
17084       *   Stubs are generated automatically by the linker if the following
17085           two conditions are met:
17086
17087           -<The address of a label is taken by means of the "gs" modifier>
17088               (short for generate stubs) like so:
17089
17090                       LDI r24, lo8(gs(<func>))
17091                       LDI r25, hi8(gs(<func>))
17092
17093           -<The final location of that label is in a code segment>
17094               outside the segment where the stubs are located.
17095
17096       *   The compiler emits such "gs" modifiers for code labels in the
17097           following situations:
17098
17099           -<Taking address of a function or code label.>
17100           -<Computed goto.>
17101           -<If prologue-save function is used, see -mcall-prologues>
17102               command-line option.
17103
17104           -<Switch/case dispatch tables. If you do not want such dispatch>
17105               tables you can specify the -fno-jump-tables command-line
17106               option.
17107
17108           -<C and C++ constructors/destructors called during
17109           startup/shutdown.>
17110           -<If the tools hit a "gs()" modifier explained above.>
17111       *   Jumping to non-symbolic addresses like so is not supported:
17112
17113                   int main (void)
17114                   {
17115                       /* Call function at word address 0x2 */
17116                       return ((int(*)(void)) 0x2)();
17117                   }
17118
17119           Instead, a stub has to be set up, i.e. the function has to be
17120           called through a symbol ("func_4" in the example):
17121
17122                   int main (void)
17123                   {
17124                       extern int func_4 (void);
17125
17126                       /* Call function at byte address 0x4 */
17127                       return func_4();
17128                   }
17129
17130           and the application be linked with -Wl,--defsym,func_4=0x4.
17131           Alternatively, "func_4" can be defined in the linker script.
17132
17133       Handling of the "RAMPD", "RAMPX", "RAMPY" and "RAMPZ" Special Function
17134       Registers
17135
17136       Some AVR devices support memories larger than the 64@tie{}KiB range
17137       that can be accessed with 16-bit pointers.  To access memory locations
17138       outside this 64@tie{}KiB range, the content of a "RAMP" register is
17139       used as high part of the address: The "X", "Y", "Z" address register is
17140       concatenated with the "RAMPX", "RAMPY", "RAMPZ" special function
17141       register, respectively, to get a wide address. Similarly, "RAMPD" is
17142       used together with direct addressing.
17143
17144       *   The startup code initializes the "RAMP" special function registers
17145           with zero.
17146
17147       *   If a AVR Named Address Spaces,named address space other than
17148           generic or "__flash" is used, then "RAMPZ" is set as needed before
17149           the operation.
17150
17151       *   If the device supports RAM larger than 64@tie{}KiB and the compiler
17152           needs to change "RAMPZ" to accomplish an operation, "RAMPZ" is
17153           reset to zero after the operation.
17154
17155       *   If the device comes with a specific "RAMP" register, the ISR
17156           prologue/epilogue saves/restores that SFR and initializes it with
17157           zero in case the ISR code might (implicitly) use it.
17158
17159       *   RAM larger than 64@tie{}KiB is not supported by GCC for AVR
17160           targets.  If you use inline assembler to read from locations
17161           outside the 16-bit address range and change one of the "RAMP"
17162           registers, you must reset it to zero after the access.
17163
17164       AVR Built-in Macros
17165
17166       GCC defines several built-in macros so that the user code can test for
17167       the presence or absence of features.  Almost any of the following
17168       built-in macros are deduced from device capabilities and thus triggered
17169       by the -mmcu= command-line option.
17170
17171       For even more AVR-specific built-in macros see AVR Named Address Spaces
17172       and AVR Built-in Functions.
17173
17174       "__AVR_ARCH__"
17175           Build-in macro that resolves to a decimal number that identifies
17176           the architecture and depends on the -mmcu=mcu option.  Possible
17177           values are:
17178
17179           2, 25, 3, 31, 35, 4, 5, 51, 6
17180
17181           for mcu="avr2", "avr25", "avr3", "avr31", "avr35", "avr4", "avr5",
17182           "avr51", "avr6",
17183
17184           respectively and
17185
17186           100, 102, 103, 104, 105, 106, 107
17187
17188           for mcu="avrtiny", "avrxmega2", "avrxmega3", "avrxmega4",
17189           "avrxmega5", "avrxmega6", "avrxmega7", respectively.  If mcu
17190           specifies a device, this built-in macro is set accordingly. For
17191           example, with -mmcu=atmega8 the macro is defined to 4.
17192
17193       "__AVR_Device__"
17194           Setting -mmcu=device defines this built-in macro which reflects the
17195           device's name. For example, -mmcu=atmega8 defines the built-in
17196           macro "__AVR_ATmega8__", -mmcu=attiny261a defines
17197           "__AVR_ATtiny261A__", etc.
17198
17199           The built-in macros' names follow the scheme "__AVR_Device__" where
17200           Device is the device name as from the AVR user manual. The
17201           difference between Device in the built-in macro and device in
17202           -mmcu=device is that the latter is always lowercase.
17203
17204           If device is not a device but only a core architecture like avr51,
17205           this macro is not defined.
17206
17207       "__AVR_DEVICE_NAME__"
17208           Setting -mmcu=device defines this built-in macro to the device's
17209           name. For example, with -mmcu=atmega8 the macro is defined to
17210           "atmega8".
17211
17212           If device is not a device but only a core architecture like avr51,
17213           this macro is not defined.
17214
17215       "__AVR_XMEGA__"
17216           The device / architecture belongs to the XMEGA family of devices.
17217
17218       "__AVR_HAVE_ELPM__"
17219           The device has the "ELPM" instruction.
17220
17221       "__AVR_HAVE_ELPMX__"
17222           The device has the "ELPM Rn,Z" and "ELPM Rn,Z+" instructions.
17223
17224       "__AVR_HAVE_MOVW__"
17225           The device has the "MOVW" instruction to perform 16-bit register-
17226           register moves.
17227
17228       "__AVR_HAVE_LPMX__"
17229           The device has the "LPM Rn,Z" and "LPM Rn,Z+" instructions.
17230
17231       "__AVR_HAVE_MUL__"
17232           The device has a hardware multiplier.
17233
17234       "__AVR_HAVE_JMP_CALL__"
17235           The device has the "JMP" and "CALL" instructions.  This is the case
17236           for devices with more than 8@tie{}KiB of program memory.
17237
17238       "__AVR_HAVE_EIJMP_EICALL__"
17239       "__AVR_3_BYTE_PC__"
17240           The device has the "EIJMP" and "EICALL" instructions.  This is the
17241           case for devices with more than 128@tie{}KiB of program memory.
17242           This also means that the program counter (PC) is 3@tie{}bytes wide.
17243
17244       "__AVR_2_BYTE_PC__"
17245           The program counter (PC) is 2@tie{}bytes wide. This is the case for
17246           devices with up to 128@tie{}KiB of program memory.
17247
17248       "__AVR_HAVE_8BIT_SP__"
17249       "__AVR_HAVE_16BIT_SP__"
17250           The stack pointer (SP) register is treated as 8-bit respectively
17251           16-bit register by the compiler.  The definition of these macros is
17252           affected by -mtiny-stack.
17253
17254       "__AVR_HAVE_SPH__"
17255       "__AVR_SP8__"
17256           The device has the SPH (high part of stack pointer) special
17257           function register or has an 8-bit stack pointer, respectively.  The
17258           definition of these macros is affected by -mmcu= and in the cases
17259           of -mmcu=avr2 and -mmcu=avr25 also by -msp8.
17260
17261       "__AVR_HAVE_RAMPD__"
17262       "__AVR_HAVE_RAMPX__"
17263       "__AVR_HAVE_RAMPY__"
17264       "__AVR_HAVE_RAMPZ__"
17265           The device has the "RAMPD", "RAMPX", "RAMPY", "RAMPZ" special
17266           function register, respectively.
17267
17268       "__NO_INTERRUPTS__"
17269           This macro reflects the -mno-interrupts command-line option.
17270
17271       "__AVR_ERRATA_SKIP__"
17272       "__AVR_ERRATA_SKIP_JMP_CALL__"
17273           Some AVR devices (AT90S8515, ATmega103) must not skip 32-bit
17274           instructions because of a hardware erratum.  Skip instructions are
17275           "SBRS", "SBRC", "SBIS", "SBIC" and "CPSE".  The second macro is
17276           only defined if "__AVR_HAVE_JMP_CALL__" is also set.
17277
17278       "__AVR_ISA_RMW__"
17279           The device has Read-Modify-Write instructions (XCH, LAC, LAS and
17280           LAT).
17281
17282       "__AVR_SFR_OFFSET__=offset"
17283           Instructions that can address I/O special function registers
17284           directly like "IN", "OUT", "SBI", etc. may use a different address
17285           as if addressed by an instruction to access RAM like "LD" or "STS".
17286           This offset depends on the device architecture and has to be
17287           subtracted from the RAM address in order to get the respective
17288           I/O@tie{}address.
17289
17290       "__AVR_SHORT_CALLS__"
17291           The -mshort-calls command line option is set.
17292
17293       "__AVR_PM_BASE_ADDRESS__=addr"
17294           Some devices support reading from flash memory by means of "LD*"
17295           instructions.  The flash memory is seen in the data address space
17296           at an offset of "__AVR_PM_BASE_ADDRESS__".  If this macro is not
17297           defined, this feature is not available.  If defined, the address
17298           space is linear and there is no need to put ".rodata" into RAM.
17299           This is handled by the default linker description file, and is
17300           currently available for "avrtiny" and "avrxmega3".  Even more
17301           convenient, there is no need to use address spaces like "__flash"
17302           or features like attribute "progmem" and "pgm_read_*".
17303
17304       "__WITH_AVRLIBC__"
17305           The compiler is configured to be used together with AVR-Libc.  See
17306           the --with-avrlibc configure option.
17307
17308       "__HAVE_DOUBLE_MULTILIB__"
17309           Defined if -mdouble= acts as a multilib option.
17310
17311       "__HAVE_DOUBLE32__"
17312       "__HAVE_DOUBLE64__"
17313           Defined if the compiler supports 32-bit double resp. 64-bit double.
17314           The actual layout is specified by option -mdouble=.
17315
17316       "__DEFAULT_DOUBLE__"
17317           The size in bits of "double" if -mdouble= is not set.  To test the
17318           layout of "double" in a program, use the built-in macro
17319           "__SIZEOF_DOUBLE__".
17320
17321       "__HAVE_LONG_DOUBLE32__"
17322       "__HAVE_LONG_DOUBLE64__"
17323       "__HAVE_LONG_DOUBLE_MULTILIB__"
17324       "__DEFAULT_LONG_DOUBLE__"
17325           Same as above, but for "long double" instead of "double".
17326
17327       "__WITH_DOUBLE_COMPARISON__"
17328           Reflects the "--with-double-comparison={tristate|bool|libf7}"
17329           configure option ("https://gcc.gnu.org/install/configure.html#avr")
17330           and is defined to 2 or 3.
17331
17332       "__WITH_LIBF7_LIBGCC__"
17333       "__WITH_LIBF7_MATH__"
17334       "__WITH_LIBF7_MATH_SYMBOLS__"
17335           Reflects the "--with-libf7={libgcc|math|math-symbols}"
17336           configure option
17337           ("https://gcc.gnu.org/install/configure.html#avr").
17338
17339   Blackfin Options
17340       -mcpu=cpu[-sirevision]
17341           Specifies the name of the target Blackfin processor.  Currently,
17342           cpu can be one of bf512, bf514, bf516, bf518, bf522, bf523, bf524,
17343           bf525, bf526, bf527, bf531, bf532, bf533, bf534, bf536, bf537,
17344           bf538, bf539, bf542, bf544, bf547, bf548, bf549, bf542m, bf544m,
17345           bf547m, bf548m, bf549m, bf561, bf592.
17346
17347           The optional sirevision specifies the silicon revision of the
17348           target Blackfin processor.  Any workarounds available for the
17349           targeted silicon revision are enabled.  If sirevision is none, no
17350           workarounds are enabled.  If sirevision is any, all workarounds for
17351           the targeted processor are enabled.  The "__SILICON_REVISION__"
17352           macro is defined to two hexadecimal digits representing the major
17353           and minor numbers in the silicon revision.  If sirevision is none,
17354           the "__SILICON_REVISION__" is not defined.  If sirevision is any,
17355           the "__SILICON_REVISION__" is defined to be 0xffff.  If this
17356           optional sirevision is not used, GCC assumes the latest known
17357           silicon revision of the targeted Blackfin processor.
17358
17359           GCC defines a preprocessor macro for the specified cpu.  For the
17360           bfin-elf toolchain, this option causes the hardware BSP provided by
17361           libgloss to be linked in if -msim is not given.
17362
17363           Without this option, bf532 is used as the processor by default.
17364
17365           Note that support for bf561 is incomplete.  For bf561, only the
17366           preprocessor macro is defined.
17367
17368       -msim
17369           Specifies that the program will be run on the simulator.  This
17370           causes the simulator BSP provided by libgloss to be linked in.
17371           This option has effect only for bfin-elf toolchain.  Certain other
17372           options, such as -mid-shared-library and -mfdpic, imply -msim.
17373
17374       -momit-leaf-frame-pointer
17375           Don't keep the frame pointer in a register for leaf functions.
17376           This avoids the instructions to save, set up and restore frame
17377           pointers and makes an extra register available in leaf functions.
17378
17379       -mspecld-anomaly
17380           When enabled, the compiler ensures that the generated code does not
17381           contain speculative loads after jump instructions. If this option
17382           is used, "__WORKAROUND_SPECULATIVE_LOADS" is defined.
17383
17384       -mno-specld-anomaly
17385           Don't generate extra code to prevent speculative loads from
17386           occurring.
17387
17388       -mcsync-anomaly
17389           When enabled, the compiler ensures that the generated code does not
17390           contain CSYNC or SSYNC instructions too soon after conditional
17391           branches.  If this option is used, "__WORKAROUND_SPECULATIVE_SYNCS"
17392           is defined.
17393
17394       -mno-csync-anomaly
17395           Don't generate extra code to prevent CSYNC or SSYNC instructions
17396           from occurring too soon after a conditional branch.
17397
17398       -mlow64k
17399           When enabled, the compiler is free to take advantage of the
17400           knowledge that the entire program fits into the low 64k of memory.
17401
17402       -mno-low64k
17403           Assume that the program is arbitrarily large.  This is the default.
17404
17405       -mstack-check-l1
17406           Do stack checking using information placed into L1 scratchpad
17407           memory by the uClinux kernel.
17408
17409       -mid-shared-library
17410           Generate code that supports shared libraries via the library ID
17411           method.  This allows for execute in place and shared libraries in
17412           an environment without virtual memory management.  This option
17413           implies -fPIC.  With a bfin-elf target, this option implies -msim.
17414
17415       -mno-id-shared-library
17416           Generate code that doesn't assume ID-based shared libraries are
17417           being used.  This is the default.
17418
17419       -mleaf-id-shared-library
17420           Generate code that supports shared libraries via the library ID
17421           method, but assumes that this library or executable won't link
17422           against any other ID shared libraries.  That allows the compiler to
17423           use faster code for jumps and calls.
17424
17425       -mno-leaf-id-shared-library
17426           Do not assume that the code being compiled won't link against any
17427           ID shared libraries.  Slower code is generated for jump and call
17428           insns.
17429
17430       -mshared-library-id=n
17431           Specifies the identification number of the ID-based shared library
17432           being compiled.  Specifying a value of 0 generates more compact
17433           code; specifying other values forces the allocation of that number
17434           to the current library but is no more space- or time-efficient than
17435           omitting this option.
17436
17437       -msep-data
17438           Generate code that allows the data segment to be located in a
17439           different area of memory from the text segment.  This allows for
17440           execute in place in an environment without virtual memory
17441           management by eliminating relocations against the text section.
17442
17443       -mno-sep-data
17444           Generate code that assumes that the data segment follows the text
17445           segment.  This is the default.
17446
17447       -mlong-calls
17448       -mno-long-calls
17449           Tells the compiler to perform function calls by first loading the
17450           address of the function into a register and then performing a
17451           subroutine call on this register.  This switch is needed if the
17452           target function lies outside of the 24-bit addressing range of the
17453           offset-based version of subroutine call instruction.
17454
17455           This feature is not enabled by default.  Specifying -mno-long-calls
17456           restores the default behavior.  Note these switches have no effect
17457           on how the compiler generates code to handle function calls via
17458           function pointers.
17459
17460       -mfast-fp
17461           Link with the fast floating-point library. This library relaxes
17462           some of the IEEE floating-point standard's rules for checking
17463           inputs against Not-a-Number (NAN), in the interest of performance.
17464
17465       -minline-plt
17466           Enable inlining of PLT entries in function calls to functions that
17467           are not known to bind locally.  It has no effect without -mfdpic.
17468
17469       -mmulticore
17470           Build a standalone application for multicore Blackfin processors.
17471           This option causes proper start files and link scripts supporting
17472           multicore to be used, and defines the macro "__BFIN_MULTICORE".  It
17473           can only be used with -mcpu=bf561[-sirevision].
17474
17475           This option can be used with -mcorea or -mcoreb, which selects the
17476           one-application-per-core programming model.  Without -mcorea or
17477           -mcoreb, the single-application/dual-core programming model is
17478           used. In this model, the main function of Core B should be named as
17479           "coreb_main".
17480
17481           If this option is not used, the single-core application programming
17482           model is used.
17483
17484       -mcorea
17485           Build a standalone application for Core A of BF561 when using the
17486           one-application-per-core programming model. Proper start files and
17487           link scripts are used to support Core A, and the macro
17488           "__BFIN_COREA" is defined.  This option can only be used in
17489           conjunction with -mmulticore.
17490
17491       -mcoreb
17492           Build a standalone application for Core B of BF561 when using the
17493           one-application-per-core programming model. Proper start files and
17494           link scripts are used to support Core B, and the macro
17495           "__BFIN_COREB" is defined. When this option is used, "coreb_main"
17496           should be used instead of "main".  This option can only be used in
17497           conjunction with -mmulticore.
17498
17499       -msdram
17500           Build a standalone application for SDRAM. Proper start files and
17501           link scripts are used to put the application into SDRAM, and the
17502           macro "__BFIN_SDRAM" is defined.  The loader should initialize
17503           SDRAM before loading the application.
17504
17505       -micplb
17506           Assume that ICPLBs are enabled at run time.  This has an effect on
17507           certain anomaly workarounds.  For Linux targets, the default is to
17508           assume ICPLBs are enabled; for standalone applications the default
17509           is off.
17510
17511   C6X Options
17512       -march=name
17513           This specifies the name of the target architecture.  GCC uses this
17514           name to determine what kind of instructions it can emit when
17515           generating assembly code.  Permissible names are: c62x, c64x,
17516           c64x+, c67x, c67x+, c674x.
17517
17518       -mbig-endian
17519           Generate code for a big-endian target.
17520
17521       -mlittle-endian
17522           Generate code for a little-endian target.  This is the default.
17523
17524       -msim
17525           Choose startup files and linker script suitable for the simulator.
17526
17527       -msdata=default
17528           Put small global and static data in the ".neardata" section, which
17529           is pointed to by register "B14".  Put small uninitialized global
17530           and static data in the ".bss" section, which is adjacent to the
17531           ".neardata" section.  Put small read-only data into the ".rodata"
17532           section.  The corresponding sections used for large pieces of data
17533           are ".fardata", ".far" and ".const".
17534
17535       -msdata=all
17536           Put all data, not just small objects, into the sections reserved
17537           for small data, and use addressing relative to the "B14" register
17538           to access them.
17539
17540       -msdata=none
17541           Make no use of the sections reserved for small data, and use
17542           absolute addresses to access all data.  Put all initialized global
17543           and static data in the ".fardata" section, and all uninitialized
17544           data in the ".far" section.  Put all constant data into the
17545           ".const" section.
17546
17547   CRIS Options
17548       These options are defined specifically for the CRIS ports.
17549
17550       -march=architecture-type
17551       -mcpu=architecture-type
17552           Generate code for the specified architecture.  The choices for
17553           architecture-type are v3, v8 and v10 for respectively ETRAX 4,
17554           ETRAX 100, and ETRAX 100 LX.  Default is v0 except for cris-axis-
17555           linux-gnu, where the default is v10.
17556
17557       -mtune=architecture-type
17558           Tune to architecture-type everything applicable about the generated
17559           code, except for the ABI and the set of available instructions.
17560           The choices for architecture-type are the same as for
17561           -march=architecture-type.
17562
17563       -mmax-stack-frame=n
17564           Warn when the stack frame of a function exceeds n bytes.
17565
17566       -metrax4
17567       -metrax100
17568           The options -metrax4 and -metrax100 are synonyms for -march=v3 and
17569           -march=v8 respectively.
17570
17571       -mmul-bug-workaround
17572       -mno-mul-bug-workaround
17573           Work around a bug in the "muls" and "mulu" instructions for CPU
17574           models where it applies.  This option is active by default.
17575
17576       -mpdebug
17577           Enable CRIS-specific verbose debug-related information in the
17578           assembly code.  This option also has the effect of turning off the
17579           #NO_APP formatted-code indicator to the assembler at the beginning
17580           of the assembly file.
17581
17582       -mcc-init
17583           Do not use condition-code results from previous instruction; always
17584           emit compare and test instructions before use of condition codes.
17585
17586       -mno-side-effects
17587           Do not emit instructions with side effects in addressing modes
17588           other than post-increment.
17589
17590       -mstack-align
17591       -mno-stack-align
17592       -mdata-align
17593       -mno-data-align
17594       -mconst-align
17595       -mno-const-align
17596           These options (no- options) arrange (eliminate arrangements) for
17597           the stack frame, individual data and constants to be aligned for
17598           the maximum single data access size for the chosen CPU model.  The
17599           default is to arrange for 32-bit alignment.  ABI details such as
17600           structure layout are not affected by these options.
17601
17602       -m32-bit
17603       -m16-bit
17604       -m8-bit
17605           Similar to the stack- data- and const-align options above, these
17606           options arrange for stack frame, writable data and constants to all
17607           be 32-bit, 16-bit or 8-bit aligned.  The default is 32-bit
17608           alignment.
17609
17610       -mno-prologue-epilogue
17611       -mprologue-epilogue
17612           With -mno-prologue-epilogue, the normal function prologue and
17613           epilogue which set up the stack frame are omitted and no return
17614           instructions or return sequences are generated in the code.  Use
17615           this option only together with visual inspection of the compiled
17616           code: no warnings or errors are generated when call-saved registers
17617           must be saved, or storage for local variables needs to be
17618           allocated.
17619
17620       -mno-gotplt
17621       -mgotplt
17622           With -fpic and -fPIC, don't generate (do generate) instruction
17623           sequences that load addresses for functions from the PLT part of
17624           the GOT rather than (traditional on other architectures) calls to
17625           the PLT.  The default is -mgotplt.
17626
17627       -melf
17628           Legacy no-op option only recognized with the cris-axis-elf and
17629           cris-axis-linux-gnu targets.
17630
17631       -mlinux
17632           Legacy no-op option only recognized with the cris-axis-linux-gnu
17633           target.
17634
17635       -sim
17636           This option, recognized for the cris-axis-elf, arranges to link
17637           with input-output functions from a simulator library.  Code,
17638           initialized data and zero-initialized data are allocated
17639           consecutively.
17640
17641       -sim2
17642           Like -sim, but pass linker options to locate initialized data at
17643           0x40000000 and zero-initialized data at 0x80000000.
17644
17645   CR16 Options
17646       These options are defined specifically for the CR16 ports.
17647
17648       -mmac
17649           Enable the use of multiply-accumulate instructions. Disabled by
17650           default.
17651
17652       -mcr16cplus
17653       -mcr16c
17654           Generate code for CR16C or CR16C+ architecture. CR16C+ architecture
17655           is default.
17656
17657       -msim
17658           Links the library libsim.a which is in compatible with simulator.
17659           Applicable to ELF compiler only.
17660
17661       -mint32
17662           Choose integer type as 32-bit wide.
17663
17664       -mbit-ops
17665           Generates "sbit"/"cbit" instructions for bit manipulations.
17666
17667       -mdata-model=model
17668           Choose a data model. The choices for model are near, far or medium.
17669           medium is default.  However, far is not valid with -mcr16c, as the
17670           CR16C architecture does not support the far data model.
17671
17672   C-SKY Options
17673       GCC supports these options when compiling for C-SKY V2 processors.
17674
17675       -march=arch
17676           Specify the C-SKY target architecture.  Valid values for arch are:
17677           ck801, ck802, ck803, ck807, and ck810.  The default is ck810.
17678
17679       -mcpu=cpu
17680           Specify the C-SKY target processor.  Valid values for cpu are:
17681           ck801, ck801t, ck802, ck802t, ck802j, ck803, ck803h, ck803t,
17682           ck803ht, ck803f, ck803fh, ck803e, ck803eh, ck803et, ck803eht,
17683           ck803ef, ck803efh, ck803ft, ck803eft, ck803efht, ck803r1, ck803hr1,
17684           ck803tr1, ck803htr1, ck803fr1, ck803fhr1, ck803er1, ck803ehr1,
17685           ck803etr1, ck803ehtr1, ck803efr1, ck803efhr1, ck803ftr1,
17686           ck803eftr1, ck803efhtr1, ck803s, ck803st, ck803se, ck803sf,
17687           ck803sef, ck803seft, ck807e, ck807ef, ck807, ck807f, ck810e,
17688           ck810et, ck810ef, ck810eft, ck810, ck810v, ck810f, ck810t, ck810fv,
17689           ck810tv, ck810ft, and ck810ftv.
17690
17691       -mbig-endian
17692       -EB
17693       -mlittle-endian
17694       -EL Select big- or little-endian code.  The default is little-endian.
17695
17696       -mhard-float
17697       -msoft-float
17698           Select hardware or software floating-point implementations.  The
17699           default is soft float.
17700
17701       -mdouble-float
17702       -mno-double-float
17703           When -mhard-float is in effect, enable generation of double-
17704           precision float instructions.  This is the default except when
17705           compiling for CK803.
17706
17707       -mfdivdu
17708       -mno-fdivdu
17709           When -mhard-float is in effect, enable generation of "frecipd",
17710           "fsqrtd", and "fdivd" instructions.  This is the default except
17711           when compiling for CK803.
17712
17713       -mfpu=fpu
17714           Select the floating-point processor.  This option can only be used
17715           with -mhard-float.  Values for fpu are fpv2_sf (equivalent to
17716           -mno-double-float -mno-fdivdu), fpv2 (-mdouble-float -mno-divdu),
17717           and fpv2_divd (-mdouble-float -mdivdu).
17718
17719       -melrw
17720       -mno-elrw
17721           Enable the extended "lrw" instruction.  This option defaults to on
17722           for CK801 and off otherwise.
17723
17724       -mistack
17725       -mno-istack
17726           Enable interrupt stack instructions; the default is off.
17727
17728           The -mistack option is required to handle the "interrupt" and "isr"
17729           function attributes.
17730
17731       -mmp
17732           Enable multiprocessor instructions; the default is off.
17733
17734       -mcp
17735           Enable coprocessor instructions; the default is off.
17736
17737       -mcache
17738           Enable coprocessor instructions; the default is off.
17739
17740       -msecurity
17741           Enable C-SKY security instructions; the default is off.
17742
17743       -mtrust
17744           Enable C-SKY trust instructions; the default is off.
17745
17746       -mdsp
17747       -medsp
17748       -mvdsp
17749           Enable C-SKY DSP, Enhanced DSP, or Vector DSP instructions,
17750           respectively.  All of these options default to off.
17751
17752       -mdiv
17753       -mno-div
17754           Generate divide instructions.  Default is off.
17755
17756       -msmart
17757       -mno-smart
17758           Generate code for Smart Mode, using only registers numbered 0-7 to
17759           allow use of 16-bit instructions.  This option is ignored for CK801
17760           where this is the required behavior, and it defaults to on for
17761           CK802.  For other targets, the default is off.
17762
17763       -mhigh-registers
17764       -mno-high-registers
17765           Generate code using the high registers numbered 16-31.  This option
17766           is not supported on CK801, CK802, or CK803, and is enabled by
17767           default for other processors.
17768
17769       -manchor
17770       -mno-anchor
17771           Generate code using global anchor symbol addresses.
17772
17773       -mpushpop
17774       -mno-pushpop
17775           Generate code using "push" and "pop" instructions.  This option
17776           defaults to on.
17777
17778       -mmultiple-stld
17779       -mstm
17780       -mno-multiple-stld
17781       -mno-stm
17782           Generate code using "stm" and "ldm" instructions.  This option
17783           isn't supported on CK801 but is enabled by default on other
17784           processors.
17785
17786       -mconstpool
17787       -mno-constpool
17788           Create constant pools in the compiler instead of deferring it to
17789           the assembler.  This option is the default and required for correct
17790           code generation on CK801 and CK802, and is optional on other
17791           processors.
17792
17793       -mstack-size
17794       -mno-stack-size
17795           Emit ".stack_size" directives for each function in the assembly
17796           output.  This option defaults to off.
17797
17798       -mccrt
17799       -mno-ccrt
17800           Generate code for the C-SKY compiler runtime instead of libgcc.
17801           This option defaults to off.
17802
17803       -mbranch-cost=n
17804           Set the branch costs to roughly "n" instructions.  The default is
17805           1.
17806
17807       -msched-prolog
17808       -mno-sched-prolog
17809           Permit scheduling of function prologue and epilogue sequences.
17810           Using this option can result in code that is not compliant with the
17811           C-SKY V2 ABI prologue requirements and that cannot be debugged or
17812           backtraced.  It is disabled by default.
17813
17814   Darwin Options
17815       These options are defined for all architectures running the Darwin
17816       operating system.
17817
17818       FSF GCC on Darwin does not create "fat" object files; it creates an
17819       object file for the single architecture that GCC was built to target.
17820       Apple's GCC on Darwin does create "fat" files if multiple -arch options
17821       are used; it does so by running the compiler or linker multiple times
17822       and joining the results together with lipo.
17823
17824       The subtype of the file created (like ppc7400 or ppc970 or i686) is
17825       determined by the flags that specify the ISA that GCC is targeting,
17826       like -mcpu or -march.  The -force_cpusubtype_ALL option can be used to
17827       override this.
17828
17829       The Darwin tools vary in their behavior when presented with an ISA
17830       mismatch.  The assembler, as, only permits instructions to be used that
17831       are valid for the subtype of the file it is generating, so you cannot
17832       put 64-bit instructions in a ppc750 object file.  The linker for shared
17833       libraries, /usr/bin/libtool, fails and prints an error if asked to
17834       create a shared library with a less restrictive subtype than its input
17835       files (for instance, trying to put a ppc970 object file in a ppc7400
17836       library).  The linker for executables, ld, quietly gives the executable
17837       the most restrictive subtype of any of its input files.
17838
17839       -Fdir
17840           Add the framework directory dir to the head of the list of
17841           directories to be searched for header files.  These directories are
17842           interleaved with those specified by -I options and are scanned in a
17843           left-to-right order.
17844
17845           A framework directory is a directory with frameworks in it.  A
17846           framework is a directory with a Headers and/or PrivateHeaders
17847           directory contained directly in it that ends in .framework.  The
17848           name of a framework is the name of this directory excluding the
17849           .framework.  Headers associated with the framework are found in one
17850           of those two directories, with Headers being searched first.  A
17851           subframework is a framework directory that is in a framework's
17852           Frameworks directory.  Includes of subframework headers can only
17853           appear in a header of a framework that contains the subframework,
17854           or in a sibling subframework header.  Two subframeworks are
17855           siblings if they occur in the same framework.  A subframework
17856           should not have the same name as a framework; a warning is issued
17857           if this is violated.  Currently a subframework cannot have
17858           subframeworks; in the future, the mechanism may be extended to
17859           support this.  The standard frameworks can be found in
17860           /System/Library/Frameworks and /Library/Frameworks.  An example
17861           include looks like "#include <Framework/header.h>", where Framework
17862           denotes the name of the framework and header.h is found in the
17863           PrivateHeaders or Headers directory.
17864
17865       -iframeworkdir
17866           Like -F except the directory is a treated as a system directory.
17867           The main difference between this -iframework and -F is that with
17868           -iframework the compiler does not warn about constructs contained
17869           within header files found via dir.  This option is valid only for
17870           the C family of languages.
17871
17872       -gused
17873           Emit debugging information for symbols that are used.  For stabs
17874           debugging format, this enables -feliminate-unused-debug-symbols.
17875           This is by default ON.
17876
17877       -gfull
17878           Emit debugging information for all symbols and types.
17879
17880       -mmacosx-version-min=version
17881           The earliest version of MacOS X that this executable will run on is
17882           version.  Typical values of version include 10.1, 10.2, and 10.3.9.
17883
17884           If the compiler was built to use the system's headers by default,
17885           then the default for this option is the system version on which the
17886           compiler is running, otherwise the default is to make choices that
17887           are compatible with as many systems and code bases as possible.
17888
17889       -mkernel
17890           Enable kernel development mode.  The -mkernel option sets -static,
17891           -fno-common, -fno-use-cxa-atexit, -fno-exceptions,
17892           -fno-non-call-exceptions, -fapple-kext, -fno-weak and -fno-rtti
17893           where applicable.  This mode also sets -mno-altivec, -msoft-float,
17894           -fno-builtin and -mlong-branch for PowerPC targets.
17895
17896       -mone-byte-bool
17897           Override the defaults for "bool" so that "sizeof(bool)==1".  By
17898           default "sizeof(bool)" is 4 when compiling for Darwin/PowerPC and 1
17899           when compiling for Darwin/x86, so this option has no effect on x86.
17900
17901           Warning: The -mone-byte-bool switch causes GCC to generate code
17902           that is not binary compatible with code generated without that
17903           switch.  Using this switch may require recompiling all other
17904           modules in a program, including system libraries.  Use this switch
17905           to conform to a non-default data model.
17906
17907       -mfix-and-continue
17908       -ffix-and-continue
17909       -findirect-data
17910           Generate code suitable for fast turnaround development, such as to
17911           allow GDB to dynamically load .o files into already-running
17912           programs.  -findirect-data and -ffix-and-continue are provided for
17913           backwards compatibility.
17914
17915       -all_load
17916           Loads all members of static archive libraries.  See man ld(1) for
17917           more information.
17918
17919       -arch_errors_fatal
17920           Cause the errors having to do with files that have the wrong
17921           architecture to be fatal.
17922
17923       -bind_at_load
17924           Causes the output file to be marked such that the dynamic linker
17925           will bind all undefined references when the file is loaded or
17926           launched.
17927
17928       -bundle
17929           Produce a Mach-o bundle format file.  See man ld(1) for more
17930           information.
17931
17932       -bundle_loader executable
17933           This option specifies the executable that will load the build
17934           output file being linked.  See man ld(1) for more information.
17935
17936       -dynamiclib
17937           When passed this option, GCC produces a dynamic library instead of
17938           an executable when linking, using the Darwin libtool command.
17939
17940       -force_cpusubtype_ALL
17941           This causes GCC's output file to have the ALL subtype, instead of
17942           one controlled by the -mcpu or -march option.
17943
17944       -allowable_client  client_name
17945       -client_name
17946       -compatibility_version
17947       -current_version
17948       -dead_strip
17949       -dependency-file
17950       -dylib_file
17951       -dylinker_install_name
17952       -dynamic
17953       -exported_symbols_list
17954       -filelist
17955       -flat_namespace
17956       -force_flat_namespace
17957       -headerpad_max_install_names
17958       -image_base
17959       -init
17960       -install_name
17961       -keep_private_externs
17962       -multi_module
17963       -multiply_defined
17964       -multiply_defined_unused
17965       -noall_load
17966       -no_dead_strip_inits_and_terms
17967       -nofixprebinding
17968       -nomultidefs
17969       -noprebind
17970       -noseglinkedit
17971       -pagezero_size
17972       -prebind
17973       -prebind_all_twolevel_modules
17974       -private_bundle
17975       -read_only_relocs
17976       -sectalign
17977       -sectobjectsymbols
17978       -whyload
17979       -seg1addr
17980       -sectcreate
17981       -sectobjectsymbols
17982       -sectorder
17983       -segaddr
17984       -segs_read_only_addr
17985       -segs_read_write_addr
17986       -seg_addr_table
17987       -seg_addr_table_filename
17988       -seglinkedit
17989       -segprot
17990       -segs_read_only_addr
17991       -segs_read_write_addr
17992       -single_module
17993       -static
17994       -sub_library
17995       -sub_umbrella
17996       -twolevel_namespace
17997       -umbrella
17998       -undefined
17999       -unexported_symbols_list
18000       -weak_reference_mismatches
18001       -whatsloaded
18002           These options are passed to the Darwin linker.  The Darwin linker
18003           man page describes them in detail.
18004
18005   DEC Alpha Options
18006       These -m options are defined for the DEC Alpha implementations:
18007
18008       -mno-soft-float
18009       -msoft-float
18010           Use (do not use) the hardware floating-point instructions for
18011           floating-point operations.  When -msoft-float is specified,
18012           functions in libgcc.a are used to perform floating-point
18013           operations.  Unless they are replaced by routines that emulate the
18014           floating-point operations, or compiled in such a way as to call
18015           such emulations routines, these routines issue floating-point
18016           operations.   If you are compiling for an Alpha without floating-
18017           point operations, you must ensure that the library is built so as
18018           not to call them.
18019
18020           Note that Alpha implementations without floating-point operations
18021           are required to have floating-point registers.
18022
18023       -mfp-reg
18024       -mno-fp-regs
18025           Generate code that uses (does not use) the floating-point register
18026           set.  -mno-fp-regs implies -msoft-float.  If the floating-point
18027           register set is not used, floating-point operands are passed in
18028           integer registers as if they were integers and floating-point
18029           results are passed in $0 instead of $f0.  This is a non-standard
18030           calling sequence, so any function with a floating-point argument or
18031           return value called by code compiled with -mno-fp-regs must also be
18032           compiled with that option.
18033
18034           A typical use of this option is building a kernel that does not
18035           use, and hence need not save and restore, any floating-point
18036           registers.
18037
18038       -mieee
18039           The Alpha architecture implements floating-point hardware optimized
18040           for maximum performance.  It is mostly compliant with the IEEE
18041           floating-point standard.  However, for full compliance, software
18042           assistance is required.  This option generates code fully IEEE-
18043           compliant code except that the inexact-flag is not maintained (see
18044           below).  If this option is turned on, the preprocessor macro
18045           "_IEEE_FP" is defined during compilation.  The resulting code is
18046           less efficient but is able to correctly support denormalized
18047           numbers and exceptional IEEE values such as not-a-number and
18048           plus/minus infinity.  Other Alpha compilers call this option
18049           -ieee_with_no_inexact.
18050
18051       -mieee-with-inexact
18052           This is like -mieee except the generated code also maintains the
18053           IEEE inexact-flag.  Turning on this option causes the generated
18054           code to implement fully-compliant IEEE math.  In addition to
18055           "_IEEE_FP", "_IEEE_FP_EXACT" is defined as a preprocessor macro.
18056           On some Alpha implementations the resulting code may execute
18057           significantly slower than the code generated by default.  Since
18058           there is very little code that depends on the inexact-flag, you
18059           should normally not specify this option.  Other Alpha compilers
18060           call this option -ieee_with_inexact.
18061
18062       -mfp-trap-mode=trap-mode
18063           This option controls what floating-point related traps are enabled.
18064           Other Alpha compilers call this option -fptm trap-mode.  The trap
18065           mode can be set to one of four values:
18066
18067           n   This is the default (normal) setting.  The only traps that are
18068               enabled are the ones that cannot be disabled in software (e.g.,
18069               division by zero trap).
18070
18071           u   In addition to the traps enabled by n, underflow traps are
18072               enabled as well.
18073
18074           su  Like u, but the instructions are marked to be safe for software
18075               completion (see Alpha architecture manual for details).
18076
18077           sui Like su, but inexact traps are enabled as well.
18078
18079       -mfp-rounding-mode=rounding-mode
18080           Selects the IEEE rounding mode.  Other Alpha compilers call this
18081           option -fprm rounding-mode.  The rounding-mode can be one of:
18082
18083           n   Normal IEEE rounding mode.  Floating-point numbers are rounded
18084               towards the nearest machine number or towards the even machine
18085               number in case of a tie.
18086
18087           m   Round towards minus infinity.
18088
18089           c   Chopped rounding mode.  Floating-point numbers are rounded
18090               towards zero.
18091
18092           d   Dynamic rounding mode.  A field in the floating-point control
18093               register (fpcr, see Alpha architecture reference manual)
18094               controls the rounding mode in effect.  The C library
18095               initializes this register for rounding towards plus infinity.
18096               Thus, unless your program modifies the fpcr, d corresponds to
18097               round towards plus infinity.
18098
18099       -mtrap-precision=trap-precision
18100           In the Alpha architecture, floating-point traps are imprecise.
18101           This means without software assistance it is impossible to recover
18102           from a floating trap and program execution normally needs to be
18103           terminated.  GCC can generate code that can assist operating system
18104           trap handlers in determining the exact location that caused a
18105           floating-point trap.  Depending on the requirements of an
18106           application, different levels of precisions can be selected:
18107
18108           p   Program precision.  This option is the default and means a trap
18109               handler can only identify which program caused a floating-point
18110               exception.
18111
18112           f   Function precision.  The trap handler can determine the
18113               function that caused a floating-point exception.
18114
18115           i   Instruction precision.  The trap handler can determine the
18116               exact instruction that caused a floating-point exception.
18117
18118           Other Alpha compilers provide the equivalent options called
18119           -scope_safe and -resumption_safe.
18120
18121       -mieee-conformant
18122           This option marks the generated code as IEEE conformant.  You must
18123           not use this option unless you also specify -mtrap-precision=i and
18124           either -mfp-trap-mode=su or -mfp-trap-mode=sui.  Its only effect is
18125           to emit the line .eflag 48 in the function prologue of the
18126           generated assembly file.
18127
18128       -mbuild-constants
18129           Normally GCC examines a 32- or 64-bit integer constant to see if it
18130           can construct it from smaller constants in two or three
18131           instructions.  If it cannot, it outputs the constant as a literal
18132           and generates code to load it from the data segment at run time.
18133
18134           Use this option to require GCC to construct all integer constants
18135           using code, even if it takes more instructions (the maximum is
18136           six).
18137
18138           You typically use this option to build a shared library dynamic
18139           loader.  Itself a shared library, it must relocate itself in memory
18140           before it can find the variables and constants in its own data
18141           segment.
18142
18143       -mbwx
18144       -mno-bwx
18145       -mcix
18146       -mno-cix
18147       -mfix
18148       -mno-fix
18149       -mmax
18150       -mno-max
18151           Indicate whether GCC should generate code to use the optional BWX,
18152           CIX, FIX and MAX instruction sets.  The default is to use the
18153           instruction sets supported by the CPU type specified via -mcpu=
18154           option or that of the CPU on which GCC was built if none is
18155           specified.
18156
18157       -mfloat-vax
18158       -mfloat-ieee
18159           Generate code that uses (does not use) VAX F and G floating-point
18160           arithmetic instead of IEEE single and double precision.
18161
18162       -mexplicit-relocs
18163       -mno-explicit-relocs
18164           Older Alpha assemblers provided no way to generate symbol
18165           relocations except via assembler macros.  Use of these macros does
18166           not allow optimal instruction scheduling.  GNU binutils as of
18167           version 2.12 supports a new syntax that allows the compiler to
18168           explicitly mark which relocations should apply to which
18169           instructions.  This option is mostly useful for debugging, as GCC
18170           detects the capabilities of the assembler when it is built and sets
18171           the default accordingly.
18172
18173       -msmall-data
18174       -mlarge-data
18175           When -mexplicit-relocs is in effect, static data is accessed via
18176           gp-relative relocations.  When -msmall-data is used, objects 8
18177           bytes long or smaller are placed in a small data area (the ".sdata"
18178           and ".sbss" sections) and are accessed via 16-bit relocations off
18179           of the $gp register.  This limits the size of the small data area
18180           to 64KB, but allows the variables to be directly accessed via a
18181           single instruction.
18182
18183           The default is -mlarge-data.  With this option the data area is
18184           limited to just below 2GB.  Programs that require more than 2GB of
18185           data must use "malloc" or "mmap" to allocate the data in the heap
18186           instead of in the program's data segment.
18187
18188           When generating code for shared libraries, -fpic implies
18189           -msmall-data and -fPIC implies -mlarge-data.
18190
18191       -msmall-text
18192       -mlarge-text
18193           When -msmall-text is used, the compiler assumes that the code of
18194           the entire program (or shared library) fits in 4MB, and is thus
18195           reachable with a branch instruction.  When -msmall-data is used,
18196           the compiler can assume that all local symbols share the same $gp
18197           value, and thus reduce the number of instructions required for a
18198           function call from 4 to 1.
18199
18200           The default is -mlarge-text.
18201
18202       -mcpu=cpu_type
18203           Set the instruction set and instruction scheduling parameters for
18204           machine type cpu_type.  You can specify either the EV style name or
18205           the corresponding chip number.  GCC supports scheduling parameters
18206           for the EV4, EV5 and EV6 family of processors and chooses the
18207           default values for the instruction set from the processor you
18208           specify.  If you do not specify a processor type, GCC defaults to
18209           the processor on which the compiler was built.
18210
18211           Supported values for cpu_type are
18212
18213           ev4
18214           ev45
18215           21064
18216               Schedules as an EV4 and has no instruction set extensions.
18217
18218           ev5
18219           21164
18220               Schedules as an EV5 and has no instruction set extensions.
18221
18222           ev56
18223           21164a
18224               Schedules as an EV5 and supports the BWX extension.
18225
18226           pca56
18227           21164pc
18228           21164PC
18229               Schedules as an EV5 and supports the BWX and MAX extensions.
18230
18231           ev6
18232           21264
18233               Schedules as an EV6 and supports the BWX, FIX, and MAX
18234               extensions.
18235
18236           ev67
18237           21264a
18238               Schedules as an EV6 and supports the BWX, CIX, FIX, and MAX
18239               extensions.
18240
18241           Native toolchains also support the value native, which selects the
18242           best architecture option for the host processor.  -mcpu=native has
18243           no effect if GCC does not recognize the processor.
18244
18245       -mtune=cpu_type
18246           Set only the instruction scheduling parameters for machine type
18247           cpu_type.  The instruction set is not changed.
18248
18249           Native toolchains also support the value native, which selects the
18250           best architecture option for the host processor.  -mtune=native has
18251           no effect if GCC does not recognize the processor.
18252
18253       -mmemory-latency=time
18254           Sets the latency the scheduler should assume for typical memory
18255           references as seen by the application.  This number is highly
18256           dependent on the memory access patterns used by the application and
18257           the size of the external cache on the machine.
18258
18259           Valid options for time are
18260
18261           number
18262               A decimal number representing clock cycles.
18263
18264           L1
18265           L2
18266           L3
18267           main
18268               The compiler contains estimates of the number of clock cycles
18269               for "typical" EV4 & EV5 hardware for the Level 1, 2 & 3 caches
18270               (also called Dcache, Scache, and Bcache), as well as to main
18271               memory.  Note that L3 is only valid for EV5.
18272
18273   eBPF Options
18274       -mframe-limit=bytes
18275           This specifies the hard limit for frame sizes, in bytes.
18276           Currently, the value that can be specified should be less than or
18277           equal to 32767.  Defaults to whatever limit is imposed by the
18278           version of the Linux kernel targeted.
18279
18280       -mkernel=version
18281           This specifies the minimum version of the kernel that will run the
18282           compiled program.  GCC uses this version to determine which
18283           instructions to use, what kernel helpers to allow, etc.  Currently,
18284           version can be one of 4.0, 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7, 4.8,
18285           4.9, 4.10, 4.11, 4.12, 4.13, 4.14, 4.15, 4.16, 4.17, 4.18, 4.19,
18286           4.20, 5.0, 5.1, 5.2, latest and native.
18287
18288       -mbig-endian
18289           Generate code for a big-endian target.
18290
18291       -mlittle-endian
18292           Generate code for a little-endian target.  This is the default.
18293
18294       -mxbpf
18295           Generate code for an expanded version of BPF, which relaxes some of
18296           the restrictions imposed by the BPF architecture:
18297
18298           -<Save and restore callee-saved registers at function entry and>
18299               exit, respectively.
18300
18301   FR30 Options
18302       These options are defined specifically for the FR30 port.
18303
18304       -msmall-model
18305           Use the small address space model.  This can produce smaller code,
18306           but it does assume that all symbolic values and addresses fit into
18307           a 20-bit range.
18308
18309       -mno-lsim
18310           Assume that runtime support has been provided and so there is no
18311           need to include the simulator library (libsim.a) on the linker
18312           command line.
18313
18314   FT32 Options
18315       These options are defined specifically for the FT32 port.
18316
18317       -msim
18318           Specifies that the program will be run on the simulator.  This
18319           causes an alternate runtime startup and library to be linked.  You
18320           must not use this option when generating programs that will run on
18321           real hardware; you must provide your own runtime library for
18322           whatever I/O functions are needed.
18323
18324       -mlra
18325           Enable Local Register Allocation.  This is still experimental for
18326           FT32, so by default the compiler uses standard reload.
18327
18328       -mnodiv
18329           Do not use div and mod instructions.
18330
18331       -mft32b
18332           Enable use of the extended instructions of the FT32B processor.
18333
18334       -mcompress
18335           Compress all code using the Ft32B code compression scheme.
18336
18337       -mnopm
18338           Do not generate code that reads program memory.
18339
18340   FRV Options
18341       -mgpr-32
18342           Only use the first 32 general-purpose registers.
18343
18344       -mgpr-64
18345           Use all 64 general-purpose registers.
18346
18347       -mfpr-32
18348           Use only the first 32 floating-point registers.
18349
18350       -mfpr-64
18351           Use all 64 floating-point registers.
18352
18353       -mhard-float
18354           Use hardware instructions for floating-point operations.
18355
18356       -msoft-float
18357           Use library routines for floating-point operations.
18358
18359       -malloc-cc
18360           Dynamically allocate condition code registers.
18361
18362       -mfixed-cc
18363           Do not try to dynamically allocate condition code registers, only
18364           use "icc0" and "fcc0".
18365
18366       -mdword
18367           Change ABI to use double word insns.
18368
18369       -mno-dword
18370           Do not use double word instructions.
18371
18372       -mdouble
18373           Use floating-point double instructions.
18374
18375       -mno-double
18376           Do not use floating-point double instructions.
18377
18378       -mmedia
18379           Use media instructions.
18380
18381       -mno-media
18382           Do not use media instructions.
18383
18384       -mmuladd
18385           Use multiply and add/subtract instructions.
18386
18387       -mno-muladd
18388           Do not use multiply and add/subtract instructions.
18389
18390       -mfdpic
18391           Select the FDPIC ABI, which uses function descriptors to represent
18392           pointers to functions.  Without any PIC/PIE-related options, it
18393           implies -fPIE.  With -fpic or -fpie, it assumes GOT entries and
18394           small data are within a 12-bit range from the GOT base address;
18395           with -fPIC or -fPIE, GOT offsets are computed with 32 bits.  With a
18396           bfin-elf target, this option implies -msim.
18397
18398       -minline-plt
18399           Enable inlining of PLT entries in function calls to functions that
18400           are not known to bind locally.  It has no effect without -mfdpic.
18401           It's enabled by default if optimizing for speed and compiling for
18402           shared libraries (i.e., -fPIC or -fpic), or when an optimization
18403           option such as -O3 or above is present in the command line.
18404
18405       -mTLS
18406           Assume a large TLS segment when generating thread-local code.
18407
18408       -mtls
18409           Do not assume a large TLS segment when generating thread-local
18410           code.
18411
18412       -mgprel-ro
18413           Enable the use of "GPREL" relocations in the FDPIC ABI for data
18414           that is known to be in read-only sections.  It's enabled by
18415           default, except for -fpic or -fpie: even though it may help make
18416           the global offset table smaller, it trades 1 instruction for 4.
18417           With -fPIC or -fPIE, it trades 3 instructions for 4, one of which
18418           may be shared by multiple symbols, and it avoids the need for a GOT
18419           entry for the referenced symbol, so it's more likely to be a win.
18420           If it is not, -mno-gprel-ro can be used to disable it.
18421
18422       -multilib-library-pic
18423           Link with the (library, not FD) pic libraries.  It's implied by
18424           -mlibrary-pic, as well as by -fPIC and -fpic without -mfdpic.  You
18425           should never have to use it explicitly.
18426
18427       -mlinked-fp
18428           Follow the EABI requirement of always creating a frame pointer
18429           whenever a stack frame is allocated.  This option is enabled by
18430           default and can be disabled with -mno-linked-fp.
18431
18432       -mlong-calls
18433           Use indirect addressing to call functions outside the current
18434           compilation unit.  This allows the functions to be placed anywhere
18435           within the 32-bit address space.
18436
18437       -malign-labels
18438           Try to align labels to an 8-byte boundary by inserting NOPs into
18439           the previous packet.  This option only has an effect when VLIW
18440           packing is enabled.  It doesn't create new packets; it merely adds
18441           NOPs to existing ones.
18442
18443       -mlibrary-pic
18444           Generate position-independent EABI code.
18445
18446       -macc-4
18447           Use only the first four media accumulator registers.
18448
18449       -macc-8
18450           Use all eight media accumulator registers.
18451
18452       -mpack
18453           Pack VLIW instructions.
18454
18455       -mno-pack
18456           Do not pack VLIW instructions.
18457
18458       -mno-eflags
18459           Do not mark ABI switches in e_flags.
18460
18461       -mcond-move
18462           Enable the use of conditional-move instructions (default).
18463
18464           This switch is mainly for debugging the compiler and will likely be
18465           removed in a future version.
18466
18467       -mno-cond-move
18468           Disable the use of conditional-move instructions.
18469
18470           This switch is mainly for debugging the compiler and will likely be
18471           removed in a future version.
18472
18473       -mscc
18474           Enable the use of conditional set instructions (default).
18475
18476           This switch is mainly for debugging the compiler and will likely be
18477           removed in a future version.
18478
18479       -mno-scc
18480           Disable the use of conditional set instructions.
18481
18482           This switch is mainly for debugging the compiler and will likely be
18483           removed in a future version.
18484
18485       -mcond-exec
18486           Enable the use of conditional execution (default).
18487
18488           This switch is mainly for debugging the compiler and will likely be
18489           removed in a future version.
18490
18491       -mno-cond-exec
18492           Disable the use of conditional execution.
18493
18494           This switch is mainly for debugging the compiler and will likely be
18495           removed in a future version.
18496
18497       -mvliw-branch
18498           Run a pass to pack branches into VLIW instructions (default).
18499
18500           This switch is mainly for debugging the compiler and will likely be
18501           removed in a future version.
18502
18503       -mno-vliw-branch
18504           Do not run a pass to pack branches into VLIW instructions.
18505
18506           This switch is mainly for debugging the compiler and will likely be
18507           removed in a future version.
18508
18509       -mmulti-cond-exec
18510           Enable optimization of "&&" and "||" in conditional execution
18511           (default).
18512
18513           This switch is mainly for debugging the compiler and will likely be
18514           removed in a future version.
18515
18516       -mno-multi-cond-exec
18517           Disable optimization of "&&" and "||" in conditional execution.
18518
18519           This switch is mainly for debugging the compiler and will likely be
18520           removed in a future version.
18521
18522       -mnested-cond-exec
18523           Enable nested conditional execution optimizations (default).
18524
18525           This switch is mainly for debugging the compiler and will likely be
18526           removed in a future version.
18527
18528       -mno-nested-cond-exec
18529           Disable nested conditional execution optimizations.
18530
18531           This switch is mainly for debugging the compiler and will likely be
18532           removed in a future version.
18533
18534       -moptimize-membar
18535           This switch removes redundant "membar" instructions from the
18536           compiler-generated code.  It is enabled by default.
18537
18538       -mno-optimize-membar
18539           This switch disables the automatic removal of redundant "membar"
18540           instructions from the generated code.
18541
18542       -mtomcat-stats
18543           Cause gas to print out tomcat statistics.
18544
18545       -mcpu=cpu
18546           Select the processor type for which to generate code.  Possible
18547           values are frv, fr550, tomcat, fr500, fr450, fr405, fr400, fr300
18548           and simple.
18549
18550   GNU/Linux Options
18551       These -m options are defined for GNU/Linux targets:
18552
18553       -mglibc
18554           Use the GNU C library.  This is the default except on
18555           *-*-linux-*uclibc*, *-*-linux-*musl* and *-*-linux-*android*
18556           targets.
18557
18558       -muclibc
18559           Use uClibc C library.  This is the default on *-*-linux-*uclibc*
18560           targets.
18561
18562       -mmusl
18563           Use the musl C library.  This is the default on *-*-linux-*musl*
18564           targets.
18565
18566       -mbionic
18567           Use Bionic C library.  This is the default on *-*-linux-*android*
18568           targets.
18569
18570       -mandroid
18571           Compile code compatible with Android platform.  This is the default
18572           on *-*-linux-*android* targets.
18573
18574           When compiling, this option enables -mbionic, -fPIC,
18575           -fno-exceptions and -fno-rtti by default.  When linking, this
18576           option makes the GCC driver pass Android-specific options to the
18577           linker.  Finally, this option causes the preprocessor macro
18578           "__ANDROID__" to be defined.
18579
18580       -tno-android-cc
18581           Disable compilation effects of -mandroid, i.e., do not enable
18582           -mbionic, -fPIC, -fno-exceptions and -fno-rtti by default.
18583
18584       -tno-android-ld
18585           Disable linking effects of -mandroid, i.e., pass standard Linux
18586           linking options to the linker.
18587
18588   H8/300 Options
18589       These -m options are defined for the H8/300 implementations:
18590
18591       -mrelax
18592           Shorten some address references at link time, when possible; uses
18593           the linker option -relax.
18594
18595       -mh Generate code for the H8/300H.
18596
18597       -ms Generate code for the H8S.
18598
18599       -mn Generate code for the H8S and H8/300H in the normal mode.  This
18600           switch must be used either with -mh or -ms.
18601
18602       -ms2600
18603           Generate code for the H8S/2600.  This switch must be used with -ms.
18604
18605       -mexr
18606           Extended registers are stored on stack before execution of function
18607           with monitor attribute. Default option is -mexr.  This option is
18608           valid only for H8S targets.
18609
18610       -mno-exr
18611           Extended registers are not stored on stack before execution of
18612           function with monitor attribute. Default option is -mno-exr.  This
18613           option is valid only for H8S targets.
18614
18615       -mint32
18616           Make "int" data 32 bits by default.
18617
18618       -malign-300
18619           On the H8/300H and H8S, use the same alignment rules as for the
18620           H8/300.  The default for the H8/300H and H8S is to align longs and
18621           floats on 4-byte boundaries.  -malign-300 causes them to be aligned
18622           on 2-byte boundaries.  This option has no effect on the H8/300.
18623
18624   HPPA Options
18625       These -m options are defined for the HPPA family of computers:
18626
18627       -march=architecture-type
18628           Generate code for the specified architecture.  The choices for
18629           architecture-type are 1.0 for PA 1.0, 1.1 for PA 1.1, and 2.0 for
18630           PA 2.0 processors.  Refer to /usr/lib/sched.models on an HP-UX
18631           system to determine the proper architecture option for your
18632           machine.  Code compiled for lower numbered architectures runs on
18633           higher numbered architectures, but not the other way around.
18634
18635       -mpa-risc-1-0
18636       -mpa-risc-1-1
18637       -mpa-risc-2-0
18638           Synonyms for -march=1.0, -march=1.1, and -march=2.0 respectively.
18639
18640       -mcaller-copies
18641           The caller copies function arguments passed by hidden reference.
18642           This option should be used with care as it is not compatible with
18643           the default 32-bit runtime.  However, only aggregates larger than
18644           eight bytes are passed by hidden reference and the option provides
18645           better compatibility with OpenMP.
18646
18647       -mjump-in-delay
18648           This option is ignored and provided for compatibility purposes
18649           only.
18650
18651       -mdisable-fpregs
18652           Prevent floating-point registers from being used in any manner.
18653           This is necessary for compiling kernels that perform lazy context
18654           switching of floating-point registers.  If you use this option and
18655           attempt to perform floating-point operations, the compiler aborts.
18656
18657       -mdisable-indexing
18658           Prevent the compiler from using indexing address modes.  This
18659           avoids some rather obscure problems when compiling MIG generated
18660           code under MACH.
18661
18662       -mno-space-regs
18663           Generate code that assumes the target has no space registers.  This
18664           allows GCC to generate faster indirect calls and use unscaled index
18665           address modes.
18666
18667           Such code is suitable for level 0 PA systems and kernels.
18668
18669       -mfast-indirect-calls
18670           Generate code that assumes calls never cross space boundaries.
18671           This allows GCC to emit code that performs faster indirect calls.
18672
18673           This option does not work in the presence of shared libraries or
18674           nested functions.
18675
18676       -mfixed-range=register-range
18677           Generate code treating the given register range as fixed registers.
18678           A fixed register is one that the register allocator cannot use.
18679           This is useful when compiling kernel code.  A register range is
18680           specified as two registers separated by a dash.  Multiple register
18681           ranges can be specified separated by a comma.
18682
18683       -mlong-load-store
18684           Generate 3-instruction load and store sequences as sometimes
18685           required by the HP-UX 10 linker.  This is equivalent to the +k
18686           option to the HP compilers.
18687
18688       -mportable-runtime
18689           Use the portable calling conventions proposed by HP for ELF
18690           systems.
18691
18692       -mgas
18693           Enable the use of assembler directives only GAS understands.
18694
18695       -mschedule=cpu-type
18696           Schedule code according to the constraints for the machine type
18697           cpu-type.  The choices for cpu-type are 700 7100, 7100LC, 7200,
18698           7300 and 8000.  Refer to /usr/lib/sched.models on an HP-UX system
18699           to determine the proper scheduling option for your machine.  The
18700           default scheduling is 8000.
18701
18702       -mlinker-opt
18703           Enable the optimization pass in the HP-UX linker.  Note this makes
18704           symbolic debugging impossible.  It also triggers a bug in the HP-UX
18705           8 and HP-UX 9 linkers in which they give bogus error messages when
18706           linking some programs.
18707
18708       -msoft-float
18709           Generate output containing library calls for floating point.
18710           Warning: the requisite libraries are not available for all HPPA
18711           targets.  Normally the facilities of the machine's usual C compiler
18712           are used, but this cannot be done directly in cross-compilation.
18713           You must make your own arrangements to provide suitable library
18714           functions for cross-compilation.
18715
18716           -msoft-float changes the calling convention in the output file;
18717           therefore, it is only useful if you compile all of a program with
18718           this option.  In particular, you need to compile libgcc.a, the
18719           library that comes with GCC, with -msoft-float in order for this to
18720           work.
18721
18722       -msio
18723           Generate the predefine, "_SIO", for server IO.  The default is
18724           -mwsio.  This generates the predefines, "__hp9000s700",
18725           "__hp9000s700__" and "_WSIO", for workstation IO.  These options
18726           are available under HP-UX and HI-UX.
18727
18728       -mgnu-ld
18729           Use options specific to GNU ld.  This passes -shared to ld when
18730           building a shared library.  It is the default when GCC is
18731           configured, explicitly or implicitly, with the GNU linker.  This
18732           option does not affect which ld is called; it only changes what
18733           parameters are passed to that ld.  The ld that is called is
18734           determined by the --with-ld configure option, GCC's program search
18735           path, and finally by the user's PATH.  The linker used by GCC can
18736           be printed using which `gcc -print-prog-name=ld`.  This option is
18737           only available on the 64-bit HP-UX GCC, i.e. configured with
18738           hppa*64*-*-hpux*.
18739
18740       -mhp-ld
18741           Use options specific to HP ld.  This passes -b to ld when building
18742           a shared library and passes +Accept TypeMismatch to ld on all
18743           links.  It is the default when GCC is configured, explicitly or
18744           implicitly, with the HP linker.  This option does not affect which
18745           ld is called; it only changes what parameters are passed to that
18746           ld.  The ld that is called is determined by the --with-ld configure
18747           option, GCC's program search path, and finally by the user's PATH.
18748           The linker used by GCC can be printed using which `gcc
18749           -print-prog-name=ld`.  This option is only available on the 64-bit
18750           HP-UX GCC, i.e. configured with hppa*64*-*-hpux*.
18751
18752       -mlong-calls
18753           Generate code that uses long call sequences.  This ensures that a
18754           call is always able to reach linker generated stubs.  The default
18755           is to generate long calls only when the distance from the call site
18756           to the beginning of the function or translation unit, as the case
18757           may be, exceeds a predefined limit set by the branch type being
18758           used.  The limits for normal calls are 7,600,000 and 240,000 bytes,
18759           respectively for the PA 2.0 and PA 1.X architectures.  Sibcalls are
18760           always limited at 240,000 bytes.
18761
18762           Distances are measured from the beginning of functions when using
18763           the -ffunction-sections option, or when using the -mgas and
18764           -mno-portable-runtime options together under HP-UX with the SOM
18765           linker.
18766
18767           It is normally not desirable to use this option as it degrades
18768           performance.  However, it may be useful in large applications,
18769           particularly when partial linking is used to build the application.
18770
18771           The types of long calls used depends on the capabilities of the
18772           assembler and linker, and the type of code being generated.  The
18773           impact on systems that support long absolute calls, and long pic
18774           symbol-difference or pc-relative calls should be relatively small.
18775           However, an indirect call is used on 32-bit ELF systems in pic code
18776           and it is quite long.
18777
18778       -munix=unix-std
18779           Generate compiler predefines and select a startfile for the
18780           specified UNIX standard.  The choices for unix-std are 93, 95 and
18781           98.  93 is supported on all HP-UX versions.  95 is available on HP-
18782           UX 10.10 and later.  98 is available on HP-UX 11.11 and later.  The
18783           default values are 93 for HP-UX 10.00, 95 for HP-UX 10.10 though to
18784           11.00, and 98 for HP-UX 11.11 and later.
18785
18786           -munix=93 provides the same predefines as GCC 3.3 and 3.4.
18787           -munix=95 provides additional predefines for "XOPEN_UNIX" and
18788           "_XOPEN_SOURCE_EXTENDED", and the startfile unix95.o.  -munix=98
18789           provides additional predefines for "_XOPEN_UNIX",
18790           "_XOPEN_SOURCE_EXTENDED", "_INCLUDE__STDC_A1_SOURCE" and
18791           "_INCLUDE_XOPEN_SOURCE_500", and the startfile unix98.o.
18792
18793           It is important to note that this option changes the interfaces for
18794           various library routines.  It also affects the operational behavior
18795           of the C library.  Thus, extreme care is needed in using this
18796           option.
18797
18798           Library code that is intended to operate with more than one UNIX
18799           standard must test, set and restore the variable
18800           "__xpg4_extended_mask" as appropriate.  Most GNU software doesn't
18801           provide this capability.
18802
18803       -nolibdld
18804           Suppress the generation of link options to search libdld.sl when
18805           the -static option is specified on HP-UX 10 and later.
18806
18807       -static
18808           The HP-UX implementation of setlocale in libc has a dependency on
18809           libdld.sl.  There isn't an archive version of libdld.sl.  Thus,
18810           when the -static option is specified, special link options are
18811           needed to resolve this dependency.
18812
18813           On HP-UX 10 and later, the GCC driver adds the necessary options to
18814           link with libdld.sl when the -static option is specified.  This
18815           causes the resulting binary to be dynamic.  On the 64-bit port, the
18816           linkers generate dynamic binaries by default in any case.  The
18817           -nolibdld option can be used to prevent the GCC driver from adding
18818           these link options.
18819
18820       -threads
18821           Add support for multithreading with the dce thread library under
18822           HP-UX.  This option sets flags for both the preprocessor and
18823           linker.
18824
18825   IA-64 Options
18826       These are the -m options defined for the Intel IA-64 architecture.
18827
18828       -mbig-endian
18829           Generate code for a big-endian target.  This is the default for HP-
18830           UX.
18831
18832       -mlittle-endian
18833           Generate code for a little-endian target.  This is the default for
18834           AIX5 and GNU/Linux.
18835
18836       -mgnu-as
18837       -mno-gnu-as
18838           Generate (or don't) code for the GNU assembler.  This is the
18839           default.
18840
18841       -mgnu-ld
18842       -mno-gnu-ld
18843           Generate (or don't) code for the GNU linker.  This is the default.
18844
18845       -mno-pic
18846           Generate code that does not use a global pointer register.  The
18847           result is not position independent code, and violates the IA-64
18848           ABI.
18849
18850       -mvolatile-asm-stop
18851       -mno-volatile-asm-stop
18852           Generate (or don't) a stop bit immediately before and after
18853           volatile asm statements.
18854
18855       -mregister-names
18856       -mno-register-names
18857           Generate (or don't) in, loc, and out register names for the stacked
18858           registers.  This may make assembler output more readable.
18859
18860       -mno-sdata
18861       -msdata
18862           Disable (or enable) optimizations that use the small data section.
18863           This may be useful for working around optimizer bugs.
18864
18865       -mconstant-gp
18866           Generate code that uses a single constant global pointer value.
18867           This is useful when compiling kernel code.
18868
18869       -mauto-pic
18870           Generate code that is self-relocatable.  This implies
18871           -mconstant-gp.  This is useful when compiling firmware code.
18872
18873       -minline-float-divide-min-latency
18874           Generate code for inline divides of floating-point values using the
18875           minimum latency algorithm.
18876
18877       -minline-float-divide-max-throughput
18878           Generate code for inline divides of floating-point values using the
18879           maximum throughput algorithm.
18880
18881       -mno-inline-float-divide
18882           Do not generate inline code for divides of floating-point values.
18883
18884       -minline-int-divide-min-latency
18885           Generate code for inline divides of integer values using the
18886           minimum latency algorithm.
18887
18888       -minline-int-divide-max-throughput
18889           Generate code for inline divides of integer values using the
18890           maximum throughput algorithm.
18891
18892       -mno-inline-int-divide
18893           Do not generate inline code for divides of integer values.
18894
18895       -minline-sqrt-min-latency
18896           Generate code for inline square roots using the minimum latency
18897           algorithm.
18898
18899       -minline-sqrt-max-throughput
18900           Generate code for inline square roots using the maximum throughput
18901           algorithm.
18902
18903       -mno-inline-sqrt
18904           Do not generate inline code for "sqrt".
18905
18906       -mfused-madd
18907       -mno-fused-madd
18908           Do (don't) generate code that uses the fused multiply/add or
18909           multiply/subtract instructions.  The default is to use these
18910           instructions.
18911
18912       -mno-dwarf2-asm
18913       -mdwarf2-asm
18914           Don't (or do) generate assembler code for the DWARF line number
18915           debugging info.  This may be useful when not using the GNU
18916           assembler.
18917
18918       -mearly-stop-bits
18919       -mno-early-stop-bits
18920           Allow stop bits to be placed earlier than immediately preceding the
18921           instruction that triggered the stop bit.  This can improve
18922           instruction scheduling, but does not always do so.
18923
18924       -mfixed-range=register-range
18925           Generate code treating the given register range as fixed registers.
18926           A fixed register is one that the register allocator cannot use.
18927           This is useful when compiling kernel code.  A register range is
18928           specified as two registers separated by a dash.  Multiple register
18929           ranges can be specified separated by a comma.
18930
18931       -mtls-size=tls-size
18932           Specify bit size of immediate TLS offsets.  Valid values are 14,
18933           22, and 64.
18934
18935       -mtune=cpu-type
18936           Tune the instruction scheduling for a particular CPU, Valid values
18937           are itanium, itanium1, merced, itanium2, and mckinley.
18938
18939       -milp32
18940       -mlp64
18941           Generate code for a 32-bit or 64-bit environment.  The 32-bit
18942           environment sets int, long and pointer to 32 bits.  The 64-bit
18943           environment sets int to 32 bits and long and pointer to 64 bits.
18944           These are HP-UX specific flags.
18945
18946       -mno-sched-br-data-spec
18947       -msched-br-data-spec
18948           (Dis/En)able data speculative scheduling before reload.  This
18949           results in generation of "ld.a" instructions and the corresponding
18950           check instructions ("ld.c" / "chk.a").  The default setting is
18951           disabled.
18952
18953       -msched-ar-data-spec
18954       -mno-sched-ar-data-spec
18955           (En/Dis)able data speculative scheduling after reload.  This
18956           results in generation of "ld.a" instructions and the corresponding
18957           check instructions ("ld.c" / "chk.a").  The default setting is
18958           enabled.
18959
18960       -mno-sched-control-spec
18961       -msched-control-spec
18962           (Dis/En)able control speculative scheduling.  This feature is
18963           available only during region scheduling (i.e. before reload).  This
18964           results in generation of the "ld.s" instructions and the
18965           corresponding check instructions "chk.s".  The default setting is
18966           disabled.
18967
18968       -msched-br-in-data-spec
18969       -mno-sched-br-in-data-spec
18970           (En/Dis)able speculative scheduling of the instructions that are
18971           dependent on the data speculative loads before reload.  This is
18972           effective only with -msched-br-data-spec enabled.  The default
18973           setting is enabled.
18974
18975       -msched-ar-in-data-spec
18976       -mno-sched-ar-in-data-spec
18977           (En/Dis)able speculative scheduling of the instructions that are
18978           dependent on the data speculative loads after reload.  This is
18979           effective only with -msched-ar-data-spec enabled.  The default
18980           setting is enabled.
18981
18982       -msched-in-control-spec
18983       -mno-sched-in-control-spec
18984           (En/Dis)able speculative scheduling of the instructions that are
18985           dependent on the control speculative loads.  This is effective only
18986           with -msched-control-spec enabled.  The default setting is enabled.
18987
18988       -mno-sched-prefer-non-data-spec-insns
18989       -msched-prefer-non-data-spec-insns
18990           If enabled, data-speculative instructions are chosen for schedule
18991           only if there are no other choices at the moment.  This makes the
18992           use of the data speculation much more conservative.  The default
18993           setting is disabled.
18994
18995       -mno-sched-prefer-non-control-spec-insns
18996       -msched-prefer-non-control-spec-insns
18997           If enabled, control-speculative instructions are chosen for
18998           schedule only if there are no other choices at the moment.  This
18999           makes the use of the control speculation much more conservative.
19000           The default setting is disabled.
19001
19002       -mno-sched-count-spec-in-critical-path
19003       -msched-count-spec-in-critical-path
19004           If enabled, speculative dependencies are considered during
19005           computation of the instructions priorities.  This makes the use of
19006           the speculation a bit more conservative.  The default setting is
19007           disabled.
19008
19009       -msched-spec-ldc
19010           Use a simple data speculation check.  This option is on by default.
19011
19012       -msched-control-spec-ldc
19013           Use a simple check for control speculation.  This option is on by
19014           default.
19015
19016       -msched-stop-bits-after-every-cycle
19017           Place a stop bit after every cycle when scheduling.  This option is
19018           on by default.
19019
19020       -msched-fp-mem-deps-zero-cost
19021           Assume that floating-point stores and loads are not likely to cause
19022           a conflict when placed into the same instruction group.  This
19023           option is disabled by default.
19024
19025       -msel-sched-dont-check-control-spec
19026           Generate checks for control speculation in selective scheduling.
19027           This flag is disabled by default.
19028
19029       -msched-max-memory-insns=max-insns
19030           Limit on the number of memory insns per instruction group, giving
19031           lower priority to subsequent memory insns attempting to schedule in
19032           the same instruction group. Frequently useful to prevent cache bank
19033           conflicts.  The default value is 1.
19034
19035       -msched-max-memory-insns-hard-limit
19036           Makes the limit specified by msched-max-memory-insns a hard limit,
19037           disallowing more than that number in an instruction group.
19038           Otherwise, the limit is "soft", meaning that non-memory operations
19039           are preferred when the limit is reached, but memory operations may
19040           still be scheduled.
19041
19042   LM32 Options
19043       These -m options are defined for the LatticeMico32 architecture:
19044
19045       -mbarrel-shift-enabled
19046           Enable barrel-shift instructions.
19047
19048       -mdivide-enabled
19049           Enable divide and modulus instructions.
19050
19051       -mmultiply-enabled
19052           Enable multiply instructions.
19053
19054       -msign-extend-enabled
19055           Enable sign extend instructions.
19056
19057       -muser-enabled
19058           Enable user-defined instructions.
19059
19060   M32C Options
19061       -mcpu=name
19062           Select the CPU for which code is generated.  name may be one of r8c
19063           for the R8C/Tiny series, m16c for the M16C (up to /60) series,
19064           m32cm for the M16C/80 series, or m32c for the M32C/80 series.
19065
19066       -msim
19067           Specifies that the program will be run on the simulator.  This
19068           causes an alternate runtime library to be linked in which supports,
19069           for example, file I/O.  You must not use this option when
19070           generating programs that will run on real hardware; you must
19071           provide your own runtime library for whatever I/O functions are
19072           needed.
19073
19074       -memregs=number
19075           Specifies the number of memory-based pseudo-registers GCC uses
19076           during code generation.  These pseudo-registers are used like real
19077           registers, so there is a tradeoff between GCC's ability to fit the
19078           code into available registers, and the performance penalty of using
19079           memory instead of registers.  Note that all modules in a program
19080           must be compiled with the same value for this option.  Because of
19081           that, you must not use this option with GCC's default runtime
19082           libraries.
19083
19084   M32R/D Options
19085       These -m options are defined for Renesas M32R/D architectures:
19086
19087       -m32r2
19088           Generate code for the M32R/2.
19089
19090       -m32rx
19091           Generate code for the M32R/X.
19092
19093       -m32r
19094           Generate code for the M32R.  This is the default.
19095
19096       -mmodel=small
19097           Assume all objects live in the lower 16MB of memory (so that their
19098           addresses can be loaded with the "ld24" instruction), and assume
19099           all subroutines are reachable with the "bl" instruction.  This is
19100           the default.
19101
19102           The addressability of a particular object can be set with the
19103           "model" attribute.
19104
19105       -mmodel=medium
19106           Assume objects may be anywhere in the 32-bit address space (the
19107           compiler generates "seth/add3" instructions to load their
19108           addresses), and assume all subroutines are reachable with the "bl"
19109           instruction.
19110
19111       -mmodel=large
19112           Assume objects may be anywhere in the 32-bit address space (the
19113           compiler generates "seth/add3" instructions to load their
19114           addresses), and assume subroutines may not be reachable with the
19115           "bl" instruction (the compiler generates the much slower
19116           "seth/add3/jl" instruction sequence).
19117
19118       -msdata=none
19119           Disable use of the small data area.  Variables are put into one of
19120           ".data", ".bss", or ".rodata" (unless the "section" attribute has
19121           been specified).  This is the default.
19122
19123           The small data area consists of sections ".sdata" and ".sbss".
19124           Objects may be explicitly put in the small data area with the
19125           "section" attribute using one of these sections.
19126
19127       -msdata=sdata
19128           Put small global and static data in the small data area, but do not
19129           generate special code to reference them.
19130
19131       -msdata=use
19132           Put small global and static data in the small data area, and
19133           generate special instructions to reference them.
19134
19135       -G num
19136           Put global and static objects less than or equal to num bytes into
19137           the small data or BSS sections instead of the normal data or BSS
19138           sections.  The default value of num is 8.  The -msdata option must
19139           be set to one of sdata or use for this option to have any effect.
19140
19141           All modules should be compiled with the same -G num value.
19142           Compiling with different values of num may or may not work; if it
19143           doesn't the linker gives an error message---incorrect code is not
19144           generated.
19145
19146       -mdebug
19147           Makes the M32R-specific code in the compiler display some
19148           statistics that might help in debugging programs.
19149
19150       -malign-loops
19151           Align all loops to a 32-byte boundary.
19152
19153       -mno-align-loops
19154           Do not enforce a 32-byte alignment for loops.  This is the default.
19155
19156       -missue-rate=number
19157           Issue number instructions per cycle.  number can only be 1 or 2.
19158
19159       -mbranch-cost=number
19160           number can only be 1 or 2.  If it is 1 then branches are preferred
19161           over conditional code, if it is 2, then the opposite applies.
19162
19163       -mflush-trap=number
19164           Specifies the trap number to use to flush the cache.  The default
19165           is 12.  Valid numbers are between 0 and 15 inclusive.
19166
19167       -mno-flush-trap
19168           Specifies that the cache cannot be flushed by using a trap.
19169
19170       -mflush-func=name
19171           Specifies the name of the operating system function to call to
19172           flush the cache.  The default is _flush_cache, but a function call
19173           is only used if a trap is not available.
19174
19175       -mno-flush-func
19176           Indicates that there is no OS function for flushing the cache.
19177
19178   M680x0 Options
19179       These are the -m options defined for M680x0 and ColdFire processors.
19180       The default settings depend on which architecture was selected when the
19181       compiler was configured; the defaults for the most common choices are
19182       given below.
19183
19184       -march=arch
19185           Generate code for a specific M680x0 or ColdFire instruction set
19186           architecture.  Permissible values of arch for M680x0 architectures
19187           are: 68000, 68010, 68020, 68030, 68040, 68060 and cpu32.  ColdFire
19188           architectures are selected according to Freescale's ISA
19189           classification and the permissible values are: isaa, isaaplus, isab
19190           and isac.
19191
19192           GCC defines a macro "__mcfarch__" whenever it is generating code
19193           for a ColdFire target.  The arch in this macro is one of the -march
19194           arguments given above.
19195
19196           When used together, -march and -mtune select code that runs on a
19197           family of similar processors but that is optimized for a particular
19198           microarchitecture.
19199
19200       -mcpu=cpu
19201           Generate code for a specific M680x0 or ColdFire processor.  The
19202           M680x0 cpus are: 68000, 68010, 68020, 68030, 68040, 68060, 68302,
19203           68332 and cpu32.  The ColdFire cpus are given by the table below,
19204           which also classifies the CPUs into families:
19205
19206           Family : -mcpu arguments
19207           51 : 51 51ac 51ag 51cn 51em 51je 51jf 51jg 51jm 51mm 51qe 51qm
19208           5206 : 5202 5204 5206
19209           5206e : 5206e
19210           5208 : 5207 5208
19211           5211a : 5210a 5211a
19212           5213 : 5211 5212 5213
19213           5216 : 5214 5216
19214           52235 : 52230 52231 52232 52233 52234 52235
19215           5225 : 5224 5225
19216           52259 : 52252 52254 52255 52256 52258 52259
19217           5235 : 5232 5233 5234 5235 523x
19218           5249 : 5249
19219           5250 : 5250
19220           5271 : 5270 5271
19221           5272 : 5272
19222           5275 : 5274 5275
19223           5282 : 5280 5281 5282 528x
19224           53017 : 53011 53012 53013 53014 53015 53016 53017
19225           5307 : 5307
19226           5329 : 5327 5328 5329 532x
19227           5373 : 5372 5373 537x
19228           5407 : 5407
19229           5475 : 5470 5471 5472 5473 5474 5475 547x 5480 5481 5482 5483 5484
19230           5485
19231
19232           -mcpu=cpu overrides -march=arch if arch is compatible with cpu.
19233           Other combinations of -mcpu and -march are rejected.
19234
19235           GCC defines the macro "__mcf_cpu_cpu" when ColdFire target cpu is
19236           selected.  It also defines "__mcf_family_family", where the value
19237           of family is given by the table above.
19238
19239       -mtune=tune
19240           Tune the code for a particular microarchitecture within the
19241           constraints set by -march and -mcpu.  The M680x0 microarchitectures
19242           are: 68000, 68010, 68020, 68030, 68040, 68060 and cpu32.  The
19243           ColdFire microarchitectures are: cfv1, cfv2, cfv3, cfv4 and cfv4e.
19244
19245           You can also use -mtune=68020-40 for code that needs to run
19246           relatively well on 68020, 68030 and 68040 targets.  -mtune=68020-60
19247           is similar but includes 68060 targets as well.  These two options
19248           select the same tuning decisions as -m68020-40 and -m68020-60
19249           respectively.
19250
19251           GCC defines the macros "__mcarch" and "__mcarch__" when tuning for
19252           680x0 architecture arch.  It also defines "mcarch" unless either
19253           -ansi or a non-GNU -std option is used.  If GCC is tuning for a
19254           range of architectures, as selected by -mtune=68020-40 or
19255           -mtune=68020-60, it defines the macros for every architecture in
19256           the range.
19257
19258           GCC also defines the macro "__muarch__" when tuning for ColdFire
19259           microarchitecture uarch, where uarch is one of the arguments given
19260           above.
19261
19262       -m68000
19263       -mc68000
19264           Generate output for a 68000.  This is the default when the compiler
19265           is configured for 68000-based systems.  It is equivalent to
19266           -march=68000.
19267
19268           Use this option for microcontrollers with a 68000 or EC000 core,
19269           including the 68008, 68302, 68306, 68307, 68322, 68328 and 68356.
19270
19271       -m68010
19272           Generate output for a 68010.  This is the default when the compiler
19273           is configured for 68010-based systems.  It is equivalent to
19274           -march=68010.
19275
19276       -m68020
19277       -mc68020
19278           Generate output for a 68020.  This is the default when the compiler
19279           is configured for 68020-based systems.  It is equivalent to
19280           -march=68020.
19281
19282       -m68030
19283           Generate output for a 68030.  This is the default when the compiler
19284           is configured for 68030-based systems.  It is equivalent to
19285           -march=68030.
19286
19287       -m68040
19288           Generate output for a 68040.  This is the default when the compiler
19289           is configured for 68040-based systems.  It is equivalent to
19290           -march=68040.
19291
19292           This option inhibits the use of 68881/68882 instructions that have
19293           to be emulated by software on the 68040.  Use this option if your
19294           68040 does not have code to emulate those instructions.
19295
19296       -m68060
19297           Generate output for a 68060.  This is the default when the compiler
19298           is configured for 68060-based systems.  It is equivalent to
19299           -march=68060.
19300
19301           This option inhibits the use of 68020 and 68881/68882 instructions
19302           that have to be emulated by software on the 68060.  Use this option
19303           if your 68060 does not have code to emulate those instructions.
19304
19305       -mcpu32
19306           Generate output for a CPU32.  This is the default when the compiler
19307           is configured for CPU32-based systems.  It is equivalent to
19308           -march=cpu32.
19309
19310           Use this option for microcontrollers with a CPU32 or CPU32+ core,
19311           including the 68330, 68331, 68332, 68333, 68334, 68336, 68340,
19312           68341, 68349 and 68360.
19313
19314       -m5200
19315           Generate output for a 520X ColdFire CPU.  This is the default when
19316           the compiler is configured for 520X-based systems.  It is
19317           equivalent to -mcpu=5206, and is now deprecated in favor of that
19318           option.
19319
19320           Use this option for microcontroller with a 5200 core, including the
19321           MCF5202, MCF5203, MCF5204 and MCF5206.
19322
19323       -m5206e
19324           Generate output for a 5206e ColdFire CPU.  The option is now
19325           deprecated in favor of the equivalent -mcpu=5206e.
19326
19327       -m528x
19328           Generate output for a member of the ColdFire 528X family.  The
19329           option is now deprecated in favor of the equivalent -mcpu=528x.
19330
19331       -m5307
19332           Generate output for a ColdFire 5307 CPU.  The option is now
19333           deprecated in favor of the equivalent -mcpu=5307.
19334
19335       -m5407
19336           Generate output for a ColdFire 5407 CPU.  The option is now
19337           deprecated in favor of the equivalent -mcpu=5407.
19338
19339       -mcfv4e
19340           Generate output for a ColdFire V4e family CPU (e.g. 547x/548x).
19341           This includes use of hardware floating-point instructions.  The
19342           option is equivalent to -mcpu=547x, and is now deprecated in favor
19343           of that option.
19344
19345       -m68020-40
19346           Generate output for a 68040, without using any of the new
19347           instructions.  This results in code that can run relatively
19348           efficiently on either a 68020/68881 or a 68030 or a 68040.  The
19349           generated code does use the 68881 instructions that are emulated on
19350           the 68040.
19351
19352           The option is equivalent to -march=68020 -mtune=68020-40.
19353
19354       -m68020-60
19355           Generate output for a 68060, without using any of the new
19356           instructions.  This results in code that can run relatively
19357           efficiently on either a 68020/68881 or a 68030 or a 68040.  The
19358           generated code does use the 68881 instructions that are emulated on
19359           the 68060.
19360
19361           The option is equivalent to -march=68020 -mtune=68020-60.
19362
19363       -mhard-float
19364       -m68881
19365           Generate floating-point instructions.  This is the default for
19366           68020 and above, and for ColdFire devices that have an FPU.  It
19367           defines the macro "__HAVE_68881__" on M680x0 targets and
19368           "__mcffpu__" on ColdFire targets.
19369
19370       -msoft-float
19371           Do not generate floating-point instructions; use library calls
19372           instead.  This is the default for 68000, 68010, and 68832 targets.
19373           It is also the default for ColdFire devices that have no FPU.
19374
19375       -mdiv
19376       -mno-div
19377           Generate (do not generate) ColdFire hardware divide and remainder
19378           instructions.  If -march is used without -mcpu, the default is "on"
19379           for ColdFire architectures and "off" for M680x0 architectures.
19380           Otherwise, the default is taken from the target CPU (either the
19381           default CPU, or the one specified by -mcpu).  For example, the
19382           default is "off" for -mcpu=5206 and "on" for -mcpu=5206e.
19383
19384           GCC defines the macro "__mcfhwdiv__" when this option is enabled.
19385
19386       -mshort
19387           Consider type "int" to be 16 bits wide, like "short int".
19388           Additionally, parameters passed on the stack are also aligned to a
19389           16-bit boundary even on targets whose API mandates promotion to
19390           32-bit.
19391
19392       -mno-short
19393           Do not consider type "int" to be 16 bits wide.  This is the
19394           default.
19395
19396       -mnobitfield
19397       -mno-bitfield
19398           Do not use the bit-field instructions.  The -m68000, -mcpu32 and
19399           -m5200 options imply -mnobitfield.
19400
19401       -mbitfield
19402           Do use the bit-field instructions.  The -m68020 option implies
19403           -mbitfield.  This is the default if you use a configuration
19404           designed for a 68020.
19405
19406       -mrtd
19407           Use a different function-calling convention, in which functions
19408           that take a fixed number of arguments return with the "rtd"
19409           instruction, which pops their arguments while returning.  This
19410           saves one instruction in the caller since there is no need to pop
19411           the arguments there.
19412
19413           This calling convention is incompatible with the one normally used
19414           on Unix, so you cannot use it if you need to call libraries
19415           compiled with the Unix compiler.
19416
19417           Also, you must provide function prototypes for all functions that
19418           take variable numbers of arguments (including "printf"); otherwise
19419           incorrect code is generated for calls to those functions.
19420
19421           In addition, seriously incorrect code results if you call a
19422           function with too many arguments.  (Normally, extra arguments are
19423           harmlessly ignored.)
19424
19425           The "rtd" instruction is supported by the 68010, 68020, 68030,
19426           68040, 68060 and CPU32 processors, but not by the 68000 or 5200.
19427
19428           The default is -mno-rtd.
19429
19430       -malign-int
19431       -mno-align-int
19432           Control whether GCC aligns "int", "long", "long long", "float",
19433           "double", and "long double" variables on a 32-bit boundary
19434           (-malign-int) or a 16-bit boundary (-mno-align-int).  Aligning
19435           variables on 32-bit boundaries produces code that runs somewhat
19436           faster on processors with 32-bit busses at the expense of more
19437           memory.
19438
19439           Warning: if you use the -malign-int switch, GCC aligns structures
19440           containing the above types differently than most published
19441           application binary interface specifications for the m68k.
19442
19443           Use the pc-relative addressing mode of the 68000 directly, instead
19444           of using a global offset table.  At present, this option implies
19445           -fpic, allowing at most a 16-bit offset for pc-relative addressing.
19446           -fPIC is not presently supported with -mpcrel, though this could be
19447           supported for 68020 and higher processors.
19448
19449       -mno-strict-align
19450       -mstrict-align
19451           Do not (do) assume that unaligned memory references are handled by
19452           the system.
19453
19454       -msep-data
19455           Generate code that allows the data segment to be located in a
19456           different area of memory from the text segment.  This allows for
19457           execute-in-place in an environment without virtual memory
19458           management.  This option implies -fPIC.
19459
19460       -mno-sep-data
19461           Generate code that assumes that the data segment follows the text
19462           segment.  This is the default.
19463
19464       -mid-shared-library
19465           Generate code that supports shared libraries via the library ID
19466           method.  This allows for execute-in-place and shared libraries in
19467           an environment without virtual memory management.  This option
19468           implies -fPIC.
19469
19470       -mno-id-shared-library
19471           Generate code that doesn't assume ID-based shared libraries are
19472           being used.  This is the default.
19473
19474       -mshared-library-id=n
19475           Specifies the identification number of the ID-based shared library
19476           being compiled.  Specifying a value of 0 generates more compact
19477           code; specifying other values forces the allocation of that number
19478           to the current library, but is no more space- or time-efficient
19479           than omitting this option.
19480
19481       -mxgot
19482       -mno-xgot
19483           When generating position-independent code for ColdFire, generate
19484           code that works if the GOT has more than 8192 entries.  This code
19485           is larger and slower than code generated without this option.  On
19486           M680x0 processors, this option is not needed; -fPIC suffices.
19487
19488           GCC normally uses a single instruction to load values from the GOT.
19489           While this is relatively efficient, it only works if the GOT is
19490           smaller than about 64k.  Anything larger causes the linker to
19491           report an error such as:
19492
19493                   relocation truncated to fit: R_68K_GOT16O foobar
19494
19495           If this happens, you should recompile your code with -mxgot.  It
19496           should then work with very large GOTs.  However, code generated
19497           with -mxgot is less efficient, since it takes 4 instructions to
19498           fetch the value of a global symbol.
19499
19500           Note that some linkers, including newer versions of the GNU linker,
19501           can create multiple GOTs and sort GOT entries.  If you have such a
19502           linker, you should only need to use -mxgot when compiling a single
19503           object file that accesses more than 8192 GOT entries.  Very few do.
19504
19505           These options have no effect unless GCC is generating position-
19506           independent code.
19507
19508       -mlong-jump-table-offsets
19509           Use 32-bit offsets in "switch" tables.  The default is to use
19510           16-bit offsets.
19511
19512   MCore Options
19513       These are the -m options defined for the Motorola M*Core processors.
19514
19515       -mhardlit
19516       -mno-hardlit
19517           Inline constants into the code stream if it can be done in two
19518           instructions or less.
19519
19520       -mdiv
19521       -mno-div
19522           Use the divide instruction.  (Enabled by default).
19523
19524       -mrelax-immediate
19525       -mno-relax-immediate
19526           Allow arbitrary-sized immediates in bit operations.
19527
19528       -mwide-bitfields
19529       -mno-wide-bitfields
19530           Always treat bit-fields as "int"-sized.
19531
19532       -m4byte-functions
19533       -mno-4byte-functions
19534           Force all functions to be aligned to a 4-byte boundary.
19535
19536       -mcallgraph-data
19537       -mno-callgraph-data
19538           Emit callgraph information.
19539
19540       -mslow-bytes
19541       -mno-slow-bytes
19542           Prefer word access when reading byte quantities.
19543
19544       -mlittle-endian
19545       -mbig-endian
19546           Generate code for a little-endian target.
19547
19548       -m210
19549       -m340
19550           Generate code for the 210 processor.
19551
19552       -mno-lsim
19553           Assume that runtime support has been provided and so omit the
19554           simulator library (libsim.a) from the linker command line.
19555
19556       -mstack-increment=size
19557           Set the maximum amount for a single stack increment operation.
19558           Large values can increase the speed of programs that contain
19559           functions that need a large amount of stack space, but they can
19560           also trigger a segmentation fault if the stack is extended too
19561           much.  The default value is 0x1000.
19562
19563   MeP Options
19564       -mabsdiff
19565           Enables the "abs" instruction, which is the absolute difference
19566           between two registers.
19567
19568       -mall-opts
19569           Enables all the optional instructions---average, multiply, divide,
19570           bit operations, leading zero, absolute difference, min/max, clip,
19571           and saturation.
19572
19573       -maverage
19574           Enables the "ave" instruction, which computes the average of two
19575           registers.
19576
19577       -mbased=n
19578           Variables of size n bytes or smaller are placed in the ".based"
19579           section by default.  Based variables use the $tp register as a base
19580           register, and there is a 128-byte limit to the ".based" section.
19581
19582       -mbitops
19583           Enables the bit operation instructions---bit test ("btstm"), set
19584           ("bsetm"), clear ("bclrm"), invert ("bnotm"), and test-and-set
19585           ("tas").
19586
19587       -mc=name
19588           Selects which section constant data is placed in.  name may be
19589           tiny, near, or far.
19590
19591       -mclip
19592           Enables the "clip" instruction.  Note that -mclip is not useful
19593           unless you also provide -mminmax.
19594
19595       -mconfig=name
19596           Selects one of the built-in core configurations.  Each MeP chip has
19597           one or more modules in it; each module has a core CPU and a variety
19598           of coprocessors, optional instructions, and peripherals.  The
19599           "MeP-Integrator" tool, not part of GCC, provides these
19600           configurations through this option; using this option is the same
19601           as using all the corresponding command-line options.  The default
19602           configuration is default.
19603
19604       -mcop
19605           Enables the coprocessor instructions.  By default, this is a 32-bit
19606           coprocessor.  Note that the coprocessor is normally enabled via the
19607           -mconfig= option.
19608
19609       -mcop32
19610           Enables the 32-bit coprocessor's instructions.
19611
19612       -mcop64
19613           Enables the 64-bit coprocessor's instructions.
19614
19615       -mivc2
19616           Enables IVC2 scheduling.  IVC2 is a 64-bit VLIW coprocessor.
19617
19618       -mdc
19619           Causes constant variables to be placed in the ".near" section.
19620
19621       -mdiv
19622           Enables the "div" and "divu" instructions.
19623
19624       -meb
19625           Generate big-endian code.
19626
19627       -mel
19628           Generate little-endian code.
19629
19630       -mio-volatile
19631           Tells the compiler that any variable marked with the "io" attribute
19632           is to be considered volatile.
19633
19634       -ml Causes variables to be assigned to the ".far" section by default.
19635
19636       -mleadz
19637           Enables the "leadz" (leading zero) instruction.
19638
19639       -mm Causes variables to be assigned to the ".near" section by default.
19640
19641       -mminmax
19642           Enables the "min" and "max" instructions.
19643
19644       -mmult
19645           Enables the multiplication and multiply-accumulate instructions.
19646
19647       -mno-opts
19648           Disables all the optional instructions enabled by -mall-opts.
19649
19650       -mrepeat
19651           Enables the "repeat" and "erepeat" instructions, used for low-
19652           overhead looping.
19653
19654       -ms Causes all variables to default to the ".tiny" section.  Note that
19655           there is a 65536-byte limit to this section.  Accesses to these
19656           variables use the %gp base register.
19657
19658       -msatur
19659           Enables the saturation instructions.  Note that the compiler does
19660           not currently generate these itself, but this option is included
19661           for compatibility with other tools, like "as".
19662
19663       -msdram
19664           Link the SDRAM-based runtime instead of the default ROM-based
19665           runtime.
19666
19667       -msim
19668           Link the simulator run-time libraries.
19669
19670       -msimnovec
19671           Link the simulator runtime libraries, excluding built-in support
19672           for reset and exception vectors and tables.
19673
19674       -mtf
19675           Causes all functions to default to the ".far" section.  Without
19676           this option, functions default to the ".near" section.
19677
19678       -mtiny=n
19679           Variables that are n bytes or smaller are allocated to the ".tiny"
19680           section.  These variables use the $gp base register.  The default
19681           for this option is 4, but note that there's a 65536-byte limit to
19682           the ".tiny" section.
19683
19684   MicroBlaze Options
19685       -msoft-float
19686           Use software emulation for floating point (default).
19687
19688       -mhard-float
19689           Use hardware floating-point instructions.
19690
19691       -mmemcpy
19692           Do not optimize block moves, use "memcpy".
19693
19694       -mno-clearbss
19695           This option is deprecated.  Use -fno-zero-initialized-in-bss
19696           instead.
19697
19698       -mcpu=cpu-type
19699           Use features of, and schedule code for, the given CPU.  Supported
19700           values are in the format vX.YY.Z, where X is a major version, YY is
19701           the minor version, and Z is compatibility code.  Example values are
19702           v3.00.a, v4.00.b, v5.00.a, v5.00.b, v6.00.a.
19703
19704       -mxl-soft-mul
19705           Use software multiply emulation (default).
19706
19707       -mxl-soft-div
19708           Use software emulation for divides (default).
19709
19710       -mxl-barrel-shift
19711           Use the hardware barrel shifter.
19712
19713       -mxl-pattern-compare
19714           Use pattern compare instructions.
19715
19716       -msmall-divides
19717           Use table lookup optimization for small signed integer divisions.
19718
19719       -mxl-stack-check
19720           This option is deprecated.  Use -fstack-check instead.
19721
19722       -mxl-gp-opt
19723           Use GP-relative ".sdata"/".sbss" sections.
19724
19725       -mxl-multiply-high
19726           Use multiply high instructions for high part of 32x32 multiply.
19727
19728       -mxl-float-convert
19729           Use hardware floating-point conversion instructions.
19730
19731       -mxl-float-sqrt
19732           Use hardware floating-point square root instruction.
19733
19734       -mbig-endian
19735           Generate code for a big-endian target.
19736
19737       -mlittle-endian
19738           Generate code for a little-endian target.
19739
19740       -mxl-reorder
19741           Use reorder instructions (swap and byte reversed load/store).
19742
19743       -mxl-mode-app-model
19744           Select application model app-model.  Valid models are
19745
19746           executable
19747               normal executable (default), uses startup code crt0.o.
19748
19749           -mpic-data-is-text-relative
19750               Assume that the displacement between the text and data segments
19751               is fixed at static link time.  This allows data to be
19752               referenced by offset from start of text address instead of GOT
19753               since PC-relative addressing is not supported.
19754
19755           xmdstub
19756               for use with Xilinx Microprocessor Debugger (XMD) based
19757               software intrusive debug agent called xmdstub. This uses
19758               startup file crt1.o and sets the start address of the program
19759               to 0x800.
19760
19761           bootstrap
19762               for applications that are loaded using a bootloader.  This
19763               model uses startup file crt2.o which does not contain a
19764               processor reset vector handler. This is suitable for
19765               transferring control on a processor reset to the bootloader
19766               rather than the application.
19767
19768           novectors
19769               for applications that do not require any of the MicroBlaze
19770               vectors. This option may be useful for applications running
19771               within a monitoring application. This model uses crt3.o as a
19772               startup file.
19773
19774           Option -xl-mode-app-model is a deprecated alias for -mxl-mode-app-
19775           model.
19776
19777   MIPS Options
19778       -EB Generate big-endian code.
19779
19780       -EL Generate little-endian code.  This is the default for mips*el-*-*
19781           configurations.
19782
19783       -march=arch
19784           Generate code that runs on arch, which can be the name of a generic
19785           MIPS ISA, or the name of a particular processor.  The ISA names
19786           are: mips1, mips2, mips3, mips4, mips32, mips32r2, mips32r3,
19787           mips32r5, mips32r6, mips64, mips64r2, mips64r3, mips64r5 and
19788           mips64r6.  The processor names are: 4kc, 4km, 4kp, 4ksc, 4kec,
19789           4kem, 4kep, 4ksd, 5kc, 5kf, 20kc, 24kc, 24kf2_1, 24kf1_1, 24kec,
19790           24kef2_1, 24kef1_1, 34kc, 34kf2_1, 34kf1_1, 34kn, 74kc, 74kf2_1,
19791           74kf1_1, 74kf3_2, 1004kc, 1004kf2_1, 1004kf1_1, i6400, i6500,
19792           interaptiv, loongson2e, loongson2f, loongson3a, gs464, gs464e,
19793           gs264e, m4k, m14k, m14kc, m14ke, m14kec, m5100, m5101, octeon,
19794           octeon+, octeon2, octeon3, orion, p5600, p6600, r2000, r3000,
19795           r3900, r4000, r4400, r4600, r4650, r4700, r5900, r6000, r8000,
19796           rm7000, rm9000, r10000, r12000, r14000, r16000, sb1, sr71000,
19797           vr4100, vr4111, vr4120, vr4130, vr4300, vr5000, vr5400, vr5500, xlr
19798           and xlp.  The special value from-abi selects the most compatible
19799           architecture for the selected ABI (that is, mips1 for 32-bit ABIs
19800           and mips3 for 64-bit ABIs).
19801
19802           The native Linux/GNU toolchain also supports the value native,
19803           which selects the best architecture option for the host processor.
19804           -march=native has no effect if GCC does not recognize the
19805           processor.
19806
19807           In processor names, a final 000 can be abbreviated as k (for
19808           example, -march=r2k).  Prefixes are optional, and vr may be written
19809           r.
19810
19811           Names of the form nf2_1 refer to processors with FPUs clocked at
19812           half the rate of the core, names of the form nf1_1 refer to
19813           processors with FPUs clocked at the same rate as the core, and
19814           names of the form nf3_2 refer to processors with FPUs clocked a
19815           ratio of 3:2 with respect to the core.  For compatibility reasons,
19816           nf is accepted as a synonym for nf2_1 while nx and bfx are accepted
19817           as synonyms for nf1_1.
19818
19819           GCC defines two macros based on the value of this option.  The
19820           first is "_MIPS_ARCH", which gives the name of target architecture,
19821           as a string.  The second has the form "_MIPS_ARCH_foo", where foo
19822           is the capitalized value of "_MIPS_ARCH".  For example,
19823           -march=r2000 sets "_MIPS_ARCH" to "r2000" and defines the macro
19824           "_MIPS_ARCH_R2000".
19825
19826           Note that the "_MIPS_ARCH" macro uses the processor names given
19827           above.  In other words, it has the full prefix and does not
19828           abbreviate 000 as k.  In the case of from-abi, the macro names the
19829           resolved architecture (either "mips1" or "mips3").  It names the
19830           default architecture when no -march option is given.
19831
19832       -mtune=arch
19833           Optimize for arch.  Among other things, this option controls the
19834           way instructions are scheduled, and the perceived cost of
19835           arithmetic operations.  The list of arch values is the same as for
19836           -march.
19837
19838           When this option is not used, GCC optimizes for the processor
19839           specified by -march.  By using -march and -mtune together, it is
19840           possible to generate code that runs on a family of processors, but
19841           optimize the code for one particular member of that family.
19842
19843           -mtune defines the macros "_MIPS_TUNE" and "_MIPS_TUNE_foo", which
19844           work in the same way as the -march ones described above.
19845
19846       -mips1
19847           Equivalent to -march=mips1.
19848
19849       -mips2
19850           Equivalent to -march=mips2.
19851
19852       -mips3
19853           Equivalent to -march=mips3.
19854
19855       -mips4
19856           Equivalent to -march=mips4.
19857
19858       -mips32
19859           Equivalent to -march=mips32.
19860
19861       -mips32r3
19862           Equivalent to -march=mips32r3.
19863
19864       -mips32r5
19865           Equivalent to -march=mips32r5.
19866
19867       -mips32r6
19868           Equivalent to -march=mips32r6.
19869
19870       -mips64
19871           Equivalent to -march=mips64.
19872
19873       -mips64r2
19874           Equivalent to -march=mips64r2.
19875
19876       -mips64r3
19877           Equivalent to -march=mips64r3.
19878
19879       -mips64r5
19880           Equivalent to -march=mips64r5.
19881
19882       -mips64r6
19883           Equivalent to -march=mips64r6.
19884
19885       -mips16
19886       -mno-mips16
19887           Generate (do not generate) MIPS16 code.  If GCC is targeting a
19888           MIPS32 or MIPS64 architecture, it makes use of the MIPS16e ASE.
19889
19890           MIPS16 code generation can also be controlled on a per-function
19891           basis by means of "mips16" and "nomips16" attributes.
19892
19893       -mflip-mips16
19894           Generate MIPS16 code on alternating functions.  This option is
19895           provided for regression testing of mixed MIPS16/non-MIPS16 code
19896           generation, and is not intended for ordinary use in compiling user
19897           code.
19898
19899       -minterlink-compressed
19900       -mno-interlink-compressed
19901           Require (do not require) that code using the standard
19902           (uncompressed) MIPS ISA be link-compatible with MIPS16 and
19903           microMIPS code, and vice versa.
19904
19905           For example, code using the standard ISA encoding cannot jump
19906           directly to MIPS16 or microMIPS code; it must either use a call or
19907           an indirect jump.  -minterlink-compressed therefore disables direct
19908           jumps unless GCC knows that the target of the jump is not
19909           compressed.
19910
19911       -minterlink-mips16
19912       -mno-interlink-mips16
19913           Aliases of -minterlink-compressed and -mno-interlink-compressed.
19914           These options predate the microMIPS ASE and are retained for
19915           backwards compatibility.
19916
19917       -mabi=32
19918       -mabi=o64
19919       -mabi=n32
19920       -mabi=64
19921       -mabi=eabi
19922           Generate code for the given ABI.
19923
19924           Note that the EABI has a 32-bit and a 64-bit variant.  GCC normally
19925           generates 64-bit code when you select a 64-bit architecture, but
19926           you can use -mgp32 to get 32-bit code instead.
19927
19928           For information about the O64 ABI, see
19929           <http://gcc.gnu.org/projects/mipso64-abi.html>.
19930
19931           GCC supports a variant of the o32 ABI in which floating-point
19932           registers are 64 rather than 32 bits wide.  You can select this
19933           combination with -mabi=32 -mfp64.  This ABI relies on the "mthc1"
19934           and "mfhc1" instructions and is therefore only supported for
19935           MIPS32R2, MIPS32R3 and MIPS32R5 processors.
19936
19937           The register assignments for arguments and return values remain the
19938           same, but each scalar value is passed in a single 64-bit register
19939           rather than a pair of 32-bit registers.  For example, scalar
19940           floating-point values are returned in $f0 only, not a $f0/$f1 pair.
19941           The set of call-saved registers also remains the same in that the
19942           even-numbered double-precision registers are saved.
19943
19944           Two additional variants of the o32 ABI are supported to enable a
19945           transition from 32-bit to 64-bit registers.  These are FPXX
19946           (-mfpxx) and FP64A (-mfp64 -mno-odd-spreg).  The FPXX extension
19947           mandates that all code must execute correctly when run using 32-bit
19948           or 64-bit registers.  The code can be interlinked with either FP32
19949           or FP64, but not both.  The FP64A extension is similar to the FP64
19950           extension but forbids the use of odd-numbered single-precision
19951           registers.  This can be used in conjunction with the "FRE" mode of
19952           FPUs in MIPS32R5 processors and allows both FP32 and FP64A code to
19953           interlink and run in the same process without changing FPU modes.
19954
19955       -mabicalls
19956       -mno-abicalls
19957           Generate (do not generate) code that is suitable for SVR4-style
19958           dynamic objects.  -mabicalls is the default for SVR4-based systems.
19959
19960       -mshared
19961       -mno-shared
19962           Generate (do not generate) code that is fully position-independent,
19963           and that can therefore be linked into shared libraries.  This
19964           option only affects -mabicalls.
19965
19966           All -mabicalls code has traditionally been position-independent,
19967           regardless of options like -fPIC and -fpic.  However, as an
19968           extension, the GNU toolchain allows executables to use absolute
19969           accesses for locally-binding symbols.  It can also use shorter GP
19970           initialization sequences and generate direct calls to locally-
19971           defined functions.  This mode is selected by -mno-shared.
19972
19973           -mno-shared depends on binutils 2.16 or higher and generates
19974           objects that can only be linked by the GNU linker.  However, the
19975           option does not affect the ABI of the final executable; it only
19976           affects the ABI of relocatable objects.  Using -mno-shared
19977           generally makes executables both smaller and quicker.
19978
19979           -mshared is the default.
19980
19981       -mplt
19982       -mno-plt
19983           Assume (do not assume) that the static and dynamic linkers support
19984           PLTs and copy relocations.  This option only affects -mno-shared
19985           -mabicalls.  For the n64 ABI, this option has no effect without
19986           -msym32.
19987
19988           You can make -mplt the default by configuring GCC with
19989           --with-mips-plt.  The default is -mno-plt otherwise.
19990
19991       -mxgot
19992       -mno-xgot
19993           Lift (do not lift) the usual restrictions on the size of the global
19994           offset table.
19995
19996           GCC normally uses a single instruction to load values from the GOT.
19997           While this is relatively efficient, it only works if the GOT is
19998           smaller than about 64k.  Anything larger causes the linker to
19999           report an error such as:
20000
20001                   relocation truncated to fit: R_MIPS_GOT16 foobar
20002
20003           If this happens, you should recompile your code with -mxgot.  This
20004           works with very large GOTs, although the code is also less
20005           efficient, since it takes three instructions to fetch the value of
20006           a global symbol.
20007
20008           Note that some linkers can create multiple GOTs.  If you have such
20009           a linker, you should only need to use -mxgot when a single object
20010           file accesses more than 64k's worth of GOT entries.  Very few do.
20011
20012           These options have no effect unless GCC is generating position
20013           independent code.
20014
20015       -mgp32
20016           Assume that general-purpose registers are 32 bits wide.
20017
20018       -mgp64
20019           Assume that general-purpose registers are 64 bits wide.
20020
20021       -mfp32
20022           Assume that floating-point registers are 32 bits wide.
20023
20024       -mfp64
20025           Assume that floating-point registers are 64 bits wide.
20026
20027       -mfpxx
20028           Do not assume the width of floating-point registers.
20029
20030       -mhard-float
20031           Use floating-point coprocessor instructions.
20032
20033       -msoft-float
20034           Do not use floating-point coprocessor instructions.  Implement
20035           floating-point calculations using library calls instead.
20036
20037       -mno-float
20038           Equivalent to -msoft-float, but additionally asserts that the
20039           program being compiled does not perform any floating-point
20040           operations.  This option is presently supported only by some bare-
20041           metal MIPS configurations, where it may select a special set of
20042           libraries that lack all floating-point support (including, for
20043           example, the floating-point "printf" formats).  If code compiled
20044           with -mno-float accidentally contains floating-point operations, it
20045           is likely to suffer a link-time or run-time failure.
20046
20047       -msingle-float
20048           Assume that the floating-point coprocessor only supports single-
20049           precision operations.
20050
20051       -mdouble-float
20052           Assume that the floating-point coprocessor supports double-
20053           precision operations.  This is the default.
20054
20055       -modd-spreg
20056       -mno-odd-spreg
20057           Enable the use of odd-numbered single-precision floating-point
20058           registers for the o32 ABI.  This is the default for processors that
20059           are known to support these registers.  When using the o32 FPXX ABI,
20060           -mno-odd-spreg is set by default.
20061
20062       -mabs=2008
20063       -mabs=legacy
20064           These options control the treatment of the special not-a-number
20065           (NaN) IEEE 754 floating-point data with the "abs.fmt" and "neg.fmt"
20066           machine instructions.
20067
20068           By default or when -mabs=legacy is used the legacy treatment is
20069           selected.  In this case these instructions are considered
20070           arithmetic and avoided where correct operation is required and the
20071           input operand might be a NaN.  A longer sequence of instructions
20072           that manipulate the sign bit of floating-point datum manually is
20073           used instead unless the -ffinite-math-only option has also been
20074           specified.
20075
20076           The -mabs=2008 option selects the IEEE 754-2008 treatment.  In this
20077           case these instructions are considered non-arithmetic and therefore
20078           operating correctly in all cases, including in particular where the
20079           input operand is a NaN.  These instructions are therefore always
20080           used for the respective operations.
20081
20082       -mnan=2008
20083       -mnan=legacy
20084           These options control the encoding of the special not-a-number
20085           (NaN) IEEE 754 floating-point data.
20086
20087           The -mnan=legacy option selects the legacy encoding.  In this case
20088           quiet NaNs (qNaNs) are denoted by the first bit of their trailing
20089           significand field being 0, whereas signaling NaNs (sNaNs) are
20090           denoted by the first bit of their trailing significand field being
20091           1.
20092
20093           The -mnan=2008 option selects the IEEE 754-2008 encoding.  In this
20094           case qNaNs are denoted by the first bit of their trailing
20095           significand field being 1, whereas sNaNs are denoted by the first
20096           bit of their trailing significand field being 0.
20097
20098           The default is -mnan=legacy unless GCC has been configured with
20099           --with-nan=2008.
20100
20101       -mllsc
20102       -mno-llsc
20103           Use (do not use) ll, sc, and sync instructions to implement atomic
20104           memory built-in functions.  When neither option is specified, GCC
20105           uses the instructions if the target architecture supports them.
20106
20107           -mllsc is useful if the runtime environment can emulate the
20108           instructions and -mno-llsc can be useful when compiling for
20109           nonstandard ISAs.  You can make either option the default by
20110           configuring GCC with --with-llsc and --without-llsc respectively.
20111           --with-llsc is the default for some configurations; see the
20112           installation documentation for details.
20113
20114       -mdsp
20115       -mno-dsp
20116           Use (do not use) revision 1 of the MIPS DSP ASE.
20117             This option defines the preprocessor macro "__mips_dsp".  It also
20118           defines "__mips_dsp_rev" to 1.
20119
20120       -mdspr2
20121       -mno-dspr2
20122           Use (do not use) revision 2 of the MIPS DSP ASE.
20123             This option defines the preprocessor macros "__mips_dsp" and
20124           "__mips_dspr2".  It also defines "__mips_dsp_rev" to 2.
20125
20126       -msmartmips
20127       -mno-smartmips
20128           Use (do not use) the MIPS SmartMIPS ASE.
20129
20130       -mpaired-single
20131       -mno-paired-single
20132           Use (do not use) paired-single floating-point instructions.
20133             This option requires hardware floating-point support to be
20134           enabled.
20135
20136       -mdmx
20137       -mno-mdmx
20138           Use (do not use) MIPS Digital Media Extension instructions.  This
20139           option can only be used when generating 64-bit code and requires
20140           hardware floating-point support to be enabled.
20141
20142       -mips3d
20143       -mno-mips3d
20144           Use (do not use) the MIPS-3D ASE.  The option -mips3d implies
20145           -mpaired-single.
20146
20147       -mmicromips
20148       -mno-micromips
20149           Generate (do not generate) microMIPS code.
20150
20151           MicroMIPS code generation can also be controlled on a per-function
20152           basis by means of "micromips" and "nomicromips" attributes.
20153
20154       -mmt
20155       -mno-mt
20156           Use (do not use) MT Multithreading instructions.
20157
20158       -mmcu
20159       -mno-mcu
20160           Use (do not use) the MIPS MCU ASE instructions.
20161
20162       -meva
20163       -mno-eva
20164           Use (do not use) the MIPS Enhanced Virtual Addressing instructions.
20165
20166       -mvirt
20167       -mno-virt
20168           Use (do not use) the MIPS Virtualization (VZ) instructions.
20169
20170       -mxpa
20171       -mno-xpa
20172           Use (do not use) the MIPS eXtended Physical Address (XPA)
20173           instructions.
20174
20175       -mcrc
20176       -mno-crc
20177           Use (do not use) the MIPS Cyclic Redundancy Check (CRC)
20178           instructions.
20179
20180       -mginv
20181       -mno-ginv
20182           Use (do not use) the MIPS Global INValidate (GINV) instructions.
20183
20184       -mloongson-mmi
20185       -mno-loongson-mmi
20186           Use (do not use) the MIPS Loongson MultiMedia extensions
20187           Instructions (MMI).
20188
20189       -mloongson-ext
20190       -mno-loongson-ext
20191           Use (do not use) the MIPS Loongson EXTensions (EXT) instructions.
20192
20193       -mloongson-ext2
20194       -mno-loongson-ext2
20195           Use (do not use) the MIPS Loongson EXTensions r2 (EXT2)
20196           instructions.
20197
20198       -mlong64
20199           Force "long" types to be 64 bits wide.  See -mlong32 for an
20200           explanation of the default and the way that the pointer size is
20201           determined.
20202
20203       -mlong32
20204           Force "long", "int", and pointer types to be 32 bits wide.
20205
20206           The default size of "int"s, "long"s and pointers depends on the
20207           ABI.  All the supported ABIs use 32-bit "int"s.  The n64 ABI uses
20208           64-bit "long"s, as does the 64-bit EABI; the others use 32-bit
20209           "long"s.  Pointers are the same size as "long"s, or the same size
20210           as integer registers, whichever is smaller.
20211
20212       -msym32
20213       -mno-sym32
20214           Assume (do not assume) that all symbols have 32-bit values,
20215           regardless of the selected ABI.  This option is useful in
20216           combination with -mabi=64 and -mno-abicalls because it allows GCC
20217           to generate shorter and faster references to symbolic addresses.
20218
20219       -G num
20220           Put definitions of externally-visible data in a small data section
20221           if that data is no bigger than num bytes.  GCC can then generate
20222           more efficient accesses to the data; see -mgpopt for details.
20223
20224           The default -G option depends on the configuration.
20225
20226       -mlocal-sdata
20227       -mno-local-sdata
20228           Extend (do not extend) the -G behavior to local data too, such as
20229           to static variables in C.  -mlocal-sdata is the default for all
20230           configurations.
20231
20232           If the linker complains that an application is using too much small
20233           data, you might want to try rebuilding the less performance-
20234           critical parts with -mno-local-sdata.  You might also want to build
20235           large libraries with -mno-local-sdata, so that the libraries leave
20236           more room for the main program.
20237
20238       -mextern-sdata
20239       -mno-extern-sdata
20240           Assume (do not assume) that externally-defined data is in a small
20241           data section if the size of that data is within the -G limit.
20242           -mextern-sdata is the default for all configurations.
20243
20244           If you compile a module Mod with -mextern-sdata -G num -mgpopt, and
20245           Mod references a variable Var that is no bigger than num bytes, you
20246           must make sure that Var is placed in a small data section.  If Var
20247           is defined by another module, you must either compile that module
20248           with a high-enough -G setting or attach a "section" attribute to
20249           Var's definition.  If Var is common, you must link the application
20250           with a high-enough -G setting.
20251
20252           The easiest way of satisfying these restrictions is to compile and
20253           link every module with the same -G option.  However, you may wish
20254           to build a library that supports several different small data
20255           limits.  You can do this by compiling the library with the highest
20256           supported -G setting and additionally using -mno-extern-sdata to
20257           stop the library from making assumptions about externally-defined
20258           data.
20259
20260       -mgpopt
20261       -mno-gpopt
20262           Use (do not use) GP-relative accesses for symbols that are known to
20263           be in a small data section; see -G, -mlocal-sdata and
20264           -mextern-sdata.  -mgpopt is the default for all configurations.
20265
20266           -mno-gpopt is useful for cases where the $gp register might not
20267           hold the value of "_gp".  For example, if the code is part of a
20268           library that might be used in a boot monitor, programs that call
20269           boot monitor routines pass an unknown value in $gp.  (In such
20270           situations, the boot monitor itself is usually compiled with -G0.)
20271
20272           -mno-gpopt implies -mno-local-sdata and -mno-extern-sdata.
20273
20274       -membedded-data
20275       -mno-embedded-data
20276           Allocate variables to the read-only data section first if possible,
20277           then next in the small data section if possible, otherwise in data.
20278           This gives slightly slower code than the default, but reduces the
20279           amount of RAM required when executing, and thus may be preferred
20280           for some embedded systems.
20281
20282       -muninit-const-in-rodata
20283       -mno-uninit-const-in-rodata
20284           Put uninitialized "const" variables in the read-only data section.
20285           This option is only meaningful in conjunction with -membedded-data.
20286
20287       -mcode-readable=setting
20288           Specify whether GCC may generate code that reads from executable
20289           sections.  There are three possible settings:
20290
20291           -mcode-readable=yes
20292               Instructions may freely access executable sections.  This is
20293               the default setting.
20294
20295           -mcode-readable=pcrel
20296               MIPS16 PC-relative load instructions can access executable
20297               sections, but other instructions must not do so.  This option
20298               is useful on 4KSc and 4KSd processors when the code TLBs have
20299               the Read Inhibit bit set.  It is also useful on processors that
20300               can be configured to have a dual instruction/data SRAM
20301               interface and that, like the M4K, automatically redirect PC-
20302               relative loads to the instruction RAM.
20303
20304           -mcode-readable=no
20305               Instructions must not access executable sections.  This option
20306               can be useful on targets that are configured to have a dual
20307               instruction/data SRAM interface but that (unlike the M4K) do
20308               not automatically redirect PC-relative loads to the instruction
20309               RAM.
20310
20311       -msplit-addresses
20312       -mno-split-addresses
20313           Enable (disable) use of the "%hi()" and "%lo()" assembler
20314           relocation operators.  This option has been superseded by
20315           -mexplicit-relocs but is retained for backwards compatibility.
20316
20317       -mexplicit-relocs
20318       -mno-explicit-relocs
20319           Use (do not use) assembler relocation operators when dealing with
20320           symbolic addresses.  The alternative, selected by
20321           -mno-explicit-relocs, is to use assembler macros instead.
20322
20323           -mexplicit-relocs is the default if GCC was configured to use an
20324           assembler that supports relocation operators.
20325
20326       -mcheck-zero-division
20327       -mno-check-zero-division
20328           Trap (do not trap) on integer division by zero.
20329
20330           The default is -mcheck-zero-division.
20331
20332       -mdivide-traps
20333       -mdivide-breaks
20334           MIPS systems check for division by zero by generating either a
20335           conditional trap or a break instruction.  Using traps results in
20336           smaller code, but is only supported on MIPS II and later.  Also,
20337           some versions of the Linux kernel have a bug that prevents trap
20338           from generating the proper signal ("SIGFPE").  Use -mdivide-traps
20339           to allow conditional traps on architectures that support them and
20340           -mdivide-breaks to force the use of breaks.
20341
20342           The default is usually -mdivide-traps, but this can be overridden
20343           at configure time using --with-divide=breaks.  Divide-by-zero
20344           checks can be completely disabled using -mno-check-zero-division.
20345
20346       -mload-store-pairs
20347       -mno-load-store-pairs
20348           Enable (disable) an optimization that pairs consecutive load or
20349           store instructions to enable load/store bonding.  This option is
20350           enabled by default but only takes effect when the selected
20351           architecture is known to support bonding.
20352
20353       -mmemcpy
20354       -mno-memcpy
20355           Force (do not force) the use of "memcpy" for non-trivial block
20356           moves.  The default is -mno-memcpy, which allows GCC to inline most
20357           constant-sized copies.
20358
20359       -mlong-calls
20360       -mno-long-calls
20361           Disable (do not disable) use of the "jal" instruction.  Calling
20362           functions using "jal" is more efficient but requires the caller and
20363           callee to be in the same 256 megabyte segment.
20364
20365           This option has no effect on abicalls code.  The default is
20366           -mno-long-calls.
20367
20368       -mmad
20369       -mno-mad
20370           Enable (disable) use of the "mad", "madu" and "mul" instructions,
20371           as provided by the R4650 ISA.
20372
20373       -mimadd
20374       -mno-imadd
20375           Enable (disable) use of the "madd" and "msub" integer instructions.
20376           The default is -mimadd on architectures that support "madd" and
20377           "msub" except for the 74k architecture where it was found to
20378           generate slower code.
20379
20380       -mfused-madd
20381       -mno-fused-madd
20382           Enable (disable) use of the floating-point multiply-accumulate
20383           instructions, when they are available.  The default is
20384           -mfused-madd.
20385
20386           On the R8000 CPU when multiply-accumulate instructions are used,
20387           the intermediate product is calculated to infinite precision and is
20388           not subject to the FCSR Flush to Zero bit.  This may be undesirable
20389           in some circumstances.  On other processors the result is
20390           numerically identical to the equivalent computation using separate
20391           multiply, add, subtract and negate instructions.
20392
20393       -nocpp
20394           Tell the MIPS assembler to not run its preprocessor over user
20395           assembler files (with a .s suffix) when assembling them.
20396
20397       -mfix-24k
20398       -mno-fix-24k
20399           Work around the 24K E48 (lost data on stores during refill) errata.
20400           The workarounds are implemented by the assembler rather than by
20401           GCC.
20402
20403       -mfix-r4000
20404       -mno-fix-r4000
20405           Work around certain R4000 CPU errata:
20406
20407           -   A double-word or a variable shift may give an incorrect result
20408               if executed immediately after starting an integer division.
20409
20410           -   A double-word or a variable shift may give an incorrect result
20411               if executed while an integer multiplication is in progress.
20412
20413           -   An integer division may give an incorrect result if started in
20414               a delay slot of a taken branch or a jump.
20415
20416       -mfix-r4400
20417       -mno-fix-r4400
20418           Work around certain R4400 CPU errata:
20419
20420           -   A double-word or a variable shift may give an incorrect result
20421               if executed immediately after starting an integer division.
20422
20423       -mfix-r10000
20424       -mno-fix-r10000
20425           Work around certain R10000 errata:
20426
20427           -   "ll"/"sc" sequences may not behave atomically on revisions
20428               prior to 3.0.  They may deadlock on revisions 2.6 and earlier.
20429
20430           This option can only be used if the target architecture supports
20431           branch-likely instructions.  -mfix-r10000 is the default when
20432           -march=r10000 is used; -mno-fix-r10000 is the default otherwise.
20433
20434       -mfix-r5900
20435       -mno-fix-r5900
20436           Do not attempt to schedule the preceding instruction into the delay
20437           slot of a branch instruction placed at the end of a short loop of
20438           six instructions or fewer and always schedule a "nop" instruction
20439           there instead.  The short loop bug under certain conditions causes
20440           loops to execute only once or twice, due to a hardware bug in the
20441           R5900 chip.  The workaround is implemented by the assembler rather
20442           than by GCC.
20443
20444       -mfix-rm7000
20445       -mno-fix-rm7000
20446           Work around the RM7000 "dmult"/"dmultu" errata.  The workarounds
20447           are implemented by the assembler rather than by GCC.
20448
20449       -mfix-vr4120
20450       -mno-fix-vr4120
20451           Work around certain VR4120 errata:
20452
20453           -   "dmultu" does not always produce the correct result.
20454
20455           -   "div" and "ddiv" do not always produce the correct result if
20456               one of the operands is negative.
20457
20458           The workarounds for the division errata rely on special functions
20459           in libgcc.a.  At present, these functions are only provided by the
20460           "mips64vr*-elf" configurations.
20461
20462           Other VR4120 errata require a NOP to be inserted between certain
20463           pairs of instructions.  These errata are handled by the assembler,
20464           not by GCC itself.
20465
20466       -mfix-vr4130
20467           Work around the VR4130 "mflo"/"mfhi" errata.  The workarounds are
20468           implemented by the assembler rather than by GCC, although GCC
20469           avoids using "mflo" and "mfhi" if the VR4130 "macc", "macchi",
20470           "dmacc" and "dmacchi" instructions are available instead.
20471
20472       -mfix-sb1
20473       -mno-fix-sb1
20474           Work around certain SB-1 CPU core errata.  (This flag currently
20475           works around the SB-1 revision 2 "F1" and "F2" floating-point
20476           errata.)
20477
20478       -mr10k-cache-barrier=setting
20479           Specify whether GCC should insert cache barriers to avoid the side
20480           effects of speculation on R10K processors.
20481
20482           In common with many processors, the R10K tries to predict the
20483           outcome of a conditional branch and speculatively executes
20484           instructions from the "taken" branch.  It later aborts these
20485           instructions if the predicted outcome is wrong.  However, on the
20486           R10K, even aborted instructions can have side effects.
20487
20488           This problem only affects kernel stores and, depending on the
20489           system, kernel loads.  As an example, a speculatively-executed
20490           store may load the target memory into cache and mark the cache line
20491           as dirty, even if the store itself is later aborted.  If a DMA
20492           operation writes to the same area of memory before the "dirty" line
20493           is flushed, the cached data overwrites the DMA-ed data.  See the
20494           R10K processor manual for a full description, including other
20495           potential problems.
20496
20497           One workaround is to insert cache barrier instructions before every
20498           memory access that might be speculatively executed and that might
20499           have side effects even if aborted.  -mr10k-cache-barrier=setting
20500           controls GCC's implementation of this workaround.  It assumes that
20501           aborted accesses to any byte in the following regions does not have
20502           side effects:
20503
20504           1.  the memory occupied by the current function's stack frame;
20505
20506           2.  the memory occupied by an incoming stack argument;
20507
20508           3.  the memory occupied by an object with a link-time-constant
20509               address.
20510
20511           It is the kernel's responsibility to ensure that speculative
20512           accesses to these regions are indeed safe.
20513
20514           If the input program contains a function declaration such as:
20515
20516                   void foo (void);
20517
20518           then the implementation of "foo" must allow "j foo" and "jal foo"
20519           to be executed speculatively.  GCC honors this restriction for
20520           functions it compiles itself.  It expects non-GCC functions (such
20521           as hand-written assembly code) to do the same.
20522
20523           The option has three forms:
20524
20525           -mr10k-cache-barrier=load-store
20526               Insert a cache barrier before a load or store that might be
20527               speculatively executed and that might have side effects even if
20528               aborted.
20529
20530           -mr10k-cache-barrier=store
20531               Insert a cache barrier before a store that might be
20532               speculatively executed and that might have side effects even if
20533               aborted.
20534
20535           -mr10k-cache-barrier=none
20536               Disable the insertion of cache barriers.  This is the default
20537               setting.
20538
20539       -mflush-func=func
20540       -mno-flush-func
20541           Specifies the function to call to flush the I and D caches, or to
20542           not call any such function.  If called, the function must take the
20543           same arguments as the common "_flush_func", that is, the address of
20544           the memory range for which the cache is being flushed, the size of
20545           the memory range, and the number 3 (to flush both caches).  The
20546           default depends on the target GCC was configured for, but commonly
20547           is either "_flush_func" or "__cpu_flush".
20548
20549       mbranch-cost=num
20550           Set the cost of branches to roughly num "simple" instructions.
20551           This cost is only a heuristic and is not guaranteed to produce
20552           consistent results across releases.  A zero cost redundantly
20553           selects the default, which is based on the -mtune setting.
20554
20555       -mbranch-likely
20556       -mno-branch-likely
20557           Enable or disable use of Branch Likely instructions, regardless of
20558           the default for the selected architecture.  By default, Branch
20559           Likely instructions may be generated if they are supported by the
20560           selected architecture.  An exception is for the MIPS32 and MIPS64
20561           architectures and processors that implement those architectures;
20562           for those, Branch Likely instructions are not be generated by
20563           default because the MIPS32 and MIPS64 architectures specifically
20564           deprecate their use.
20565
20566       -mcompact-branches=never
20567       -mcompact-branches=optimal
20568       -mcompact-branches=always
20569           These options control which form of branches will be generated.
20570           The default is -mcompact-branches=optimal.
20571
20572           The -mcompact-branches=never option ensures that compact branch
20573           instructions will never be generated.
20574
20575           The -mcompact-branches=always option ensures that a compact branch
20576           instruction will be generated if available.  If a compact branch
20577           instruction is not available, a delay slot form of the branch will
20578           be used instead.
20579
20580           This option is supported from MIPS Release 6 onwards.
20581
20582           The -mcompact-branches=optimal option will cause a delay slot
20583           branch to be used if one is available in the current ISA and the
20584           delay slot is successfully filled.  If the delay slot is not
20585           filled, a compact branch will be chosen if one is available.
20586
20587       -mfp-exceptions
20588       -mno-fp-exceptions
20589           Specifies whether FP exceptions are enabled.  This affects how FP
20590           instructions are scheduled for some processors.  The default is
20591           that FP exceptions are enabled.
20592
20593           For instance, on the SB-1, if FP exceptions are disabled, and we
20594           are emitting 64-bit code, then we can use both FP pipes.
20595           Otherwise, we can only use one FP pipe.
20596
20597       -mvr4130-align
20598       -mno-vr4130-align
20599           The VR4130 pipeline is two-way superscalar, but can only issue two
20600           instructions together if the first one is 8-byte aligned.  When
20601           this option is enabled, GCC aligns pairs of instructions that it
20602           thinks should execute in parallel.
20603
20604           This option only has an effect when optimizing for the VR4130.  It
20605           normally makes code faster, but at the expense of making it bigger.
20606           It is enabled by default at optimization level -O3.
20607
20608       -msynci
20609       -mno-synci
20610           Enable (disable) generation of "synci" instructions on
20611           architectures that support it.  The "synci" instructions (if
20612           enabled) are generated when "__builtin___clear_cache" is compiled.
20613
20614           This option defaults to -mno-synci, but the default can be
20615           overridden by configuring GCC with --with-synci.
20616
20617           When compiling code for single processor systems, it is generally
20618           safe to use "synci".  However, on many multi-core (SMP) systems, it
20619           does not invalidate the instruction caches on all cores and may
20620           lead to undefined behavior.
20621
20622       -mrelax-pic-calls
20623       -mno-relax-pic-calls
20624           Try to turn PIC calls that are normally dispatched via register $25
20625           into direct calls.  This is only possible if the linker can resolve
20626           the destination at link time and if the destination is within range
20627           for a direct call.
20628
20629           -mrelax-pic-calls is the default if GCC was configured to use an
20630           assembler and a linker that support the ".reloc" assembly directive
20631           and -mexplicit-relocs is in effect.  With -mno-explicit-relocs,
20632           this optimization can be performed by the assembler and the linker
20633           alone without help from the compiler.
20634
20635       -mmcount-ra-address
20636       -mno-mcount-ra-address
20637           Emit (do not emit) code that allows "_mcount" to modify the calling
20638           function's return address.  When enabled, this option extends the
20639           usual "_mcount" interface with a new ra-address parameter, which
20640           has type "intptr_t *" and is passed in register $12.  "_mcount" can
20641           then modify the return address by doing both of the following:
20642
20643           *   Returning the new address in register $31.
20644
20645           *   Storing the new address in "*ra-address", if ra-address is
20646               nonnull.
20647
20648           The default is -mno-mcount-ra-address.
20649
20650       -mframe-header-opt
20651       -mno-frame-header-opt
20652           Enable (disable) frame header optimization in the o32 ABI.  When
20653           using the o32 ABI, calling functions will allocate 16 bytes on the
20654           stack for the called function to write out register arguments.
20655           When enabled, this optimization will suppress the allocation of the
20656           frame header if it can be determined that it is unused.
20657
20658           This optimization is off by default at all optimization levels.
20659
20660       -mlxc1-sxc1
20661       -mno-lxc1-sxc1
20662           When applicable, enable (disable) the generation of "lwxc1",
20663           "swxc1", "ldxc1", "sdxc1" instructions.  Enabled by default.
20664
20665       -mmadd4
20666       -mno-madd4
20667           When applicable, enable (disable) the generation of 4-operand
20668           "madd.s", "madd.d" and related instructions.  Enabled by default.
20669
20670   MMIX Options
20671       These options are defined for the MMIX:
20672
20673       -mlibfuncs
20674       -mno-libfuncs
20675           Specify that intrinsic library functions are being compiled,
20676           passing all values in registers, no matter the size.
20677
20678       -mepsilon
20679       -mno-epsilon
20680           Generate floating-point comparison instructions that compare with
20681           respect to the "rE" epsilon register.
20682
20683       -mabi=mmixware
20684       -mabi=gnu
20685           Generate code that passes function parameters and return values
20686           that (in the called function) are seen as registers $0 and up, as
20687           opposed to the GNU ABI which uses global registers $231 and up.
20688
20689       -mzero-extend
20690       -mno-zero-extend
20691           When reading data from memory in sizes shorter than 64 bits, use
20692           (do not use) zero-extending load instructions by default, rather
20693           than sign-extending ones.
20694
20695       -mknuthdiv
20696       -mno-knuthdiv
20697           Make the result of a division yielding a remainder have the same
20698           sign as the divisor.  With the default, -mno-knuthdiv, the sign of
20699           the remainder follows the sign of the dividend.  Both methods are
20700           arithmetically valid, the latter being almost exclusively used.
20701
20702       -mtoplevel-symbols
20703       -mno-toplevel-symbols
20704           Prepend (do not prepend) a : to all global symbols, so the assembly
20705           code can be used with the "PREFIX" assembly directive.
20706
20707       -melf
20708           Generate an executable in the ELF format, rather than the default
20709           mmo format used by the mmix simulator.
20710
20711       -mbranch-predict
20712       -mno-branch-predict
20713           Use (do not use) the probable-branch instructions, when static
20714           branch prediction indicates a probable branch.
20715
20716       -mbase-addresses
20717       -mno-base-addresses
20718           Generate (do not generate) code that uses base addresses.  Using a
20719           base address automatically generates a request (handled by the
20720           assembler and the linker) for a constant to be set up in a global
20721           register.  The register is used for one or more base address
20722           requests within the range 0 to 255 from the value held in the
20723           register.  The generally leads to short and fast code, but the
20724           number of different data items that can be addressed is limited.
20725           This means that a program that uses lots of static data may require
20726           -mno-base-addresses.
20727
20728       -msingle-exit
20729       -mno-single-exit
20730           Force (do not force) generated code to have a single exit point in
20731           each function.
20732
20733   MN10300 Options
20734       These -m options are defined for Matsushita MN10300 architectures:
20735
20736       -mmult-bug
20737           Generate code to avoid bugs in the multiply instructions for the
20738           MN10300 processors.  This is the default.
20739
20740       -mno-mult-bug
20741           Do not generate code to avoid bugs in the multiply instructions for
20742           the MN10300 processors.
20743
20744       -mam33
20745           Generate code using features specific to the AM33 processor.
20746
20747       -mno-am33
20748           Do not generate code using features specific to the AM33 processor.
20749           This is the default.
20750
20751       -mam33-2
20752           Generate code using features specific to the AM33/2.0 processor.
20753
20754       -mam34
20755           Generate code using features specific to the AM34 processor.
20756
20757       -mtune=cpu-type
20758           Use the timing characteristics of the indicated CPU type when
20759           scheduling instructions.  This does not change the targeted
20760           processor type.  The CPU type must be one of mn10300, am33, am33-2
20761           or am34.
20762
20763       -mreturn-pointer-on-d0
20764           When generating a function that returns a pointer, return the
20765           pointer in both "a0" and "d0".  Otherwise, the pointer is returned
20766           only in "a0", and attempts to call such functions without a
20767           prototype result in errors.  Note that this option is on by
20768           default; use -mno-return-pointer-on-d0 to disable it.
20769
20770       -mno-crt0
20771           Do not link in the C run-time initialization object file.
20772
20773       -mrelax
20774           Indicate to the linker that it should perform a relaxation
20775           optimization pass to shorten branches, calls and absolute memory
20776           addresses.  This option only has an effect when used on the command
20777           line for the final link step.
20778
20779           This option makes symbolic debugging impossible.
20780
20781       -mliw
20782           Allow the compiler to generate Long Instruction Word instructions
20783           if the target is the AM33 or later.  This is the default.  This
20784           option defines the preprocessor macro "__LIW__".
20785
20786       -mno-liw
20787           Do not allow the compiler to generate Long Instruction Word
20788           instructions.  This option defines the preprocessor macro
20789           "__NO_LIW__".
20790
20791       -msetlb
20792           Allow the compiler to generate the SETLB and Lcc instructions if
20793           the target is the AM33 or later.  This is the default.  This option
20794           defines the preprocessor macro "__SETLB__".
20795
20796       -mno-setlb
20797           Do not allow the compiler to generate SETLB or Lcc instructions.
20798           This option defines the preprocessor macro "__NO_SETLB__".
20799
20800   Moxie Options
20801       -meb
20802           Generate big-endian code.  This is the default for moxie-*-*
20803           configurations.
20804
20805       -mel
20806           Generate little-endian code.
20807
20808       -mmul.x
20809           Generate mul.x and umul.x instructions.  This is the default for
20810           moxiebox-*-* configurations.
20811
20812       -mno-crt0
20813           Do not link in the C run-time initialization object file.
20814
20815   MSP430 Options
20816       These options are defined for the MSP430:
20817
20818       -masm-hex
20819           Force assembly output to always use hex constants.  Normally such
20820           constants are signed decimals, but this option is available for
20821           testsuite and/or aesthetic purposes.
20822
20823       -mmcu=
20824           Select the MCU to target.  This is used to create a C preprocessor
20825           symbol based upon the MCU name, converted to upper case and pre-
20826           and post-fixed with __.  This in turn is used by the msp430.h
20827           header file to select an MCU-specific supplementary header file.
20828
20829           The option also sets the ISA to use.  If the MCU name is one that
20830           is known to only support the 430 ISA then that is selected,
20831           otherwise the 430X ISA is selected.  A generic MCU name of msp430
20832           can also be used to select the 430 ISA.  Similarly the generic
20833           msp430x MCU name selects the 430X ISA.
20834
20835           In addition an MCU-specific linker script is added to the linker
20836           command line.  The script's name is the name of the MCU with .ld
20837           appended.  Thus specifying -mmcu=xxx on the gcc command line
20838           defines the C preprocessor symbol "__XXX__" and cause the linker to
20839           search for a script called xxx.ld.
20840
20841           The ISA and hardware multiply supported for the different MCUs is
20842           hard-coded into GCC.  However, an external devices.csv file can be
20843           used to extend device support beyond those that have been hard-
20844           coded.
20845
20846           GCC searches for the devices.csv file using the following methods
20847           in the given precedence order, where the first method takes
20848           precendence over the second which takes precedence over the third.
20849
20850           Include path specified with "-I" and "-L"
20851               devices.csv will be searched for in each of the directories
20852               specified by include paths and linker library search paths.
20853
20854           Path specified by the environment variable MSP430_GCC_INCLUDE_DIR
20855               Define the value of the global environment variable
20856               MSP430_GCC_INCLUDE_DIR to the full path to the directory
20857               containing devices.csv, and GCC will search this directory for
20858               devices.csv.  If devices.csv is found, this directory will also
20859               be registered as an include path, and linker library path.
20860               Header files and linker scripts in this directory can therefore
20861               be used without manually specifying "-I" and "-L" on the
20862               command line.
20863
20864           The msp430-elf{,bare}/include/devices directory
20865               Finally, GCC will examine msp430-elf{,bare}/include/devices
20866               from the toolchain root directory.  This directory does not
20867               exist in a default installation, but if the user has created it
20868               and copied devices.csv there, then the MCU data will be read.
20869               As above, this directory will also be registered as an include
20870               path, and linker library path.
20871
20872           If none of the above search methods find devices.csv, then the
20873           hard-coded MCU data is used.
20874
20875       -mwarn-mcu
20876       -mno-warn-mcu
20877           This option enables or disables warnings about conflicts between
20878           the MCU name specified by the -mmcu option and the ISA set by the
20879           -mcpu option and/or the hardware multiply support set by the
20880           -mhwmult option.  It also toggles warnings about unrecognized MCU
20881           names.  This option is on by default.
20882
20883       -mcpu=
20884           Specifies the ISA to use.  Accepted values are msp430, msp430x and
20885           msp430xv2.  This option is deprecated.  The -mmcu= option should be
20886           used to select the ISA.
20887
20888       -msim
20889           Link to the simulator runtime libraries and linker script.
20890           Overrides any scripts that would be selected by the -mmcu= option.
20891
20892       -mlarge
20893           Use large-model addressing (20-bit pointers, 32-bit "size_t").
20894
20895       -msmall
20896           Use small-model addressing (16-bit pointers, 16-bit "size_t").
20897
20898       -mrelax
20899           This option is passed to the assembler and linker, and allows the
20900           linker to perform certain optimizations that cannot be done until
20901           the final link.
20902
20903       mhwmult=
20904           Describes the type of hardware multiply supported by the target.
20905           Accepted values are none for no hardware multiply, 16bit for the
20906           original 16-bit-only multiply supported by early MCUs.  32bit for
20907           the 16/32-bit multiply supported by later MCUs and f5series for the
20908           16/32-bit multiply supported by F5-series MCUs.  A value of auto
20909           can also be given.  This tells GCC to deduce the hardware multiply
20910           support based upon the MCU name provided by the -mmcu option.  If
20911           no -mmcu option is specified or if the MCU name is not recognized
20912           then no hardware multiply support is assumed.  "auto" is the
20913           default setting.
20914
20915           Hardware multiplies are normally performed by calling a library
20916           routine.  This saves space in the generated code.  When compiling
20917           at -O3 or higher however the hardware multiplier is invoked inline.
20918           This makes for bigger, but faster code.
20919
20920           The hardware multiply routines disable interrupts whilst running
20921           and restore the previous interrupt state when they finish.  This
20922           makes them safe to use inside interrupt handlers as well as in
20923           normal code.
20924
20925       -minrt
20926           Enable the use of a minimum runtime environment - no static
20927           initializers or constructors.  This is intended for memory-
20928           constrained devices.  The compiler includes special symbols in some
20929           objects that tell the linker and runtime which code fragments are
20930           required.
20931
20932       -mtiny-printf
20933           Enable reduced code size "printf" and "puts" library functions.
20934           The tiny implementations of these functions are not reentrant, so
20935           must be used with caution in multi-threaded applications.
20936
20937           Support for streams has been removed and the string to be printed
20938           will always be sent to stdout via the "write" syscall.  The string
20939           is not buffered before it is sent to write.
20940
20941           This option requires Newlib Nano IO, so GCC must be configured with
20942           --enable-newlib-nano-formatted-io.
20943
20944       -mcode-region=
20945       -mdata-region=
20946           These options tell the compiler where to place functions and data
20947           that do not have one of the "lower", "upper", "either" or "section"
20948           attributes.  Possible values are "lower", "upper", "either" or
20949           "any".  The first three behave like the corresponding attribute.
20950           The fourth possible value - "any" - is the default.  It leaves
20951           placement entirely up to the linker script and how it assigns the
20952           standard sections (".text", ".data", etc) to the memory regions.
20953
20954       -msilicon-errata=
20955           This option passes on a request to assembler to enable the fixes
20956           for the named silicon errata.
20957
20958       -msilicon-errata-warn=
20959           This option passes on a request to the assembler to enable warning
20960           messages when a silicon errata might need to be applied.
20961
20962       -mwarn-devices-csv
20963       -mno-warn-devices-csv
20964           Warn if devices.csv is not found or there are problem parsing it
20965           (default: on).
20966
20967   NDS32 Options
20968       These options are defined for NDS32 implementations:
20969
20970       -mbig-endian
20971           Generate code in big-endian mode.
20972
20973       -mlittle-endian
20974           Generate code in little-endian mode.
20975
20976       -mreduced-regs
20977           Use reduced-set registers for register allocation.
20978
20979       -mfull-regs
20980           Use full-set registers for register allocation.
20981
20982       -mcmov
20983           Generate conditional move instructions.
20984
20985       -mno-cmov
20986           Do not generate conditional move instructions.
20987
20988       -mext-perf
20989           Generate performance extension instructions.
20990
20991       -mno-ext-perf
20992           Do not generate performance extension instructions.
20993
20994       -mext-perf2
20995           Generate performance extension 2 instructions.
20996
20997       -mno-ext-perf2
20998           Do not generate performance extension 2 instructions.
20999
21000       -mext-string
21001           Generate string extension instructions.
21002
21003       -mno-ext-string
21004           Do not generate string extension instructions.
21005
21006       -mv3push
21007           Generate v3 push25/pop25 instructions.
21008
21009       -mno-v3push
21010           Do not generate v3 push25/pop25 instructions.
21011
21012       -m16-bit
21013           Generate 16-bit instructions.
21014
21015       -mno-16-bit
21016           Do not generate 16-bit instructions.
21017
21018       -misr-vector-size=num
21019           Specify the size of each interrupt vector, which must be 4 or 16.
21020
21021       -mcache-block-size=num
21022           Specify the size of each cache block, which must be a power of 2
21023           between 4 and 512.
21024
21025       -march=arch
21026           Specify the name of the target architecture.
21027
21028       -mcmodel=code-model
21029           Set the code model to one of
21030
21031           small
21032               All the data and read-only data segments must be within 512KB
21033               addressing space.  The text segment must be within 16MB
21034               addressing space.
21035
21036           medium
21037               The data segment must be within 512KB while the read-only data
21038               segment can be within 4GB addressing space.  The text segment
21039               should be still within 16MB addressing space.
21040
21041           large
21042               All the text and data segments can be within 4GB addressing
21043               space.
21044
21045       -mctor-dtor
21046           Enable constructor/destructor feature.
21047
21048       -mrelax
21049           Guide linker to relax instructions.
21050
21051   Nios II Options
21052       These are the options defined for the Altera Nios II processor.
21053
21054       -G num
21055           Put global and static objects less than or equal to num bytes into
21056           the small data or BSS sections instead of the normal data or BSS
21057           sections.  The default value of num is 8.
21058
21059       -mgpopt=option
21060       -mgpopt
21061       -mno-gpopt
21062           Generate (do not generate) GP-relative accesses.  The following
21063           option names are recognized:
21064
21065           none
21066               Do not generate GP-relative accesses.
21067
21068           local
21069               Generate GP-relative accesses for small data objects that are
21070               not external, weak, or uninitialized common symbols.  Also use
21071               GP-relative addressing for objects that have been explicitly
21072               placed in a small data section via a "section" attribute.
21073
21074           global
21075               As for local, but also generate GP-relative accesses for small
21076               data objects that are external, weak, or common.  If you use
21077               this option, you must ensure that all parts of your program
21078               (including libraries) are compiled with the same -G setting.
21079
21080           data
21081               Generate GP-relative accesses for all data objects in the
21082               program.  If you use this option, the entire data and BSS
21083               segments of your program must fit in 64K of memory and you must
21084               use an appropriate linker script to allocate them within the
21085               addressable range of the global pointer.
21086
21087           all Generate GP-relative addresses for function pointers as well as
21088               data pointers.  If you use this option, the entire text, data,
21089               and BSS segments of your program must fit in 64K of memory and
21090               you must use an appropriate linker script to allocate them
21091               within the addressable range of the global pointer.
21092
21093           -mgpopt is equivalent to -mgpopt=local, and -mno-gpopt is
21094           equivalent to -mgpopt=none.
21095
21096           The default is -mgpopt except when -fpic or -fPIC is specified to
21097           generate position-independent code.  Note that the Nios II ABI does
21098           not permit GP-relative accesses from shared libraries.
21099
21100           You may need to specify -mno-gpopt explicitly when building
21101           programs that include large amounts of small data, including large
21102           GOT data sections.  In this case, the 16-bit offset for GP-relative
21103           addressing may not be large enough to allow access to the entire
21104           small data section.
21105
21106       -mgprel-sec=regexp
21107           This option specifies additional section names that can be accessed
21108           via GP-relative addressing.  It is most useful in conjunction with
21109           "section" attributes on variable declarations and a custom linker
21110           script.  The regexp is a POSIX Extended Regular Expression.
21111
21112           This option does not affect the behavior of the -G option, and the
21113           specified sections are in addition to the standard ".sdata" and
21114           ".sbss" small-data sections that are recognized by -mgpopt.
21115
21116       -mr0rel-sec=regexp
21117           This option specifies names of sections that can be accessed via a
21118           16-bit offset from "r0"; that is, in the low 32K or high 32K of the
21119           32-bit address space.  It is most useful in conjunction with
21120           "section" attributes on variable declarations and a custom linker
21121           script.  The regexp is a POSIX Extended Regular Expression.
21122
21123           In contrast to the use of GP-relative addressing for small data,
21124           zero-based addressing is never generated by default and there are
21125           no conventional section names used in standard linker scripts for
21126           sections in the low or high areas of memory.
21127
21128       -mel
21129       -meb
21130           Generate little-endian (default) or big-endian (experimental) code,
21131           respectively.
21132
21133       -march=arch
21134           This specifies the name of the target Nios II architecture.  GCC
21135           uses this name to determine what kind of instructions it can emit
21136           when generating assembly code.  Permissible names are: r1, r2.
21137
21138           The preprocessor macro "__nios2_arch__" is available to programs,
21139           with value 1 or 2, indicating the targeted ISA level.
21140
21141       -mbypass-cache
21142       -mno-bypass-cache
21143           Force all load and store instructions to always bypass cache by
21144           using I/O variants of the instructions. The default is not to
21145           bypass the cache.
21146
21147       -mno-cache-volatile
21148       -mcache-volatile
21149           Volatile memory access bypass the cache using the I/O variants of
21150           the load and store instructions. The default is not to bypass the
21151           cache.
21152
21153       -mno-fast-sw-div
21154       -mfast-sw-div
21155           Do not use table-based fast divide for small numbers. The default
21156           is to use the fast divide at -O3 and above.
21157
21158       -mno-hw-mul
21159       -mhw-mul
21160       -mno-hw-mulx
21161       -mhw-mulx
21162       -mno-hw-div
21163       -mhw-div
21164           Enable or disable emitting "mul", "mulx" and "div" family of
21165           instructions by the compiler. The default is to emit "mul" and not
21166           emit "div" and "mulx".
21167
21168       -mbmx
21169       -mno-bmx
21170       -mcdx
21171       -mno-cdx
21172           Enable or disable generation of Nios II R2 BMX (bit manipulation)
21173           and CDX (code density) instructions.  Enabling these instructions
21174           also requires -march=r2.  Since these instructions are optional
21175           extensions to the R2 architecture, the default is not to emit them.
21176
21177       -mcustom-insn=N
21178       -mno-custom-insn
21179           Each -mcustom-insn=N option enables use of a custom instruction
21180           with encoding N when generating code that uses insn.  For example,
21181           -mcustom-fadds=253 generates custom instruction 253 for single-
21182           precision floating-point add operations instead of the default
21183           behavior of using a library call.
21184
21185           The following values of insn are supported.  Except as otherwise
21186           noted, floating-point operations are expected to be implemented
21187           with normal IEEE 754 semantics and correspond directly to the C
21188           operators or the equivalent GCC built-in functions.
21189
21190           Single-precision floating point:
21191
21192           fadds, fsubs, fdivs, fmuls
21193               Binary arithmetic operations.
21194
21195           fnegs
21196               Unary negation.
21197
21198           fabss
21199               Unary absolute value.
21200
21201           fcmpeqs, fcmpges, fcmpgts, fcmples, fcmplts, fcmpnes
21202               Comparison operations.
21203
21204           fmins, fmaxs
21205               Floating-point minimum and maximum.  These instructions are
21206               only generated if -ffinite-math-only is specified.
21207
21208           fsqrts
21209               Unary square root operation.
21210
21211           fcoss, fsins, ftans, fatans, fexps, flogs
21212               Floating-point trigonometric and exponential functions.  These
21213               instructions are only generated if -funsafe-math-optimizations
21214               is also specified.
21215
21216           Double-precision floating point:
21217
21218           faddd, fsubd, fdivd, fmuld
21219               Binary arithmetic operations.
21220
21221           fnegd
21222               Unary negation.
21223
21224           fabsd
21225               Unary absolute value.
21226
21227           fcmpeqd, fcmpged, fcmpgtd, fcmpled, fcmpltd, fcmpned
21228               Comparison operations.
21229
21230           fmind, fmaxd
21231               Double-precision minimum and maximum.  These instructions are
21232               only generated if -ffinite-math-only is specified.
21233
21234           fsqrtd
21235               Unary square root operation.
21236
21237           fcosd, fsind, ftand, fatand, fexpd, flogd
21238               Double-precision trigonometric and exponential functions.
21239               These instructions are only generated if
21240               -funsafe-math-optimizations is also specified.
21241
21242           Conversions:
21243
21244           fextsd
21245               Conversion from single precision to double precision.
21246
21247           ftruncds
21248               Conversion from double precision to single precision.
21249
21250           fixsi, fixsu, fixdi, fixdu
21251               Conversion from floating point to signed or unsigned integer
21252               types, with truncation towards zero.
21253
21254           round
21255               Conversion from single-precision floating point to signed
21256               integer, rounding to the nearest integer and ties away from
21257               zero.  This corresponds to the "__builtin_lroundf" function
21258               when -fno-math-errno is used.
21259
21260           floatis, floatus, floatid, floatud
21261               Conversion from signed or unsigned integer types to floating-
21262               point types.
21263
21264           In addition, all of the following transfer instructions for
21265           internal registers X and Y must be provided to use any of the
21266           double-precision floating-point instructions.  Custom instructions
21267           taking two double-precision source operands expect the first
21268           operand in the 64-bit register X.  The other operand (or only
21269           operand of a unary operation) is given to the custom arithmetic
21270           instruction with the least significant half in source register src1
21271           and the most significant half in src2.  A custom instruction that
21272           returns a double-precision result returns the most significant 32
21273           bits in the destination register and the other half in 32-bit
21274           register Y.  GCC automatically generates the necessary code
21275           sequences to write register X and/or read register Y when double-
21276           precision floating-point instructions are used.
21277
21278           fwrx
21279               Write src1 into the least significant half of X and src2 into
21280               the most significant half of X.
21281
21282           fwry
21283               Write src1 into Y.
21284
21285           frdxhi, frdxlo
21286               Read the most or least (respectively) significant half of X and
21287               store it in dest.
21288
21289           frdy
21290               Read the value of Y and store it into dest.
21291
21292           Note that you can gain more local control over generation of Nios
21293           II custom instructions by using the "target("custom-insn=N")" and
21294           "target("no-custom-insn")" function attributes or pragmas.
21295
21296       -mcustom-fpu-cfg=name
21297           This option enables a predefined, named set of custom instruction
21298           encodings (see -mcustom-insn above).  Currently, the following sets
21299           are defined:
21300
21301           -mcustom-fpu-cfg=60-1 is equivalent to: -mcustom-fmuls=252
21302           -mcustom-fadds=253 -mcustom-fsubs=254 -fsingle-precision-constant
21303
21304           -mcustom-fpu-cfg=60-2 is equivalent to: -mcustom-fmuls=252
21305           -mcustom-fadds=253 -mcustom-fsubs=254 -mcustom-fdivs=255
21306           -fsingle-precision-constant
21307
21308           -mcustom-fpu-cfg=72-3 is equivalent to: -mcustom-floatus=243
21309           -mcustom-fixsi=244 -mcustom-floatis=245 -mcustom-fcmpgts=246
21310           -mcustom-fcmples=249 -mcustom-fcmpeqs=250 -mcustom-fcmpnes=251
21311           -mcustom-fmuls=252 -mcustom-fadds=253 -mcustom-fsubs=254
21312           -mcustom-fdivs=255 -fsingle-precision-constant
21313
21314           Custom instruction assignments given by individual -mcustom-insn=
21315           options override those given by -mcustom-fpu-cfg=, regardless of
21316           the order of the options on the command line.
21317
21318           Note that you can gain more local control over selection of a FPU
21319           configuration by using the "target("custom-fpu-cfg=name")" function
21320           attribute or pragma.
21321
21322       These additional -m options are available for the Altera Nios II ELF
21323       (bare-metal) target:
21324
21325       -mhal
21326           Link with HAL BSP.  This suppresses linking with the GCC-provided C
21327           runtime startup and termination code, and is typically used in
21328           conjunction with -msys-crt0= to specify the location of the
21329           alternate startup code provided by the HAL BSP.
21330
21331       -msmallc
21332           Link with a limited version of the C library, -lsmallc, rather than
21333           Newlib.
21334
21335       -msys-crt0=startfile
21336           startfile is the file name of the startfile (crt0) to use when
21337           linking.  This option is only useful in conjunction with -mhal.
21338
21339       -msys-lib=systemlib
21340           systemlib is the library name of the library that provides low-
21341           level system calls required by the C library, e.g. "read" and
21342           "write".  This option is typically used to link with a library
21343           provided by a HAL BSP.
21344
21345   Nvidia PTX Options
21346       These options are defined for Nvidia PTX:
21347
21348       -m32
21349       -m64
21350           Generate code for 32-bit or 64-bit ABI.
21351
21352       -misa=ISA-string
21353           Generate code for given the specified PTX ISA (e.g. sm_35).  ISA
21354           strings must be lower-case.  Valid ISA strings include sm_30 and
21355           sm_35.  The default ISA is sm_30.
21356
21357       -mmainkernel
21358           Link in code for a __main kernel.  This is for stand-alone instead
21359           of offloading execution.
21360
21361       -moptimize
21362           Apply partitioned execution optimizations.  This is the default
21363           when any level of optimization is selected.
21364
21365       -msoft-stack
21366           Generate code that does not use ".local" memory directly for stack
21367           storage. Instead, a per-warp stack pointer is maintained
21368           explicitly. This enables variable-length stack allocation (with
21369           variable-length arrays or "alloca"), and when global memory is used
21370           for underlying storage, makes it possible to access automatic
21371           variables from other threads, or with atomic instructions. This
21372           code generation variant is used for OpenMP offloading, but the
21373           option is exposed on its own for the purpose of testing the
21374           compiler; to generate code suitable for linking into programs using
21375           OpenMP offloading, use option -mgomp.
21376
21377       -muniform-simt
21378           Switch to code generation variant that allows to execute all
21379           threads in each warp, while maintaining memory state and side
21380           effects as if only one thread in each warp was active outside of
21381           OpenMP SIMD regions.  All atomic operations and calls to runtime
21382           (malloc, free, vprintf) are conditionally executed (iff current
21383           lane index equals the master lane index), and the register being
21384           assigned is copied via a shuffle instruction from the master lane.
21385           Outside of SIMD regions lane 0 is the master; inside, each thread
21386           sees itself as the master.  Shared memory array "int __nvptx_uni[]"
21387           stores all-zeros or all-ones bitmasks for each warp, indicating
21388           current mode (0 outside of SIMD regions).  Each thread can bitwise-
21389           and the bitmask at position "tid.y" with current lane index to
21390           compute the master lane index.
21391
21392       -mgomp
21393           Generate code for use in OpenMP offloading: enables -msoft-stack
21394           and -muniform-simt options, and selects corresponding multilib
21395           variant.
21396
21397   OpenRISC Options
21398       These options are defined for OpenRISC:
21399
21400       -mboard=name
21401           Configure a board specific runtime.  This will be passed to the
21402           linker for newlib board library linking.  The default is "or1ksim".
21403
21404       -mnewlib
21405           This option is ignored; it is for compatibility purposes only.
21406           This used to select linker and preprocessor options for use with
21407           newlib.
21408
21409       -msoft-div
21410       -mhard-div
21411           Select software or hardware divide ("l.div", "l.divu")
21412           instructions.  This default is hardware divide.
21413
21414       -msoft-mul
21415       -mhard-mul
21416           Select software or hardware multiply ("l.mul", "l.muli")
21417           instructions.  This default is hardware multiply.
21418
21419       -msoft-float
21420       -mhard-float
21421           Select software or hardware for floating point operations.  The
21422           default is software.
21423
21424       -mdouble-float
21425           When -mhard-float is selected, enables generation of double-
21426           precision floating point instructions.  By default functions from
21427           libgcc are used to perform double-precision floating point
21428           operations.
21429
21430       -munordered-float
21431           When -mhard-float is selected, enables generation of unordered
21432           floating point compare and set flag ("lf.sfun*") instructions.  By
21433           default functions from libgcc are used to perform unordered
21434           floating point compare and set flag operations.
21435
21436       -mcmov
21437           Enable generation of conditional move ("l.cmov") instructions.  By
21438           default the equivalent will be generated using set and branch.
21439
21440       -mror
21441           Enable generation of rotate right ("l.ror") instructions.  By
21442           default functions from libgcc are used to perform rotate right
21443           operations.
21444
21445       -mrori
21446           Enable generation of rotate right with immediate ("l.rori")
21447           instructions.  By default functions from libgcc are used to perform
21448           rotate right with immediate operations.
21449
21450       -msext
21451           Enable generation of sign extension ("l.ext*") instructions.  By
21452           default memory loads are used to perform sign extension.
21453
21454       -msfimm
21455           Enable generation of compare and set flag with immediate ("l.sf*i")
21456           instructions.  By default extra instructions will be generated to
21457           store the immediate to a register first.
21458
21459       -mshftimm
21460           Enable generation of shift with immediate ("l.srai", "l.srli",
21461           "l.slli") instructions.  By default extra instructions will be
21462           generated to store the immediate to a register first.
21463
21464   PDP-11 Options
21465       These options are defined for the PDP-11:
21466
21467       -mfpu
21468           Use hardware FPP floating point.  This is the default.  (FIS
21469           floating point on the PDP-11/40 is not supported.)  Implies -m45.
21470
21471       -msoft-float
21472           Do not use hardware floating point.
21473
21474       -mac0
21475           Return floating-point results in ac0 (fr0 in Unix assembler
21476           syntax).
21477
21478       -mno-ac0
21479           Return floating-point results in memory.  This is the default.
21480
21481       -m40
21482           Generate code for a PDP-11/40.  Implies -msoft-float -mno-split.
21483
21484       -m45
21485           Generate code for a PDP-11/45.  This is the default.
21486
21487       -m10
21488           Generate code for a PDP-11/10.  Implies -msoft-float -mno-split.
21489
21490       -mint16
21491       -mno-int32
21492           Use 16-bit "int".  This is the default.
21493
21494       -mint32
21495       -mno-int16
21496           Use 32-bit "int".
21497
21498       -msplit
21499           Target has split instruction and data space.  Implies -m45.
21500
21501       -munix-asm
21502           Use Unix assembler syntax.
21503
21504       -mdec-asm
21505           Use DEC assembler syntax.
21506
21507       -mgnu-asm
21508           Use GNU assembler syntax.  This is the default.
21509
21510       -mlra
21511           Use the new LRA register allocator.  By default, the old "reload"
21512           allocator is used.
21513
21514   picoChip Options
21515       These -m options are defined for picoChip implementations:
21516
21517       -mae=ae_type
21518           Set the instruction set, register set, and instruction scheduling
21519           parameters for array element type ae_type.  Supported values for
21520           ae_type are ANY, MUL, and MAC.
21521
21522           -mae=ANY selects a completely generic AE type.  Code generated with
21523           this option runs on any of the other AE types.  The code is not as
21524           efficient as it would be if compiled for a specific AE type, and
21525           some types of operation (e.g., multiplication) do not work properly
21526           on all types of AE.
21527
21528           -mae=MUL selects a MUL AE type.  This is the most useful AE type
21529           for compiled code, and is the default.
21530
21531           -mae=MAC selects a DSP-style MAC AE.  Code compiled with this
21532           option may suffer from poor performance of byte (char)
21533           manipulation, since the DSP AE does not provide hardware support
21534           for byte load/stores.
21535
21536       -msymbol-as-address
21537           Enable the compiler to directly use a symbol name as an address in
21538           a load/store instruction, without first loading it into a register.
21539           Typically, the use of this option generates larger programs, which
21540           run faster than when the option isn't used.  However, the results
21541           vary from program to program, so it is left as a user option,
21542           rather than being permanently enabled.
21543
21544       -mno-inefficient-warnings
21545           Disables warnings about the generation of inefficient code.  These
21546           warnings can be generated, for example, when compiling code that
21547           performs byte-level memory operations on the MAC AE type.  The MAC
21548           AE has no hardware support for byte-level memory operations, so all
21549           byte load/stores must be synthesized from word load/store
21550           operations.  This is inefficient and a warning is generated to
21551           indicate that you should rewrite the code to avoid byte operations,
21552           or to target an AE type that has the necessary hardware support.
21553           This option disables these warnings.
21554
21555   PowerPC Options
21556       These are listed under
21557
21558   PRU Options
21559       These command-line options are defined for PRU target:
21560
21561       -minrt
21562           Link with a minimum runtime environment, with no support for static
21563           initializers and constructors.  Using this option can significantly
21564           reduce the size of the final ELF binary.  Beware that the compiler
21565           could still generate code with static initializers and
21566           constructors.  It is up to the programmer to ensure that the source
21567           program will not use those features.
21568
21569       -mmcu=mcu
21570           Specify the PRU MCU variant to use.  Check Newlib for the exact
21571           list of supported MCUs.
21572
21573       -mno-relax
21574           Make GCC pass the --no-relax command-line option to the linker
21575           instead of the --relax option.
21576
21577       -mloop
21578           Allow (or do not allow) GCC to use the LOOP instruction.
21579
21580       -mabi=variant
21581           Specify the ABI variant to output code for.  -mabi=ti selects the
21582           unmodified TI ABI while -mabi=gnu selects a GNU variant that copes
21583           more naturally with certain GCC assumptions.  These are the
21584           differences:
21585
21586           Function Pointer Size
21587               TI ABI specifies that function (code) pointers are 16-bit,
21588               whereas GNU supports only 32-bit data and code pointers.
21589
21590           Optional Return Value Pointer
21591               Function return values larger than 64 bits are passed by using
21592               a hidden pointer as the first argument of the function.  TI
21593               ABI, though, mandates that the pointer can be NULL in case the
21594               caller is not using the returned value.  GNU always passes and
21595               expects a valid return value pointer.
21596
21597           The current -mabi=ti implementation simply raises a compile error
21598           when any of the above code constructs is detected.  As a
21599           consequence the standard C library cannot be built and it is
21600           omitted when linking with -mabi=ti.
21601
21602           Relaxation is a GNU feature and for safety reasons is disabled when
21603           using -mabi=ti.  The TI toolchain does not emit relocations for
21604           QBBx instructions, so the GNU linker cannot adjust them when
21605           shortening adjacent LDI32 pseudo instructions.
21606
21607   RISC-V Options
21608       These command-line options are defined for RISC-V targets:
21609
21610       -mbranch-cost=n
21611           Set the cost of branches to roughly n instructions.
21612
21613       -mplt
21614       -mno-plt
21615           When generating PIC code, do or don't allow the use of PLTs.
21616           Ignored for non-PIC.  The default is -mplt.
21617
21618       -mabi=ABI-string
21619           Specify integer and floating-point calling convention.  ABI-string
21620           contains two parts: the size of integer types and the registers
21621           used for floating-point types.  For example -march=rv64ifd
21622           -mabi=lp64d means that long and pointers are 64-bit (implicitly
21623           defining int to be 32-bit), and that floating-point values up to 64
21624           bits wide are passed in F registers.  Contrast this with
21625           -march=rv64ifd -mabi=lp64f, which still allows the compiler to
21626           generate code that uses the F and D extensions but only allows
21627           floating-point values up to 32 bits long to be passed in registers;
21628           or -march=rv64ifd -mabi=lp64, in which no floating-point arguments
21629           will be passed in registers.
21630
21631           The default for this argument is system dependent, users who want a
21632           specific calling convention should specify one explicitly.  The
21633           valid calling conventions are: ilp32, ilp32f, ilp32d, lp64, lp64f,
21634           and lp64d.  Some calling conventions are impossible to implement on
21635           some ISAs: for example, -march=rv32if -mabi=ilp32d is invalid
21636           because the ABI requires 64-bit values be passed in F registers,
21637           but F registers are only 32 bits wide.  There is also the ilp32e
21638           ABI that can only be used with the rv32e architecture.  This ABI is
21639           not well specified at present, and is subject to change.
21640
21641       -mfdiv
21642       -mno-fdiv
21643           Do or don't use hardware floating-point divide and square root
21644           instructions.  This requires the F or D extensions for floating-
21645           point registers.  The default is to use them if the specified
21646           architecture has these instructions.
21647
21648       -mdiv
21649       -mno-div
21650           Do or don't use hardware instructions for integer division.  This
21651           requires the M extension.  The default is to use them if the
21652           specified architecture has these instructions.
21653
21654       -march=ISA-string
21655           Generate code for given RISC-V ISA (e.g. rv64im).  ISA strings must
21656           be lower-case.  Examples include rv64i, rv32g, rv32e, and rv32imaf.
21657
21658       -mtune=processor-string
21659           Optimize the output for the given processor, specified by
21660           microarchitecture name.  Permissible values for this option are:
21661           rocket, sifive-3-series, sifive-5-series, sifive-7-series, and
21662           size.
21663
21664           When -mtune= is not specified, the default is rocket.
21665
21666           The size choice is not intended for use by end-users.  This is used
21667           when -Os is specified.  It overrides the instruction cost info
21668           provided by -mtune=, but does not override the pipeline info.  This
21669           helps reduce code size while still giving good performance.
21670
21671       -mpreferred-stack-boundary=num
21672           Attempt to keep the stack boundary aligned to a 2 raised to num
21673           byte boundary.  If -mpreferred-stack-boundary is not specified, the
21674           default is 4 (16 bytes or 128-bits).
21675
21676           Warning: If you use this switch, then you must build all modules
21677           with the same value, including any libraries.  This includes the
21678           system libraries and startup modules.
21679
21680       -msmall-data-limit=n
21681           Put global and static data smaller than n bytes into a special
21682           section (on some targets).
21683
21684       -msave-restore
21685       -mno-save-restore
21686           Do or don't use smaller but slower prologue and epilogue code that
21687           uses library function calls.  The default is to use fast inline
21688           prologues and epilogues.
21689
21690       -mstrict-align
21691       -mno-strict-align
21692           Do not or do generate unaligned memory accesses.  The default is
21693           set depending on whether the processor we are optimizing for
21694           supports fast unaligned access or not.
21695
21696       -mcmodel=medlow
21697           Generate code for the medium-low code model. The program and its
21698           statically defined symbols must lie within a single 2 GiB address
21699           range and must lie between absolute addresses -2 GiB and +2 GiB.
21700           Programs can be statically or dynamically linked. This is the
21701           default code model.
21702
21703       -mcmodel=medany
21704           Generate code for the medium-any code model. The program and its
21705           statically defined symbols must be within any single 2 GiB address
21706           range. Programs can be statically or dynamically linked.
21707
21708       -mexplicit-relocs
21709       -mno-exlicit-relocs
21710           Use or do not use assembler relocation operators when dealing with
21711           symbolic addresses.  The alternative is to use assembler macros
21712           instead, which may limit optimization.
21713
21714       -mrelax
21715       -mno-relax
21716           Take advantage of linker relaxations to reduce the number of
21717           instructions required to materialize symbol addresses. The default
21718           is to take advantage of linker relaxations.
21719
21720       -memit-attribute
21721       -mno-emit-attribute
21722           Emit (do not emit) RISC-V attribute to record extra information
21723           into ELF objects.  This feature requires at least binutils 2.32.
21724
21725       -malign-data=type
21726           Control how GCC aligns variables and constants of array, structure,
21727           or union types.  Supported values for type are xlen which uses x
21728           register width as the alignment value, and natural which uses
21729           natural alignment.  xlen is the default.
21730
21731   RL78 Options
21732       -msim
21733           Links in additional target libraries to support operation within a
21734           simulator.
21735
21736       -mmul=none
21737       -mmul=g10
21738       -mmul=g13
21739       -mmul=g14
21740       -mmul=rl78
21741           Specifies the type of hardware multiplication and division support
21742           to be used.  The simplest is "none", which uses software for both
21743           multiplication and division.  This is the default.  The "g13" value
21744           is for the hardware multiply/divide peripheral found on the
21745           RL78/G13 (S2 core) targets.  The "g14" value selects the use of the
21746           multiplication and division instructions supported by the RL78/G14
21747           (S3 core) parts.  The value "rl78" is an alias for "g14" and the
21748           value "mg10" is an alias for "none".
21749
21750           In addition a C preprocessor macro is defined, based upon the
21751           setting of this option.  Possible values are: "__RL78_MUL_NONE__",
21752           "__RL78_MUL_G13__" or "__RL78_MUL_G14__".
21753
21754       -mcpu=g10
21755       -mcpu=g13
21756       -mcpu=g14
21757       -mcpu=rl78
21758           Specifies the RL78 core to target.  The default is the G14 core,
21759           also known as an S3 core or just RL78.  The G13 or S2 core does not
21760           have multiply or divide instructions, instead it uses a hardware
21761           peripheral for these operations.  The G10 or S1 core does not have
21762           register banks, so it uses a different calling convention.
21763
21764           If this option is set it also selects the type of hardware multiply
21765           support to use, unless this is overridden by an explicit -mmul=none
21766           option on the command line.  Thus specifying -mcpu=g13 enables the
21767           use of the G13 hardware multiply peripheral and specifying
21768           -mcpu=g10 disables the use of hardware multiplications altogether.
21769
21770           Note, although the RL78/G14 core is the default target, specifying
21771           -mcpu=g14 or -mcpu=rl78 on the command line does change the
21772           behavior of the toolchain since it also enables G14 hardware
21773           multiply support.  If these options are not specified on the
21774           command line then software multiplication routines will be used
21775           even though the code targets the RL78 core.  This is for backwards
21776           compatibility with older toolchains which did not have hardware
21777           multiply and divide support.
21778
21779           In addition a C preprocessor macro is defined, based upon the
21780           setting of this option.  Possible values are: "__RL78_G10__",
21781           "__RL78_G13__" or "__RL78_G14__".
21782
21783       -mg10
21784       -mg13
21785       -mg14
21786       -mrl78
21787           These are aliases for the corresponding -mcpu= option.  They are
21788           provided for backwards compatibility.
21789
21790       -mallregs
21791           Allow the compiler to use all of the available registers.  By
21792           default registers "r24..r31" are reserved for use in interrupt
21793           handlers.  With this option enabled these registers can be used in
21794           ordinary functions as well.
21795
21796       -m64bit-doubles
21797       -m32bit-doubles
21798           Make the "double" data type be 64 bits (-m64bit-doubles) or 32 bits
21799           (-m32bit-doubles) in size.  The default is -m32bit-doubles.
21800
21801       -msave-mduc-in-interrupts
21802       -mno-save-mduc-in-interrupts
21803           Specifies that interrupt handler functions should preserve the MDUC
21804           registers.  This is only necessary if normal code might use the
21805           MDUC registers, for example because it performs multiplication and
21806           division operations.  The default is to ignore the MDUC registers
21807           as this makes the interrupt handlers faster.  The target option
21808           -mg13 needs to be passed for this to work as this feature is only
21809           available on the G13 target (S2 core).  The MDUC registers will
21810           only be saved if the interrupt handler performs a multiplication or
21811           division operation or it calls another function.
21812
21813   IBM RS/6000 and PowerPC Options
21814       These -m options are defined for the IBM RS/6000 and PowerPC:
21815
21816       -mpowerpc-gpopt
21817       -mno-powerpc-gpopt
21818       -mpowerpc-gfxopt
21819       -mno-powerpc-gfxopt
21820       -mpowerpc64
21821       -mno-powerpc64
21822       -mmfcrf
21823       -mno-mfcrf
21824       -mpopcntb
21825       -mno-popcntb
21826       -mpopcntd
21827       -mno-popcntd
21828       -mfprnd
21829       -mno-fprnd
21830       -mcmpb
21831       -mno-cmpb
21832       -mhard-dfp
21833       -mno-hard-dfp
21834           You use these options to specify which instructions are available
21835           on the processor you are using.  The default value of these options
21836           is determined when configuring GCC.  Specifying the -mcpu=cpu_type
21837           overrides the specification of these options.  We recommend you use
21838           the -mcpu=cpu_type option rather than the options listed above.
21839
21840           Specifying -mpowerpc-gpopt allows GCC to use the optional PowerPC
21841           architecture instructions in the General Purpose group, including
21842           floating-point square root.  Specifying -mpowerpc-gfxopt allows GCC
21843           to use the optional PowerPC architecture instructions in the
21844           Graphics group, including floating-point select.
21845
21846           The -mmfcrf option allows GCC to generate the move from condition
21847           register field instruction implemented on the POWER4 processor and
21848           other processors that support the PowerPC V2.01 architecture.  The
21849           -mpopcntb option allows GCC to generate the popcount and double-
21850           precision FP reciprocal estimate instruction implemented on the
21851           POWER5 processor and other processors that support the PowerPC
21852           V2.02 architecture.  The -mpopcntd option allows GCC to generate
21853           the popcount instruction implemented on the POWER7 processor and
21854           other processors that support the PowerPC V2.06 architecture.  The
21855           -mfprnd option allows GCC to generate the FP round to integer
21856           instructions implemented on the POWER5+ processor and other
21857           processors that support the PowerPC V2.03 architecture.  The -mcmpb
21858           option allows GCC to generate the compare bytes instruction
21859           implemented on the POWER6 processor and other processors that
21860           support the PowerPC V2.05 architecture.  The -mhard-dfp option
21861           allows GCC to generate the decimal floating-point instructions
21862           implemented on some POWER processors.
21863
21864           The -mpowerpc64 option allows GCC to generate the additional 64-bit
21865           instructions that are found in the full PowerPC64 architecture and
21866           to treat GPRs as 64-bit, doubleword quantities.  GCC defaults to
21867           -mno-powerpc64.
21868
21869       -mcpu=cpu_type
21870           Set architecture type, register usage, and instruction scheduling
21871           parameters for machine type cpu_type.  Supported values for
21872           cpu_type are 401, 403, 405, 405fp, 440, 440fp, 464, 464fp, 476,
21873           476fp, 505, 601, 602, 603, 603e, 604, 604e, 620, 630, 740, 7400,
21874           7450, 750, 801, 821, 823, 860, 970, 8540, a2, e300c2, e300c3,
21875           e500mc, e500mc64, e5500, e6500, ec603e, G3, G4, G5, titan, power3,
21876           power4, power5, power5+, power6, power6x, power7, power8, power9,
21877           future, powerpc, powerpc64, powerpc64le, rs64, and native.
21878
21879           -mcpu=powerpc, -mcpu=powerpc64, and -mcpu=powerpc64le specify pure
21880           32-bit PowerPC (either endian), 64-bit big endian PowerPC and
21881           64-bit little endian PowerPC architecture machine types, with an
21882           appropriate, generic processor model assumed for scheduling
21883           purposes.
21884
21885           Specifying native as cpu type detects and selects the architecture
21886           option that corresponds to the host processor of the system
21887           performing the compilation.  -mcpu=native has no effect if GCC does
21888           not recognize the processor.
21889
21890           The other options specify a specific processor.  Code generated
21891           under those options runs best on that processor, and may not run at
21892           all on others.
21893
21894           The -mcpu options automatically enable or disable the following
21895           options:
21896
21897           -maltivec  -mfprnd  -mhard-float  -mmfcrf  -mmultiple -mpopcntb
21898           -mpopcntd  -mpowerpc64 -mpowerpc-gpopt  -mpowerpc-gfxopt -mmulhw
21899           -mdlmzb  -mmfpgpr  -mvsx -mcrypto  -mhtm  -mpower8-fusion
21900           -mpower8-vector -mquad-memory  -mquad-memory-atomic  -mfloat128
21901           -mfloat128-hardware -mprefixed -mpcrel -mmma
21902
21903           The particular options set for any particular CPU varies between
21904           compiler versions, depending on what setting seems to produce
21905           optimal code for that CPU; it doesn't necessarily reflect the
21906           actual hardware's capabilities.  If you wish to set an individual
21907           option to a particular value, you may specify it after the -mcpu
21908           option, like -mcpu=970 -mno-altivec.
21909
21910           On AIX, the -maltivec and -mpowerpc64 options are not enabled or
21911           disabled by the -mcpu option at present because AIX does not have
21912           full support for these options.  You may still enable or disable
21913           them individually if you're sure it'll work in your environment.
21914
21915       -mtune=cpu_type
21916           Set the instruction scheduling parameters for machine type
21917           cpu_type, but do not set the architecture type or register usage,
21918           as -mcpu=cpu_type does.  The same values for cpu_type are used for
21919           -mtune as for -mcpu.  If both are specified, the code generated
21920           uses the architecture and registers set by -mcpu, but the
21921           scheduling parameters set by -mtune.
21922
21923       -mcmodel=small
21924           Generate PowerPC64 code for the small model: The TOC is limited to
21925           64k.
21926
21927       -mcmodel=medium
21928           Generate PowerPC64 code for the medium model: The TOC and other
21929           static data may be up to a total of 4G in size.  This is the
21930           default for 64-bit Linux.
21931
21932       -mcmodel=large
21933           Generate PowerPC64 code for the large model: The TOC may be up to
21934           4G in size.  Other data and code is only limited by the 64-bit
21935           address space.
21936
21937       -maltivec
21938       -mno-altivec
21939           Generate code that uses (does not use) AltiVec instructions, and
21940           also enable the use of built-in functions that allow more direct
21941           access to the AltiVec instruction set.  You may also need to set
21942           -mabi=altivec to adjust the current ABI with AltiVec ABI
21943           enhancements.
21944
21945           When -maltivec is used, the element order for AltiVec intrinsics
21946           such as "vec_splat", "vec_extract", and "vec_insert" match array
21947           element order corresponding to the endianness of the target.  That
21948           is, element zero identifies the leftmost element in a vector
21949           register when targeting a big-endian platform, and identifies the
21950           rightmost element in a vector register when targeting a little-
21951           endian platform.
21952
21953       -mvrsave
21954       -mno-vrsave
21955           Generate VRSAVE instructions when generating AltiVec code.
21956
21957       -msecure-plt
21958           Generate code that allows ld and ld.so to build executables and
21959           shared libraries with non-executable ".plt" and ".got" sections.
21960           This is a PowerPC 32-bit SYSV ABI option.
21961
21962       -mbss-plt
21963           Generate code that uses a BSS ".plt" section that ld.so fills in,
21964           and requires ".plt" and ".got" sections that are both writable and
21965           executable.  This is a PowerPC 32-bit SYSV ABI option.
21966
21967       -misel
21968       -mno-isel
21969           This switch enables or disables the generation of ISEL
21970           instructions.
21971
21972       -mvsx
21973       -mno-vsx
21974           Generate code that uses (does not use) vector/scalar (VSX)
21975           instructions, and also enable the use of built-in functions that
21976           allow more direct access to the VSX instruction set.
21977
21978       -mcrypto
21979       -mno-crypto
21980           Enable the use (disable) of the built-in functions that allow
21981           direct access to the cryptographic instructions that were added in
21982           version 2.07 of the PowerPC ISA.
21983
21984       -mhtm
21985       -mno-htm
21986           Enable (disable) the use of the built-in functions that allow
21987           direct access to the Hardware Transactional Memory (HTM)
21988           instructions that were added in version 2.07 of the PowerPC ISA.
21989
21990       -mpower8-fusion
21991       -mno-power8-fusion
21992           Generate code that keeps (does not keeps) some integer operations
21993           adjacent so that the instructions can be fused together on power8
21994           and later processors.
21995
21996       -mpower8-vector
21997       -mno-power8-vector
21998           Generate code that uses (does not use) the vector and scalar
21999           instructions that were added in version 2.07 of the PowerPC ISA.
22000           Also enable the use of built-in functions that allow more direct
22001           access to the vector instructions.
22002
22003       -mquad-memory
22004       -mno-quad-memory
22005           Generate code that uses (does not use) the non-atomic quad word
22006           memory instructions.  The -mquad-memory option requires use of
22007           64-bit mode.
22008
22009       -mquad-memory-atomic
22010       -mno-quad-memory-atomic
22011           Generate code that uses (does not use) the atomic quad word memory
22012           instructions.  The -mquad-memory-atomic option requires use of
22013           64-bit mode.
22014
22015       -mfloat128
22016       -mno-float128
22017           Enable/disable the __float128 keyword for IEEE 128-bit floating
22018           point and use either software emulation for IEEE 128-bit floating
22019           point or hardware instructions.
22020
22021           The VSX instruction set (-mvsx, -mcpu=power7, -mcpu=power8), or
22022           -mcpu=power9 must be enabled to use the IEEE 128-bit floating point
22023           support.  The IEEE 128-bit floating point support only works on
22024           PowerPC Linux systems.
22025
22026           The default for -mfloat128 is enabled on PowerPC Linux systems
22027           using the VSX instruction set, and disabled on other systems.
22028
22029           If you use the ISA 3.0 instruction set (-mpower9-vector or
22030           -mcpu=power9) on a 64-bit system, the IEEE 128-bit floating point
22031           support will also enable the generation of ISA 3.0 IEEE 128-bit
22032           floating point instructions.  Otherwise, if you do not specify to
22033           generate ISA 3.0 instructions or you are targeting a 32-bit big
22034           endian system, IEEE 128-bit floating point will be done with
22035           software emulation.
22036
22037       -mfloat128-hardware
22038       -mno-float128-hardware
22039           Enable/disable using ISA 3.0 hardware instructions to support the
22040           __float128 data type.
22041
22042           The default for -mfloat128-hardware is enabled on PowerPC Linux
22043           systems using the ISA 3.0 instruction set, and disabled on other
22044           systems.
22045
22046       -m32
22047       -m64
22048           Generate code for 32-bit or 64-bit environments of Darwin and SVR4
22049           targets (including GNU/Linux).  The 32-bit environment sets int,
22050           long and pointer to 32 bits and generates code that runs on any
22051           PowerPC variant.  The 64-bit environment sets int to 32 bits and
22052           long and pointer to 64 bits, and generates code for PowerPC64, as
22053           for -mpowerpc64.
22054
22055       -mfull-toc
22056       -mno-fp-in-toc
22057       -mno-sum-in-toc
22058       -mminimal-toc
22059           Modify generation of the TOC (Table Of Contents), which is created
22060           for every executable file.  The -mfull-toc option is selected by
22061           default.  In that case, GCC allocates at least one TOC entry for
22062           each unique non-automatic variable reference in your program.  GCC
22063           also places floating-point constants in the TOC.  However, only
22064           16,384 entries are available in the TOC.
22065
22066           If you receive a linker error message that saying you have
22067           overflowed the available TOC space, you can reduce the amount of
22068           TOC space used with the -mno-fp-in-toc and -mno-sum-in-toc options.
22069           -mno-fp-in-toc prevents GCC from putting floating-point constants
22070           in the TOC and -mno-sum-in-toc forces GCC to generate code to
22071           calculate the sum of an address and a constant at run time instead
22072           of putting that sum into the TOC.  You may specify one or both of
22073           these options.  Each causes GCC to produce very slightly slower and
22074           larger code at the expense of conserving TOC space.
22075
22076           If you still run out of space in the TOC even when you specify both
22077           of these options, specify -mminimal-toc instead.  This option
22078           causes GCC to make only one TOC entry for every file.  When you
22079           specify this option, GCC produces code that is slower and larger
22080           but which uses extremely little TOC space.  You may wish to use
22081           this option only on files that contain less frequently-executed
22082           code.
22083
22084       -maix64
22085       -maix32
22086           Enable 64-bit AIX ABI and calling convention: 64-bit pointers,
22087           64-bit "long" type, and the infrastructure needed to support them.
22088           Specifying -maix64 implies -mpowerpc64, while -maix32 disables the
22089           64-bit ABI and implies -mno-powerpc64.  GCC defaults to -maix32.
22090
22091       -mxl-compat
22092       -mno-xl-compat
22093           Produce code that conforms more closely to IBM XL compiler
22094           semantics when using AIX-compatible ABI.  Pass floating-point
22095           arguments to prototyped functions beyond the register save area
22096           (RSA) on the stack in addition to argument FPRs.  Do not assume
22097           that most significant double in 128-bit long double value is
22098           properly rounded when comparing values and converting to double.
22099           Use XL symbol names for long double support routines.
22100
22101           The AIX calling convention was extended but not initially
22102           documented to handle an obscure K&R C case of calling a function
22103           that takes the address of its arguments with fewer arguments than
22104           declared.  IBM XL compilers access floating-point arguments that do
22105           not fit in the RSA from the stack when a subroutine is compiled
22106           without optimization.  Because always storing floating-point
22107           arguments on the stack is inefficient and rarely needed, this
22108           option is not enabled by default and only is necessary when calling
22109           subroutines compiled by IBM XL compilers without optimization.
22110
22111       -mpe
22112           Support IBM RS/6000 SP Parallel Environment (PE).  Link an
22113           application written to use message passing with special startup
22114           code to enable the application to run.  The system must have PE
22115           installed in the standard location (/usr/lpp/ppe.poe/), or the
22116           specs file must be overridden with the -specs= option to specify
22117           the appropriate directory location.  The Parallel Environment does
22118           not support threads, so the -mpe option and the -pthread option are
22119           incompatible.
22120
22121       -malign-natural
22122       -malign-power
22123           On AIX, 32-bit Darwin, and 64-bit PowerPC GNU/Linux, the option
22124           -malign-natural overrides the ABI-defined alignment of larger
22125           types, such as floating-point doubles, on their natural size-based
22126           boundary.  The option -malign-power instructs GCC to follow the
22127           ABI-specified alignment rules.  GCC defaults to the standard
22128           alignment defined in the ABI.
22129
22130           On 64-bit Darwin, natural alignment is the default, and
22131           -malign-power is not supported.
22132
22133       -msoft-float
22134       -mhard-float
22135           Generate code that does not use (uses) the floating-point register
22136           set.  Software floating-point emulation is provided if you use the
22137           -msoft-float option, and pass the option to GCC when linking.
22138
22139       -mmultiple
22140       -mno-multiple
22141           Generate code that uses (does not use) the load multiple word
22142           instructions and the store multiple word instructions.  These
22143           instructions are generated by default on POWER systems, and not
22144           generated on PowerPC systems.  Do not use -mmultiple on little-
22145           endian PowerPC systems, since those instructions do not work when
22146           the processor is in little-endian mode.  The exceptions are PPC740
22147           and PPC750 which permit these instructions in little-endian mode.
22148
22149       -mupdate
22150       -mno-update
22151           Generate code that uses (does not use) the load or store
22152           instructions that update the base register to the address of the
22153           calculated memory location.  These instructions are generated by
22154           default.  If you use -mno-update, there is a small window between
22155           the time that the stack pointer is updated and the address of the
22156           previous frame is stored, which means code that walks the stack
22157           frame across interrupts or signals may get corrupted data.
22158
22159       -mavoid-indexed-addresses
22160       -mno-avoid-indexed-addresses
22161           Generate code that tries to avoid (not avoid) the use of indexed
22162           load or store instructions. These instructions can incur a
22163           performance penalty on Power6 processors in certain situations,
22164           such as when stepping through large arrays that cross a 16M
22165           boundary.  This option is enabled by default when targeting Power6
22166           and disabled otherwise.
22167
22168       -mfused-madd
22169       -mno-fused-madd
22170           Generate code that uses (does not use) the floating-point multiply
22171           and accumulate instructions.  These instructions are generated by
22172           default if hardware floating point is used.  The machine-dependent
22173           -mfused-madd option is now mapped to the machine-independent
22174           -ffp-contract=fast option, and -mno-fused-madd is mapped to
22175           -ffp-contract=off.
22176
22177       -mmulhw
22178       -mno-mulhw
22179           Generate code that uses (does not use) the half-word multiply and
22180           multiply-accumulate instructions on the IBM 405, 440, 464 and 476
22181           processors.  These instructions are generated by default when
22182           targeting those processors.
22183
22184       -mdlmzb
22185       -mno-dlmzb
22186           Generate code that uses (does not use) the string-search dlmzb
22187           instruction on the IBM 405, 440, 464 and 476 processors.  This
22188           instruction is generated by default when targeting those
22189           processors.
22190
22191       -mno-bit-align
22192       -mbit-align
22193           On System V.4 and embedded PowerPC systems do not (do) force
22194           structures and unions that contain bit-fields to be aligned to the
22195           base type of the bit-field.
22196
22197           For example, by default a structure containing nothing but 8
22198           "unsigned" bit-fields of length 1 is aligned to a 4-byte boundary
22199           and has a size of 4 bytes.  By using -mno-bit-align, the structure
22200           is aligned to a 1-byte boundary and is 1 byte in size.
22201
22202       -mno-strict-align
22203       -mstrict-align
22204           On System V.4 and embedded PowerPC systems do not (do) assume that
22205           unaligned memory references are handled by the system.
22206
22207       -mrelocatable
22208       -mno-relocatable
22209           Generate code that allows (does not allow) a static executable to
22210           be relocated to a different address at run time.  A simple embedded
22211           PowerPC system loader should relocate the entire contents of
22212           ".got2" and 4-byte locations listed in the ".fixup" section, a
22213           table of 32-bit addresses generated by this option.  For this to
22214           work, all objects linked together must be compiled with
22215           -mrelocatable or -mrelocatable-lib.  -mrelocatable code aligns the
22216           stack to an 8-byte boundary.
22217
22218       -mrelocatable-lib
22219       -mno-relocatable-lib
22220           Like -mrelocatable, -mrelocatable-lib generates a ".fixup" section
22221           to allow static executables to be relocated at run time, but
22222           -mrelocatable-lib does not use the smaller stack alignment of
22223           -mrelocatable.  Objects compiled with -mrelocatable-lib may be
22224           linked with objects compiled with any combination of the
22225           -mrelocatable options.
22226
22227       -mno-toc
22228       -mtoc
22229           On System V.4 and embedded PowerPC systems do not (do) assume that
22230           register 2 contains a pointer to a global area pointing to the
22231           addresses used in the program.
22232
22233       -mlittle
22234       -mlittle-endian
22235           On System V.4 and embedded PowerPC systems compile code for the
22236           processor in little-endian mode.  The -mlittle-endian option is the
22237           same as -mlittle.
22238
22239       -mbig
22240       -mbig-endian
22241           On System V.4 and embedded PowerPC systems compile code for the
22242           processor in big-endian mode.  The -mbig-endian option is the same
22243           as -mbig.
22244
22245       -mdynamic-no-pic
22246           On Darwin and Mac OS X systems, compile code so that it is not
22247           relocatable, but that its external references are relocatable.  The
22248           resulting code is suitable for applications, but not shared
22249           libraries.
22250
22251       -msingle-pic-base
22252           Treat the register used for PIC addressing as read-only, rather
22253           than loading it in the prologue for each function.  The runtime
22254           system is responsible for initializing this register with an
22255           appropriate value before execution begins.
22256
22257       -mprioritize-restricted-insns=priority
22258           This option controls the priority that is assigned to dispatch-slot
22259           restricted instructions during the second scheduling pass.  The
22260           argument priority takes the value 0, 1, or 2 to assign no, highest,
22261           or second-highest (respectively) priority to dispatch-slot
22262           restricted instructions.
22263
22264       -msched-costly-dep=dependence_type
22265           This option controls which dependences are considered costly by the
22266           target during instruction scheduling.  The argument dependence_type
22267           takes one of the following values:
22268
22269           no  No dependence is costly.
22270
22271           all All dependences are costly.
22272
22273           true_store_to_load
22274               A true dependence from store to load is costly.
22275
22276           store_to_load
22277               Any dependence from store to load is costly.
22278
22279           number
22280               Any dependence for which the latency is greater than or equal
22281               to number is costly.
22282
22283       -minsert-sched-nops=scheme
22284           This option controls which NOP insertion scheme is used during the
22285           second scheduling pass.  The argument scheme takes one of the
22286           following values:
22287
22288           no  Don't insert NOPs.
22289
22290           pad Pad with NOPs any dispatch group that has vacant issue slots,
22291               according to the scheduler's grouping.
22292
22293           regroup_exact
22294               Insert NOPs to force costly dependent insns into separate
22295               groups.  Insert exactly as many NOPs as needed to force an insn
22296               to a new group, according to the estimated processor grouping.
22297
22298           number
22299               Insert NOPs to force costly dependent insns into separate
22300               groups.  Insert number NOPs to force an insn to a new group.
22301
22302       -mcall-sysv
22303           On System V.4 and embedded PowerPC systems compile code using
22304           calling conventions that adhere to the March 1995 draft of the
22305           System V Application Binary Interface, PowerPC processor
22306           supplement.  This is the default unless you configured GCC using
22307           powerpc-*-eabiaix.
22308
22309       -mcall-sysv-eabi
22310       -mcall-eabi
22311           Specify both -mcall-sysv and -meabi options.
22312
22313       -mcall-sysv-noeabi
22314           Specify both -mcall-sysv and -mno-eabi options.
22315
22316       -mcall-aixdesc
22317           On System V.4 and embedded PowerPC systems compile code for the AIX
22318           operating system.
22319
22320       -mcall-linux
22321           On System V.4 and embedded PowerPC systems compile code for the
22322           Linux-based GNU system.
22323
22324       -mcall-freebsd
22325           On System V.4 and embedded PowerPC systems compile code for the
22326           FreeBSD operating system.
22327
22328       -mcall-netbsd
22329           On System V.4 and embedded PowerPC systems compile code for the
22330           NetBSD operating system.
22331
22332       -mcall-openbsd
22333           On System V.4 and embedded PowerPC systems compile code for the
22334           OpenBSD operating system.
22335
22336       -mtraceback=traceback_type
22337           Select the type of traceback table. Valid values for traceback_type
22338           are full, part, and no.
22339
22340       -maix-struct-return
22341           Return all structures in memory (as specified by the AIX ABI).
22342
22343       -msvr4-struct-return
22344           Return structures smaller than 8 bytes in registers (as specified
22345           by the SVR4 ABI).
22346
22347       -mabi=abi-type
22348           Extend the current ABI with a particular extension, or remove such
22349           extension.  Valid values are altivec, no-altivec, ibmlongdouble,
22350           ieeelongdouble, elfv1, elfv2.
22351
22352       -mabi=ibmlongdouble
22353           Change the current ABI to use IBM extended-precision long double.
22354           This is not likely to work if your system defaults to using IEEE
22355           extended-precision long double.  If you change the long double type
22356           from IEEE extended-precision, the compiler will issue a warning
22357           unless you use the -Wno-psabi option.  Requires -mlong-double-128
22358           to be enabled.
22359
22360       -mabi=ieeelongdouble
22361           Change the current ABI to use IEEE extended-precision long double.
22362           This is not likely to work if your system defaults to using IBM
22363           extended-precision long double.  If you change the long double type
22364           from IBM extended-precision, the compiler will issue a warning
22365           unless you use the -Wno-psabi option.  Requires -mlong-double-128
22366           to be enabled.
22367
22368       -mabi=elfv1
22369           Change the current ABI to use the ELFv1 ABI.  This is the default
22370           ABI for big-endian PowerPC 64-bit Linux.  Overriding the default
22371           ABI requires special system support and is likely to fail in
22372           spectacular ways.
22373
22374       -mabi=elfv2
22375           Change the current ABI to use the ELFv2 ABI.  This is the default
22376           ABI for little-endian PowerPC 64-bit Linux.  Overriding the default
22377           ABI requires special system support and is likely to fail in
22378           spectacular ways.
22379
22380       -mgnu-attribute
22381       -mno-gnu-attribute
22382           Emit .gnu_attribute assembly directives to set tag/value pairs in a
22383           .gnu.attributes section that specify ABI variations in function
22384           parameters or return values.
22385
22386       -mprototype
22387       -mno-prototype
22388           On System V.4 and embedded PowerPC systems assume that all calls to
22389           variable argument functions are properly prototyped.  Otherwise,
22390           the compiler must insert an instruction before every non-prototyped
22391           call to set or clear bit 6 of the condition code register ("CR") to
22392           indicate whether floating-point values are passed in the floating-
22393           point registers in case the function takes variable arguments.
22394           With -mprototype, only calls to prototyped variable argument
22395           functions set or clear the bit.
22396
22397       -msim
22398           On embedded PowerPC systems, assume that the startup module is
22399           called sim-crt0.o and that the standard C libraries are libsim.a
22400           and libc.a.  This is the default for powerpc-*-eabisim
22401           configurations.
22402
22403       -mmvme
22404           On embedded PowerPC systems, assume that the startup module is
22405           called crt0.o and the standard C libraries are libmvme.a and
22406           libc.a.
22407
22408       -mads
22409           On embedded PowerPC systems, assume that the startup module is
22410           called crt0.o and the standard C libraries are libads.a and libc.a.
22411
22412       -myellowknife
22413           On embedded PowerPC systems, assume that the startup module is
22414           called crt0.o and the standard C libraries are libyk.a and libc.a.
22415
22416       -mvxworks
22417           On System V.4 and embedded PowerPC systems, specify that you are
22418           compiling for a VxWorks system.
22419
22420       -memb
22421           On embedded PowerPC systems, set the "PPC_EMB" bit in the ELF flags
22422           header to indicate that eabi extended relocations are used.
22423
22424       -meabi
22425       -mno-eabi
22426           On System V.4 and embedded PowerPC systems do (do not) adhere to
22427           the Embedded Applications Binary Interface (EABI), which is a set
22428           of modifications to the System V.4 specifications.  Selecting
22429           -meabi means that the stack is aligned to an 8-byte boundary, a
22430           function "__eabi" is called from "main" to set up the EABI
22431           environment, and the -msdata option can use both "r2" and "r13" to
22432           point to two separate small data areas.  Selecting -mno-eabi means
22433           that the stack is aligned to a 16-byte boundary, no EABI
22434           initialization function is called from "main", and the -msdata
22435           option only uses "r13" to point to a single small data area.  The
22436           -meabi option is on by default if you configured GCC using one of
22437           the powerpc*-*-eabi* options.
22438
22439       -msdata=eabi
22440           On System V.4 and embedded PowerPC systems, put small initialized
22441           "const" global and static data in the ".sdata2" section, which is
22442           pointed to by register "r2".  Put small initialized non-"const"
22443           global and static data in the ".sdata" section, which is pointed to
22444           by register "r13".  Put small uninitialized global and static data
22445           in the ".sbss" section, which is adjacent to the ".sdata" section.
22446           The -msdata=eabi option is incompatible with the -mrelocatable
22447           option.  The -msdata=eabi option also sets the -memb option.
22448
22449       -msdata=sysv
22450           On System V.4 and embedded PowerPC systems, put small global and
22451           static data in the ".sdata" section, which is pointed to by
22452           register "r13".  Put small uninitialized global and static data in
22453           the ".sbss" section, which is adjacent to the ".sdata" section.
22454           The -msdata=sysv option is incompatible with the -mrelocatable
22455           option.
22456
22457       -msdata=default
22458       -msdata
22459           On System V.4 and embedded PowerPC systems, if -meabi is used,
22460           compile code the same as -msdata=eabi, otherwise compile code the
22461           same as -msdata=sysv.
22462
22463       -msdata=data
22464           On System V.4 and embedded PowerPC systems, put small global data
22465           in the ".sdata" section.  Put small uninitialized global data in
22466           the ".sbss" section.  Do not use register "r13" to address small
22467           data however.  This is the default behavior unless other -msdata
22468           options are used.
22469
22470       -msdata=none
22471       -mno-sdata
22472           On embedded PowerPC systems, put all initialized global and static
22473           data in the ".data" section, and all uninitialized data in the
22474           ".bss" section.
22475
22476       -mreadonly-in-sdata
22477           Put read-only objects in the ".sdata" section as well.  This is the
22478           default.
22479
22480       -mblock-move-inline-limit=num
22481           Inline all block moves (such as calls to "memcpy" or structure
22482           copies) less than or equal to num bytes.  The minimum value for num
22483           is 32 bytes on 32-bit targets and 64 bytes on 64-bit targets.  The
22484           default value is target-specific.
22485
22486       -mblock-compare-inline-limit=num
22487           Generate non-looping inline code for all block compares (such as
22488           calls to "memcmp" or structure compares) less than or equal to num
22489           bytes. If num is 0, all inline expansion (non-loop and loop) of
22490           block compare is disabled. The default value is target-specific.
22491
22492       -mblock-compare-inline-loop-limit=num
22493           Generate an inline expansion using loop code for all block compares
22494           that are less than or equal to num bytes, but greater than the
22495           limit for non-loop inline block compare expansion. If the block
22496           length is not constant, at most num bytes will be compared before
22497           "memcmp" is called to compare the remainder of the block. The
22498           default value is target-specific.
22499
22500       -mstring-compare-inline-limit=num
22501           Compare at most num string bytes with inline code.  If the
22502           difference or end of string is not found at the end of the inline
22503           compare a call to "strcmp" or "strncmp" will take care of the rest
22504           of the comparison. The default is 64 bytes.
22505
22506       -G num
22507           On embedded PowerPC systems, put global and static items less than
22508           or equal to num bytes into the small data or BSS sections instead
22509           of the normal data or BSS section.  By default, num is 8.  The -G
22510           num switch is also passed to the linker.  All modules should be
22511           compiled with the same -G num value.
22512
22513       -mregnames
22514       -mno-regnames
22515           On System V.4 and embedded PowerPC systems do (do not) emit
22516           register names in the assembly language output using symbolic
22517           forms.
22518
22519       -mlongcall
22520       -mno-longcall
22521           By default assume that all calls are far away so that a longer and
22522           more expensive calling sequence is required.  This is required for
22523           calls farther than 32 megabytes (33,554,432 bytes) from the current
22524           location.  A short call is generated if the compiler knows the call
22525           cannot be that far away.  This setting can be overridden by the
22526           "shortcall" function attribute, or by "#pragma longcall(0)".
22527
22528           Some linkers are capable of detecting out-of-range calls and
22529           generating glue code on the fly.  On these systems, long calls are
22530           unnecessary and generate slower code.  As of this writing, the AIX
22531           linker can do this, as can the GNU linker for PowerPC/64.  It is
22532           planned to add this feature to the GNU linker for 32-bit PowerPC
22533           systems as well.
22534
22535           On PowerPC64 ELFv2 and 32-bit PowerPC systems with newer GNU
22536           linkers, GCC can generate long calls using an inline PLT call
22537           sequence (see -mpltseq).  PowerPC with -mbss-plt and PowerPC64
22538           ELFv1 (big-endian) do not support inline PLT calls.
22539
22540           On Darwin/PPC systems, "#pragma longcall" generates "jbsr callee,
22541           L42", plus a branch island (glue code).  The two target addresses
22542           represent the callee and the branch island.  The Darwin/PPC linker
22543           prefers the first address and generates a "bl callee" if the PPC
22544           "bl" instruction reaches the callee directly; otherwise, the linker
22545           generates "bl L42" to call the branch island.  The branch island is
22546           appended to the body of the calling function; it computes the full
22547           32-bit address of the callee and jumps to it.
22548
22549           On Mach-O (Darwin) systems, this option directs the compiler emit
22550           to the glue for every direct call, and the Darwin linker decides
22551           whether to use or discard it.
22552
22553           In the future, GCC may ignore all longcall specifications when the
22554           linker is known to generate glue.
22555
22556       -mpltseq
22557       -mno-pltseq
22558           Implement (do not implement) -fno-plt and long calls using an
22559           inline PLT call sequence that supports lazy linking and long calls
22560           to functions in dlopen'd shared libraries.  Inline PLT calls are
22561           only supported on PowerPC64 ELFv2 and 32-bit PowerPC systems with
22562           newer GNU linkers, and are enabled by default if the support is
22563           detected when configuring GCC, and, in the case of 32-bit PowerPC,
22564           if GCC is configured with --enable-secureplt.  -mpltseq code and
22565           -mbss-plt 32-bit PowerPC relocatable objects may not be linked
22566           together.
22567
22568       -mtls-markers
22569       -mno-tls-markers
22570           Mark (do not mark) calls to "__tls_get_addr" with a relocation
22571           specifying the function argument.  The relocation allows the linker
22572           to reliably associate function call with argument setup
22573           instructions for TLS optimization, which in turn allows GCC to
22574           better schedule the sequence.
22575
22576       -mrecip
22577       -mno-recip
22578           This option enables use of the reciprocal estimate and reciprocal
22579           square root estimate instructions with additional Newton-Raphson
22580           steps to increase precision instead of doing a divide or square
22581           root and divide for floating-point arguments.  You should use the
22582           -ffast-math option when using -mrecip (or at least
22583           -funsafe-math-optimizations, -ffinite-math-only, -freciprocal-math
22584           and -fno-trapping-math).  Note that while the throughput of the
22585           sequence is generally higher than the throughput of the non-
22586           reciprocal instruction, the precision of the sequence can be
22587           decreased by up to 2 ulp (i.e. the inverse of 1.0 equals
22588           0.99999994) for reciprocal square roots.
22589
22590       -mrecip=opt
22591           This option controls which reciprocal estimate instructions may be
22592           used.  opt is a comma-separated list of options, which may be
22593           preceded by a "!" to invert the option:
22594
22595           all Enable all estimate instructions.
22596
22597           default
22598               Enable the default instructions, equivalent to -mrecip.
22599
22600           none
22601               Disable all estimate instructions, equivalent to -mno-recip.
22602
22603           div Enable the reciprocal approximation instructions for both
22604               single and double precision.
22605
22606           divf
22607               Enable the single-precision reciprocal approximation
22608               instructions.
22609
22610           divd
22611               Enable the double-precision reciprocal approximation
22612               instructions.
22613
22614           rsqrt
22615               Enable the reciprocal square root approximation instructions
22616               for both single and double precision.
22617
22618           rsqrtf
22619               Enable the single-precision reciprocal square root
22620               approximation instructions.
22621
22622           rsqrtd
22623               Enable the double-precision reciprocal square root
22624               approximation instructions.
22625
22626           So, for example, -mrecip=all,!rsqrtd enables all of the reciprocal
22627           estimate instructions, except for the "FRSQRTE", "XSRSQRTEDP", and
22628           "XVRSQRTEDP" instructions which handle the double-precision
22629           reciprocal square root calculations.
22630
22631       -mrecip-precision
22632       -mno-recip-precision
22633           Assume (do not assume) that the reciprocal estimate instructions
22634           provide higher-precision estimates than is mandated by the PowerPC
22635           ABI.  Selecting -mcpu=power6, -mcpu=power7 or -mcpu=power8
22636           automatically selects -mrecip-precision.  The double-precision
22637           square root estimate instructions are not generated by default on
22638           low-precision machines, since they do not provide an estimate that
22639           converges after three steps.
22640
22641       -mveclibabi=type
22642           Specifies the ABI type to use for vectorizing intrinsics using an
22643           external library.  The only type supported at present is mass,
22644           which specifies to use IBM's Mathematical Acceleration Subsystem
22645           (MASS) libraries for vectorizing intrinsics using external
22646           libraries.  GCC currently emits calls to "acosd2", "acosf4",
22647           "acoshd2", "acoshf4", "asind2", "asinf4", "asinhd2", "asinhf4",
22648           "atan2d2", "atan2f4", "atand2", "atanf4", "atanhd2", "atanhf4",
22649           "cbrtd2", "cbrtf4", "cosd2", "cosf4", "coshd2", "coshf4", "erfcd2",
22650           "erfcf4", "erfd2", "erff4", "exp2d2", "exp2f4", "expd2", "expf4",
22651           "expm1d2", "expm1f4", "hypotd2", "hypotf4", "lgammad2", "lgammaf4",
22652           "log10d2", "log10f4", "log1pd2", "log1pf4", "log2d2", "log2f4",
22653           "logd2", "logf4", "powd2", "powf4", "sind2", "sinf4", "sinhd2",
22654           "sinhf4", "sqrtd2", "sqrtf4", "tand2", "tanf4", "tanhd2", and
22655           "tanhf4" when generating code for power7.  Both -ftree-vectorize
22656           and -funsafe-math-optimizations must also be enabled.  The MASS
22657           libraries must be specified at link time.
22658
22659       -mfriz
22660       -mno-friz
22661           Generate (do not generate) the "friz" instruction when the
22662           -funsafe-math-optimizations option is used to optimize rounding of
22663           floating-point values to 64-bit integer and back to floating point.
22664           The "friz" instruction does not return the same value if the
22665           floating-point number is too large to fit in an integer.
22666
22667       -mpointers-to-nested-functions
22668       -mno-pointers-to-nested-functions
22669           Generate (do not generate) code to load up the static chain
22670           register ("r11") when calling through a pointer on AIX and 64-bit
22671           Linux systems where a function pointer points to a 3-word
22672           descriptor giving the function address, TOC value to be loaded in
22673           register "r2", and static chain value to be loaded in register
22674           "r11".  The -mpointers-to-nested-functions is on by default.  You
22675           cannot call through pointers to nested functions or pointers to
22676           functions compiled in other languages that use the static chain if
22677           you use -mno-pointers-to-nested-functions.
22678
22679       -msave-toc-indirect
22680       -mno-save-toc-indirect
22681           Generate (do not generate) code to save the TOC value in the
22682           reserved stack location in the function prologue if the function
22683           calls through a pointer on AIX and 64-bit Linux systems.  If the
22684           TOC value is not saved in the prologue, it is saved just before the
22685           call through the pointer.  The -mno-save-toc-indirect option is the
22686           default.
22687
22688       -mcompat-align-parm
22689       -mno-compat-align-parm
22690           Generate (do not generate) code to pass structure parameters with a
22691           maximum alignment of 64 bits, for compatibility with older versions
22692           of GCC.
22693
22694           Older versions of GCC (prior to 4.9.0) incorrectly did not align a
22695           structure parameter on a 128-bit boundary when that structure
22696           contained a member requiring 128-bit alignment.  This is corrected
22697           in more recent versions of GCC.  This option may be used to
22698           generate code that is compatible with functions compiled with older
22699           versions of GCC.
22700
22701           The -mno-compat-align-parm option is the default.
22702
22703       -mstack-protector-guard=guard
22704       -mstack-protector-guard-reg=reg
22705       -mstack-protector-guard-offset=offset
22706       -mstack-protector-guard-symbol=symbol
22707           Generate stack protection code using canary at guard.  Supported
22708           locations are global for global canary or tls for per-thread canary
22709           in the TLS block (the default with GNU libc version 2.4 or later).
22710
22711           With the latter choice the options -mstack-protector-guard-reg=reg
22712           and -mstack-protector-guard-offset=offset furthermore specify which
22713           register to use as base register for reading the canary, and from
22714           what offset from that base register. The default for those is as
22715           specified in the relevant ABI.
22716           -mstack-protector-guard-symbol=symbol overrides the offset with a
22717           symbol reference to a canary in the TLS block.
22718
22719       -mpcrel
22720       -mno-pcrel
22721           Generate (do not generate) pc-relative addressing when the option
22722           -mcpu=future is used.  The -mpcrel option requires that the medium
22723           code model (-mcmodel=medium) and prefixed addressing (-mprefixed)
22724           options are enabled.
22725
22726       -mprefixed
22727       -mno-prefixed
22728           Generate (do not generate) addressing modes using prefixed load and
22729           store instructions when the option -mcpu=future is used.
22730
22731       -mmma
22732       -mno-mma
22733           Generate (do not generate) the MMA instructions when the option
22734           -mcpu=future is used.
22735
22736   RX Options
22737       These command-line options are defined for RX targets:
22738
22739       -m64bit-doubles
22740       -m32bit-doubles
22741           Make the "double" data type be 64 bits (-m64bit-doubles) or 32 bits
22742           (-m32bit-doubles) in size.  The default is -m32bit-doubles.  Note
22743           RX floating-point hardware only works on 32-bit values, which is
22744           why the default is -m32bit-doubles.
22745
22746       -fpu
22747       -nofpu
22748           Enables (-fpu) or disables (-nofpu) the use of RX floating-point
22749           hardware.  The default is enabled for the RX600 series and disabled
22750           for the RX200 series.
22751
22752           Floating-point instructions are only generated for 32-bit floating-
22753           point values, however, so the FPU hardware is not used for doubles
22754           if the -m64bit-doubles option is used.
22755
22756           Note If the -fpu option is enabled then -funsafe-math-optimizations
22757           is also enabled automatically.  This is because the RX FPU
22758           instructions are themselves unsafe.
22759
22760       -mcpu=name
22761           Selects the type of RX CPU to be targeted.  Currently three types
22762           are supported, the generic RX600 and RX200 series hardware and the
22763           specific RX610 CPU.  The default is RX600.
22764
22765           The only difference between RX600 and RX610 is that the RX610 does
22766           not support the "MVTIPL" instruction.
22767
22768           The RX200 series does not have a hardware floating-point unit and
22769           so -nofpu is enabled by default when this type is selected.
22770
22771       -mbig-endian-data
22772       -mlittle-endian-data
22773           Store data (but not code) in the big-endian format.  The default is
22774           -mlittle-endian-data, i.e. to store data in the little-endian
22775           format.
22776
22777       -msmall-data-limit=N
22778           Specifies the maximum size in bytes of global and static variables
22779           which can be placed into the small data area.  Using the small data
22780           area can lead to smaller and faster code, but the size of area is
22781           limited and it is up to the programmer to ensure that the area does
22782           not overflow.  Also when the small data area is used one of the
22783           RX's registers (usually "r13") is reserved for use pointing to this
22784           area, so it is no longer available for use by the compiler.  This
22785           could result in slower and/or larger code if variables are pushed
22786           onto the stack instead of being held in this register.
22787
22788           Note, common variables (variables that have not been initialized)
22789           and constants are not placed into the small data area as they are
22790           assigned to other sections in the output executable.
22791
22792           The default value is zero, which disables this feature.  Note, this
22793           feature is not enabled by default with higher optimization levels
22794           (-O2 etc) because of the potentially detrimental effects of
22795           reserving a register.  It is up to the programmer to experiment and
22796           discover whether this feature is of benefit to their program.  See
22797           the description of the -mpid option for a description of how the
22798           actual register to hold the small data area pointer is chosen.
22799
22800       -msim
22801       -mno-sim
22802           Use the simulator runtime.  The default is to use the libgloss
22803           board-specific runtime.
22804
22805       -mas100-syntax
22806       -mno-as100-syntax
22807           When generating assembler output use a syntax that is compatible
22808           with Renesas's AS100 assembler.  This syntax can also be handled by
22809           the GAS assembler, but it has some restrictions so it is not
22810           generated by default.
22811
22812       -mmax-constant-size=N
22813           Specifies the maximum size, in bytes, of a constant that can be
22814           used as an operand in a RX instruction.  Although the RX
22815           instruction set does allow constants of up to 4 bytes in length to
22816           be used in instructions, a longer value equates to a longer
22817           instruction.  Thus in some circumstances it can be beneficial to
22818           restrict the size of constants that are used in instructions.
22819           Constants that are too big are instead placed into a constant pool
22820           and referenced via register indirection.
22821
22822           The value N can be between 0 and 4.  A value of 0 (the default) or
22823           4 means that constants of any size are allowed.
22824
22825       -mrelax
22826           Enable linker relaxation.  Linker relaxation is a process whereby
22827           the linker attempts to reduce the size of a program by finding
22828           shorter versions of various instructions.  Disabled by default.
22829
22830       -mint-register=N
22831           Specify the number of registers to reserve for fast interrupt
22832           handler functions.  The value N can be between 0 and 4.  A value of
22833           1 means that register "r13" is reserved for the exclusive use of
22834           fast interrupt handlers.  A value of 2 reserves "r13" and "r12".  A
22835           value of 3 reserves "r13", "r12" and "r11", and a value of 4
22836           reserves "r13" through "r10".  A value of 0, the default, does not
22837           reserve any registers.
22838
22839       -msave-acc-in-interrupts
22840           Specifies that interrupt handler functions should preserve the
22841           accumulator register.  This is only necessary if normal code might
22842           use the accumulator register, for example because it performs
22843           64-bit multiplications.  The default is to ignore the accumulator
22844           as this makes the interrupt handlers faster.
22845
22846       -mpid
22847       -mno-pid
22848           Enables the generation of position independent data.  When enabled
22849           any access to constant data is done via an offset from a base
22850           address held in a register.  This allows the location of constant
22851           data to be determined at run time without requiring the executable
22852           to be relocated, which is a benefit to embedded applications with
22853           tight memory constraints.  Data that can be modified is not
22854           affected by this option.
22855
22856           Note, using this feature reserves a register, usually "r13", for
22857           the constant data base address.  This can result in slower and/or
22858           larger code, especially in complicated functions.
22859
22860           The actual register chosen to hold the constant data base address
22861           depends upon whether the -msmall-data-limit and/or the
22862           -mint-register command-line options are enabled.  Starting with
22863           register "r13" and proceeding downwards, registers are allocated
22864           first to satisfy the requirements of -mint-register, then -mpid and
22865           finally -msmall-data-limit.  Thus it is possible for the small data
22866           area register to be "r8" if both -mint-register=4 and -mpid are
22867           specified on the command line.
22868
22869           By default this feature is not enabled.  The default can be
22870           restored via the -mno-pid command-line option.
22871
22872       -mno-warn-multiple-fast-interrupts
22873       -mwarn-multiple-fast-interrupts
22874           Prevents GCC from issuing a warning message if it finds more than
22875           one fast interrupt handler when it is compiling a file.  The
22876           default is to issue a warning for each extra fast interrupt handler
22877           found, as the RX only supports one such interrupt.
22878
22879       -mallow-string-insns
22880       -mno-allow-string-insns
22881           Enables or disables the use of the string manipulation instructions
22882           "SMOVF", "SCMPU", "SMOVB", "SMOVU", "SUNTIL" "SWHILE" and also the
22883           "RMPA" instruction.  These instructions may prefetch data, which is
22884           not safe to do if accessing an I/O register.  (See section 12.2.7
22885           of the RX62N Group User's Manual for more information).
22886
22887           The default is to allow these instructions, but it is not possible
22888           for GCC to reliably detect all circumstances where a string
22889           instruction might be used to access an I/O register, so their use
22890           cannot be disabled automatically.  Instead it is reliant upon the
22891           programmer to use the -mno-allow-string-insns option if their
22892           program accesses I/O space.
22893
22894           When the instructions are enabled GCC defines the C preprocessor
22895           symbol "__RX_ALLOW_STRING_INSNS__", otherwise it defines the symbol
22896           "__RX_DISALLOW_STRING_INSNS__".
22897
22898       -mjsr
22899       -mno-jsr
22900           Use only (or not only) "JSR" instructions to access functions.
22901           This option can be used when code size exceeds the range of "BSR"
22902           instructions.  Note that -mno-jsr does not mean to not use "JSR"
22903           but instead means that any type of branch may be used.
22904
22905       Note: The generic GCC command-line option -ffixed-reg has special
22906       significance to the RX port when used with the "interrupt" function
22907       attribute.  This attribute indicates a function intended to process
22908       fast interrupts.  GCC ensures that it only uses the registers "r10",
22909       "r11", "r12" and/or "r13" and only provided that the normal use of the
22910       corresponding registers have been restricted via the -ffixed-reg or
22911       -mint-register command-line options.
22912
22913   S/390 and zSeries Options
22914       These are the -m options defined for the S/390 and zSeries
22915       architecture.
22916
22917       -mhard-float
22918       -msoft-float
22919           Use (do not use) the hardware floating-point instructions and
22920           registers for floating-point operations.  When -msoft-float is
22921           specified, functions in libgcc.a are used to perform floating-point
22922           operations.  When -mhard-float is specified, the compiler generates
22923           IEEE floating-point instructions.  This is the default.
22924
22925       -mhard-dfp
22926       -mno-hard-dfp
22927           Use (do not use) the hardware decimal-floating-point instructions
22928           for decimal-floating-point operations.  When -mno-hard-dfp is
22929           specified, functions in libgcc.a are used to perform decimal-
22930           floating-point operations.  When -mhard-dfp is specified, the
22931           compiler generates decimal-floating-point hardware instructions.
22932           This is the default for -march=z9-ec or higher.
22933
22934       -mlong-double-64
22935       -mlong-double-128
22936           These switches control the size of "long double" type. A size of 64
22937           bits makes the "long double" type equivalent to the "double" type.
22938           This is the default.
22939
22940       -mbackchain
22941       -mno-backchain
22942           Store (do not store) the address of the caller's frame as backchain
22943           pointer into the callee's stack frame.  A backchain may be needed
22944           to allow debugging using tools that do not understand DWARF call
22945           frame information.  When -mno-packed-stack is in effect, the
22946           backchain pointer is stored at the bottom of the stack frame; when
22947           -mpacked-stack is in effect, the backchain is placed into the
22948           topmost word of the 96/160 byte register save area.
22949
22950           In general, code compiled with -mbackchain is call-compatible with
22951           code compiled with -mmo-backchain; however, use of the backchain
22952           for debugging purposes usually requires that the whole binary is
22953           built with -mbackchain.  Note that the combination of -mbackchain,
22954           -mpacked-stack and -mhard-float is not supported.  In order to
22955           build a linux kernel use -msoft-float.
22956
22957           The default is to not maintain the backchain.
22958
22959       -mpacked-stack
22960       -mno-packed-stack
22961           Use (do not use) the packed stack layout.  When -mno-packed-stack
22962           is specified, the compiler uses the all fields of the 96/160 byte
22963           register save area only for their default purpose; unused fields
22964           still take up stack space.  When -mpacked-stack is specified,
22965           register save slots are densely packed at the top of the register
22966           save area; unused space is reused for other purposes, allowing for
22967           more efficient use of the available stack space.  However, when
22968           -mbackchain is also in effect, the topmost word of the save area is
22969           always used to store the backchain, and the return address register
22970           is always saved two words below the backchain.
22971
22972           As long as the stack frame backchain is not used, code generated
22973           with -mpacked-stack is call-compatible with code generated with
22974           -mno-packed-stack.  Note that some non-FSF releases of GCC 2.95 for
22975           S/390 or zSeries generated code that uses the stack frame backchain
22976           at run time, not just for debugging purposes.  Such code is not
22977           call-compatible with code compiled with -mpacked-stack.  Also, note
22978           that the combination of -mbackchain, -mpacked-stack and
22979           -mhard-float is not supported.  In order to build a linux kernel
22980           use -msoft-float.
22981
22982           The default is to not use the packed stack layout.
22983
22984       -msmall-exec
22985       -mno-small-exec
22986           Generate (or do not generate) code using the "bras" instruction to
22987           do subroutine calls.  This only works reliably if the total
22988           executable size does not exceed 64k.  The default is to use the
22989           "basr" instruction instead, which does not have this limitation.
22990
22991       -m64
22992       -m31
22993           When -m31 is specified, generate code compliant to the GNU/Linux
22994           for S/390 ABI.  When -m64 is specified, generate code compliant to
22995           the GNU/Linux for zSeries ABI.  This allows GCC in particular to
22996           generate 64-bit instructions.  For the s390 targets, the default is
22997           -m31, while the s390x targets default to -m64.
22998
22999       -mzarch
23000       -mesa
23001           When -mzarch is specified, generate code using the instructions
23002           available on z/Architecture.  When -mesa is specified, generate
23003           code using the instructions available on ESA/390.  Note that -mesa
23004           is not possible with -m64.  When generating code compliant to the
23005           GNU/Linux for S/390 ABI, the default is -mesa.  When generating
23006           code compliant to the GNU/Linux for zSeries ABI, the default is
23007           -mzarch.
23008
23009       -mhtm
23010       -mno-htm
23011           The -mhtm option enables a set of builtins making use of
23012           instructions available with the transactional execution facility
23013           introduced with the IBM zEnterprise EC12 machine generation S/390
23014           System z Built-in Functions.  -mhtm is enabled by default when
23015           using -march=zEC12.
23016
23017       -mvx
23018       -mno-vx
23019           When -mvx is specified, generate code using the instructions
23020           available with the vector extension facility introduced with the
23021           IBM z13 machine generation.  This option changes the ABI for some
23022           vector type values with regard to alignment and calling
23023           conventions.  In case vector type values are being used in an ABI-
23024           relevant context a GAS .gnu_attribute command will be added to mark
23025           the resulting binary with the ABI used.  -mvx is enabled by default
23026           when using -march=z13.
23027
23028       -mzvector
23029       -mno-zvector
23030           The -mzvector option enables vector language extensions and
23031           builtins using instructions available with the vector extension
23032           facility introduced with the IBM z13 machine generation.  This
23033           option adds support for vector to be used as a keyword to define
23034           vector type variables and arguments.  vector is only available when
23035           GNU extensions are enabled.  It will not be expanded when
23036           requesting strict standard compliance e.g. with -std=c99.  In
23037           addition to the GCC low-level builtins -mzvector enables a set of
23038           builtins added for compatibility with AltiVec-style implementations
23039           like Power and Cell.  In order to make use of these builtins the
23040           header file vecintrin.h needs to be included.  -mzvector is
23041           disabled by default.
23042
23043       -mmvcle
23044       -mno-mvcle
23045           Generate (or do not generate) code using the "mvcle" instruction to
23046           perform block moves.  When -mno-mvcle is specified, use a "mvc"
23047           loop instead.  This is the default unless optimizing for size.
23048
23049       -mdebug
23050       -mno-debug
23051           Print (or do not print) additional debug information when
23052           compiling.  The default is to not print debug information.
23053
23054       -march=cpu-type
23055           Generate code that runs on cpu-type, which is the name of a system
23056           representing a certain processor type.  Possible values for cpu-
23057           type are z900/arch5, z990/arch6, z9-109, z9-ec/arch7, z10/arch8,
23058           z196/arch9, zEC12, z13/arch11, z14/arch12, z15/arch13, and native.
23059
23060           The default is -march=z900.
23061
23062           Specifying native as cpu type can be used to select the best
23063           architecture option for the host processor.  -march=native has no
23064           effect if GCC does not recognize the processor.
23065
23066       -mtune=cpu-type
23067           Tune to cpu-type everything applicable about the generated code,
23068           except for the ABI and the set of available instructions.  The list
23069           of cpu-type values is the same as for -march.  The default is the
23070           value used for -march.
23071
23072       -mtpf-trace
23073       -mno-tpf-trace
23074           Generate code that adds (does not add) in TPF OS specific branches
23075           to trace routines in the operating system.  This option is off by
23076           default, even when compiling for the TPF OS.
23077
23078       -mtpf-trace-skip
23079       -mno-tpf-trace-skip
23080           Generate code that changes (does not change) the default branch
23081           targets enabled by -mtpf-trace to point to specialized trace
23082           routines providing the ability of selectively skipping function
23083           trace entries for the TPF OS.  This option is off by default, even
23084           when compiling for the TPF OS and specifying -mtpf-trace.
23085
23086       -mfused-madd
23087       -mno-fused-madd
23088           Generate code that uses (does not use) the floating-point multiply
23089           and accumulate instructions.  These instructions are generated by
23090           default if hardware floating point is used.
23091
23092       -mwarn-framesize=framesize
23093           Emit a warning if the current function exceeds the given frame
23094           size.  Because this is a compile-time check it doesn't need to be a
23095           real problem when the program runs.  It is intended to identify
23096           functions that most probably cause a stack overflow.  It is useful
23097           to be used in an environment with limited stack size e.g. the linux
23098           kernel.
23099
23100       -mwarn-dynamicstack
23101           Emit a warning if the function calls "alloca" or uses dynamically-
23102           sized arrays.  This is generally a bad idea with a limited stack
23103           size.
23104
23105       -mstack-guard=stack-guard
23106       -mstack-size=stack-size
23107           If these options are provided the S/390 back end emits additional
23108           instructions in the function prologue that trigger a trap if the
23109           stack size is stack-guard bytes above the stack-size (remember that
23110           the stack on S/390 grows downward).  If the stack-guard option is
23111           omitted the smallest power of 2 larger than the frame size of the
23112           compiled function is chosen.  These options are intended to be used
23113           to help debugging stack overflow problems.  The additionally
23114           emitted code causes only little overhead and hence can also be used
23115           in production-like systems without greater performance degradation.
23116           The given values have to be exact powers of 2 and stack-size has to
23117           be greater than stack-guard without exceeding 64k.  In order to be
23118           efficient the extra code makes the assumption that the stack starts
23119           at an address aligned to the value given by stack-size.  The stack-
23120           guard option can only be used in conjunction with stack-size.
23121
23122       -mhotpatch=pre-halfwords,post-halfwords
23123           If the hotpatch option is enabled, a "hot-patching" function
23124           prologue is generated for all functions in the compilation unit.
23125           The funtion label is prepended with the given number of two-byte
23126           NOP instructions (pre-halfwords, maximum 1000000).  After the
23127           label, 2 * post-halfwords bytes are appended, using the largest NOP
23128           like instructions the architecture allows (maximum 1000000).
23129
23130           If both arguments are zero, hotpatching is disabled.
23131
23132           This option can be overridden for individual functions with the
23133           "hotpatch" attribute.
23134
23135   Score Options
23136       These options are defined for Score implementations:
23137
23138       -meb
23139           Compile code for big-endian mode.  This is the default.
23140
23141       -mel
23142           Compile code for little-endian mode.
23143
23144       -mnhwloop
23145           Disable generation of "bcnz" instructions.
23146
23147       -muls
23148           Enable generation of unaligned load and store instructions.
23149
23150       -mmac
23151           Enable the use of multiply-accumulate instructions. Disabled by
23152           default.
23153
23154       -mscore5
23155           Specify the SCORE5 as the target architecture.
23156
23157       -mscore5u
23158           Specify the SCORE5U of the target architecture.
23159
23160       -mscore7
23161           Specify the SCORE7 as the target architecture. This is the default.
23162
23163       -mscore7d
23164           Specify the SCORE7D as the target architecture.
23165
23166   SH Options
23167       These -m options are defined for the SH implementations:
23168
23169       -m1 Generate code for the SH1.
23170
23171       -m2 Generate code for the SH2.
23172
23173       -m2e
23174           Generate code for the SH2e.
23175
23176       -m2a-nofpu
23177           Generate code for the SH2a without FPU, or for a SH2a-FPU in such a
23178           way that the floating-point unit is not used.
23179
23180       -m2a-single-only
23181           Generate code for the SH2a-FPU, in such a way that no double-
23182           precision floating-point operations are used.
23183
23184       -m2a-single
23185           Generate code for the SH2a-FPU assuming the floating-point unit is
23186           in single-precision mode by default.
23187
23188       -m2a
23189           Generate code for the SH2a-FPU assuming the floating-point unit is
23190           in double-precision mode by default.
23191
23192       -m3 Generate code for the SH3.
23193
23194       -m3e
23195           Generate code for the SH3e.
23196
23197       -m4-nofpu
23198           Generate code for the SH4 without a floating-point unit.
23199
23200       -m4-single-only
23201           Generate code for the SH4 with a floating-point unit that only
23202           supports single-precision arithmetic.
23203
23204       -m4-single
23205           Generate code for the SH4 assuming the floating-point unit is in
23206           single-precision mode by default.
23207
23208       -m4 Generate code for the SH4.
23209
23210       -m4-100
23211           Generate code for SH4-100.
23212
23213       -m4-100-nofpu
23214           Generate code for SH4-100 in such a way that the floating-point
23215           unit is not used.
23216
23217       -m4-100-single
23218           Generate code for SH4-100 assuming the floating-point unit is in
23219           single-precision mode by default.
23220
23221       -m4-100-single-only
23222           Generate code for SH4-100 in such a way that no double-precision
23223           floating-point operations are used.
23224
23225       -m4-200
23226           Generate code for SH4-200.
23227
23228       -m4-200-nofpu
23229           Generate code for SH4-200 without in such a way that the floating-
23230           point unit is not used.
23231
23232       -m4-200-single
23233           Generate code for SH4-200 assuming the floating-point unit is in
23234           single-precision mode by default.
23235
23236       -m4-200-single-only
23237           Generate code for SH4-200 in such a way that no double-precision
23238           floating-point operations are used.
23239
23240       -m4-300
23241           Generate code for SH4-300.
23242
23243       -m4-300-nofpu
23244           Generate code for SH4-300 without in such a way that the floating-
23245           point unit is not used.
23246
23247       -m4-300-single
23248           Generate code for SH4-300 in such a way that no double-precision
23249           floating-point operations are used.
23250
23251       -m4-300-single-only
23252           Generate code for SH4-300 in such a way that no double-precision
23253           floating-point operations are used.
23254
23255       -m4-340
23256           Generate code for SH4-340 (no MMU, no FPU).
23257
23258       -m4-500
23259           Generate code for SH4-500 (no FPU).  Passes -isa=sh4-nofpu to the
23260           assembler.
23261
23262       -m4a-nofpu
23263           Generate code for the SH4al-dsp, or for a SH4a in such a way that
23264           the floating-point unit is not used.
23265
23266       -m4a-single-only
23267           Generate code for the SH4a, in such a way that no double-precision
23268           floating-point operations are used.
23269
23270       -m4a-single
23271           Generate code for the SH4a assuming the floating-point unit is in
23272           single-precision mode by default.
23273
23274       -m4a
23275           Generate code for the SH4a.
23276
23277       -m4al
23278           Same as -m4a-nofpu, except that it implicitly passes -dsp to the
23279           assembler.  GCC doesn't generate any DSP instructions at the
23280           moment.
23281
23282       -mb Compile code for the processor in big-endian mode.
23283
23284       -ml Compile code for the processor in little-endian mode.
23285
23286       -mdalign
23287           Align doubles at 64-bit boundaries.  Note that this changes the
23288           calling conventions, and thus some functions from the standard C
23289           library do not work unless you recompile it first with -mdalign.
23290
23291       -mrelax
23292           Shorten some address references at link time, when possible; uses
23293           the linker option -relax.
23294
23295       -mbigtable
23296           Use 32-bit offsets in "switch" tables.  The default is to use
23297           16-bit offsets.
23298
23299       -mbitops
23300           Enable the use of bit manipulation instructions on SH2A.
23301
23302       -mfmovd
23303           Enable the use of the instruction "fmovd".  Check -mdalign for
23304           alignment constraints.
23305
23306       -mrenesas
23307           Comply with the calling conventions defined by Renesas.
23308
23309       -mno-renesas
23310           Comply with the calling conventions defined for GCC before the
23311           Renesas conventions were available.  This option is the default for
23312           all targets of the SH toolchain.
23313
23314       -mnomacsave
23315           Mark the "MAC" register as call-clobbered, even if -mrenesas is
23316           given.
23317
23318       -mieee
23319       -mno-ieee
23320           Control the IEEE compliance of floating-point comparisons, which
23321           affects the handling of cases where the result of a comparison is
23322           unordered.  By default -mieee is implicitly enabled.  If
23323           -ffinite-math-only is enabled -mno-ieee is implicitly set, which
23324           results in faster floating-point greater-equal and less-equal
23325           comparisons.  The implicit settings can be overridden by specifying
23326           either -mieee or -mno-ieee.
23327
23328       -minline-ic_invalidate
23329           Inline code to invalidate instruction cache entries after setting
23330           up nested function trampolines.  This option has no effect if
23331           -musermode is in effect and the selected code generation option
23332           (e.g. -m4) does not allow the use of the "icbi" instruction.  If
23333           the selected code generation option does not allow the use of the
23334           "icbi" instruction, and -musermode is not in effect, the inlined
23335           code manipulates the instruction cache address array directly with
23336           an associative write.  This not only requires privileged mode at
23337           run time, but it also fails if the cache line had been mapped via
23338           the TLB and has become unmapped.
23339
23340       -misize
23341           Dump instruction size and location in the assembly code.
23342
23343       -mpadstruct
23344           This option is deprecated.  It pads structures to multiple of 4
23345           bytes, which is incompatible with the SH ABI.
23346
23347       -matomic-model=model
23348           Sets the model of atomic operations and additional parameters as a
23349           comma separated list.  For details on the atomic built-in functions
23350           see __atomic Builtins.  The following models and parameters are
23351           supported:
23352
23353           none
23354               Disable compiler generated atomic sequences and emit library
23355               calls for atomic operations.  This is the default if the target
23356               is not "sh*-*-linux*".
23357
23358           soft-gusa
23359               Generate GNU/Linux compatible gUSA software atomic sequences
23360               for the atomic built-in functions.  The generated atomic
23361               sequences require additional support from the
23362               interrupt/exception handling code of the system and are only
23363               suitable for SH3* and SH4* single-core systems.  This option is
23364               enabled by default when the target is "sh*-*-linux*" and SH3*
23365               or SH4*.  When the target is SH4A, this option also partially
23366               utilizes the hardware atomic instructions "movli.l" and
23367               "movco.l" to create more efficient code, unless strict is
23368               specified.
23369
23370           soft-tcb
23371               Generate software atomic sequences that use a variable in the
23372               thread control block.  This is a variation of the gUSA
23373               sequences which can also be used on SH1* and SH2* targets.  The
23374               generated atomic sequences require additional support from the
23375               interrupt/exception handling code of the system and are only
23376               suitable for single-core systems.  When using this model, the
23377               gbr-offset= parameter has to be specified as well.
23378
23379           soft-imask
23380               Generate software atomic sequences that temporarily disable
23381               interrupts by setting "SR.IMASK = 1111".  This model works only
23382               when the program runs in privileged mode and is only suitable
23383               for single-core systems.  Additional support from the
23384               interrupt/exception handling code of the system is not
23385               required.  This model is enabled by default when the target is
23386               "sh*-*-linux*" and SH1* or SH2*.
23387
23388           hard-llcs
23389               Generate hardware atomic sequences using the "movli.l" and
23390               "movco.l" instructions only.  This is only available on SH4A
23391               and is suitable for multi-core systems.  Since the hardware
23392               instructions support only 32 bit atomic variables access to 8
23393               or 16 bit variables is emulated with 32 bit accesses.  Code
23394               compiled with this option is also compatible with other
23395               software atomic model interrupt/exception handling systems if
23396               executed on an SH4A system.  Additional support from the
23397               interrupt/exception handling code of the system is not required
23398               for this model.
23399
23400           gbr-offset=
23401               This parameter specifies the offset in bytes of the variable in
23402               the thread control block structure that should be used by the
23403               generated atomic sequences when the soft-tcb model has been
23404               selected.  For other models this parameter is ignored.  The
23405               specified value must be an integer multiple of four and in the
23406               range 0-1020.
23407
23408           strict
23409               This parameter prevents mixed usage of multiple atomic models,
23410               even if they are compatible, and makes the compiler generate
23411               atomic sequences of the specified model only.
23412
23413       -mtas
23414           Generate the "tas.b" opcode for "__atomic_test_and_set".  Notice
23415           that depending on the particular hardware and software
23416           configuration this can degrade overall performance due to the
23417           operand cache line flushes that are implied by the "tas.b"
23418           instruction.  On multi-core SH4A processors the "tas.b" instruction
23419           must be used with caution since it can result in data corruption
23420           for certain cache configurations.
23421
23422       -mprefergot
23423           When generating position-independent code, emit function calls
23424           using the Global Offset Table instead of the Procedure Linkage
23425           Table.
23426
23427       -musermode
23428       -mno-usermode
23429           Don't allow (allow) the compiler generating privileged mode code.
23430           Specifying -musermode also implies -mno-inline-ic_invalidate if the
23431           inlined code would not work in user mode.  -musermode is the
23432           default when the target is "sh*-*-linux*".  If the target is SH1*
23433           or SH2* -musermode has no effect, since there is no user mode.
23434
23435       -multcost=number
23436           Set the cost to assume for a multiply insn.
23437
23438       -mdiv=strategy
23439           Set the division strategy to be used for integer division
23440           operations.  strategy can be one of:
23441
23442           call-div1
23443               Calls a library function that uses the single-step division
23444               instruction "div1" to perform the operation.  Division by zero
23445               calculates an unspecified result and does not trap.  This is
23446               the default except for SH4, SH2A and SHcompact.
23447
23448           call-fp
23449               Calls a library function that performs the operation in double
23450               precision floating point.  Division by zero causes a floating-
23451               point exception.  This is the default for SHcompact with FPU.
23452               Specifying this for targets that do not have a double precision
23453               FPU defaults to "call-div1".
23454
23455           call-table
23456               Calls a library function that uses a lookup table for small
23457               divisors and the "div1" instruction with case distinction for
23458               larger divisors.  Division by zero calculates an unspecified
23459               result and does not trap.  This is the default for SH4.
23460               Specifying this for targets that do not have dynamic shift
23461               instructions defaults to "call-div1".
23462
23463           When a division strategy has not been specified the default
23464           strategy is selected based on the current target.  For SH2A the
23465           default strategy is to use the "divs" and "divu" instructions
23466           instead of library function calls.
23467
23468       -maccumulate-outgoing-args
23469           Reserve space once for outgoing arguments in the function prologue
23470           rather than around each call.  Generally beneficial for performance
23471           and size.  Also needed for unwinding to avoid changing the stack
23472           frame around conditional code.
23473
23474       -mdivsi3_libfunc=name
23475           Set the name of the library function used for 32-bit signed
23476           division to name.  This only affects the name used in the call
23477           division strategies, and the compiler still expects the same sets
23478           of input/output/clobbered registers as if this option were not
23479           present.
23480
23481       -mfixed-range=register-range
23482           Generate code treating the given register range as fixed registers.
23483           A fixed register is one that the register allocator cannot use.
23484           This is useful when compiling kernel code.  A register range is
23485           specified as two registers separated by a dash.  Multiple register
23486           ranges can be specified separated by a comma.
23487
23488       -mbranch-cost=num
23489           Assume num to be the cost for a branch instruction.  Higher numbers
23490           make the compiler try to generate more branch-free code if
23491           possible.  If not specified the value is selected depending on the
23492           processor type that is being compiled for.
23493
23494       -mzdcbranch
23495       -mno-zdcbranch
23496           Assume (do not assume) that zero displacement conditional branch
23497           instructions "bt" and "bf" are fast.  If -mzdcbranch is specified,
23498           the compiler prefers zero displacement branch code sequences.  This
23499           is enabled by default when generating code for SH4 and SH4A.  It
23500           can be explicitly disabled by specifying -mno-zdcbranch.
23501
23502       -mcbranch-force-delay-slot
23503           Force the usage of delay slots for conditional branches, which
23504           stuffs the delay slot with a "nop" if a suitable instruction cannot
23505           be found.  By default this option is disabled.  It can be enabled
23506           to work around hardware bugs as found in the original SH7055.
23507
23508       -mfused-madd
23509       -mno-fused-madd
23510           Generate code that uses (does not use) the floating-point multiply
23511           and accumulate instructions.  These instructions are generated by
23512           default if hardware floating point is used.  The machine-dependent
23513           -mfused-madd option is now mapped to the machine-independent
23514           -ffp-contract=fast option, and -mno-fused-madd is mapped to
23515           -ffp-contract=off.
23516
23517       -mfsca
23518       -mno-fsca
23519           Allow or disallow the compiler to emit the "fsca" instruction for
23520           sine and cosine approximations.  The option -mfsca must be used in
23521           combination with -funsafe-math-optimizations.  It is enabled by
23522           default when generating code for SH4A.  Using -mno-fsca disables
23523           sine and cosine approximations even if -funsafe-math-optimizations
23524           is in effect.
23525
23526       -mfsrra
23527       -mno-fsrra
23528           Allow or disallow the compiler to emit the "fsrra" instruction for
23529           reciprocal square root approximations.  The option -mfsrra must be
23530           used in combination with -funsafe-math-optimizations and
23531           -ffinite-math-only.  It is enabled by default when generating code
23532           for SH4A.  Using -mno-fsrra disables reciprocal square root
23533           approximations even if -funsafe-math-optimizations and
23534           -ffinite-math-only are in effect.
23535
23536       -mpretend-cmove
23537           Prefer zero-displacement conditional branches for conditional move
23538           instruction patterns.  This can result in faster code on the SH4
23539           processor.
23540
23541       -mfdpic
23542           Generate code using the FDPIC ABI.
23543
23544   Solaris 2 Options
23545       These -m options are supported on Solaris 2:
23546
23547       -mclear-hwcap
23548           -mclear-hwcap tells the compiler to remove the hardware
23549           capabilities generated by the Solaris assembler.  This is only
23550           necessary when object files use ISA extensions not supported by the
23551           current machine, but check at runtime whether or not to use them.
23552
23553       -mimpure-text
23554           -mimpure-text, used in addition to -shared, tells the compiler to
23555           not pass -z text to the linker when linking a shared object.  Using
23556           this option, you can link position-dependent code into a shared
23557           object.
23558
23559           -mimpure-text suppresses the "relocations remain against
23560           allocatable but non-writable sections" linker error message.
23561           However, the necessary relocations trigger copy-on-write, and the
23562           shared object is not actually shared across processes.  Instead of
23563           using -mimpure-text, you should compile all source code with -fpic
23564           or -fPIC.
23565
23566       These switches are supported in addition to the above on Solaris 2:
23567
23568       -pthreads
23569           This is a synonym for -pthread.
23570
23571   SPARC Options
23572       These -m options are supported on the SPARC:
23573
23574       -mno-app-regs
23575       -mapp-regs
23576           Specify -mapp-regs to generate output using the global registers 2
23577           through 4, which the SPARC SVR4 ABI reserves for applications.
23578           Like the global register 1, each global register 2 through 4 is
23579           then treated as an allocable register that is clobbered by function
23580           calls.  This is the default.
23581
23582           To be fully SVR4 ABI-compliant at the cost of some performance
23583           loss, specify -mno-app-regs.  You should compile libraries and
23584           system software with this option.
23585
23586       -mflat
23587       -mno-flat
23588           With -mflat, the compiler does not generate save/restore
23589           instructions and uses a "flat" or single register window model.
23590           This model is compatible with the regular register window model.
23591           The local registers and the input registers (0--5) are still
23592           treated as "call-saved" registers and are saved on the stack as
23593           needed.
23594
23595           With -mno-flat (the default), the compiler generates save/restore
23596           instructions (except for leaf functions).  This is the normal
23597           operating mode.
23598
23599       -mfpu
23600       -mhard-float
23601           Generate output containing floating-point instructions.  This is
23602           the default.
23603
23604       -mno-fpu
23605       -msoft-float
23606           Generate output containing library calls for floating point.
23607           Warning: the requisite libraries are not available for all SPARC
23608           targets.  Normally the facilities of the machine's usual C compiler
23609           are used, but this cannot be done directly in cross-compilation.
23610           You must make your own arrangements to provide suitable library
23611           functions for cross-compilation.  The embedded targets sparc-*-aout
23612           and sparclite-*-* do provide software floating-point support.
23613
23614           -msoft-float changes the calling convention in the output file;
23615           therefore, it is only useful if you compile all of a program with
23616           this option.  In particular, you need to compile libgcc.a, the
23617           library that comes with GCC, with -msoft-float in order for this to
23618           work.
23619
23620       -mhard-quad-float
23621           Generate output containing quad-word (long double) floating-point
23622           instructions.
23623
23624       -msoft-quad-float
23625           Generate output containing library calls for quad-word (long
23626           double) floating-point instructions.  The functions called are
23627           those specified in the SPARC ABI.  This is the default.
23628
23629           As of this writing, there are no SPARC implementations that have
23630           hardware support for the quad-word floating-point instructions.
23631           They all invoke a trap handler for one of these instructions, and
23632           then the trap handler emulates the effect of the instruction.
23633           Because of the trap handler overhead, this is much slower than
23634           calling the ABI library routines.  Thus the -msoft-quad-float
23635           option is the default.
23636
23637       -mno-unaligned-doubles
23638       -munaligned-doubles
23639           Assume that doubles have 8-byte alignment.  This is the default.
23640
23641           With -munaligned-doubles, GCC assumes that doubles have 8-byte
23642           alignment only if they are contained in another type, or if they
23643           have an absolute address.  Otherwise, it assumes they have 4-byte
23644           alignment.  Specifying this option avoids some rare compatibility
23645           problems with code generated by other compilers.  It is not the
23646           default because it results in a performance loss, especially for
23647           floating-point code.
23648
23649       -muser-mode
23650       -mno-user-mode
23651           Do not generate code that can only run in supervisor mode.  This is
23652           relevant only for the "casa" instruction emitted for the LEON3
23653           processor.  This is the default.
23654
23655       -mfaster-structs
23656       -mno-faster-structs
23657           With -mfaster-structs, the compiler assumes that structures should
23658           have 8-byte alignment.  This enables the use of pairs of "ldd" and
23659           "std" instructions for copies in structure assignment, in place of
23660           twice as many "ld" and "st" pairs.  However, the use of this
23661           changed alignment directly violates the SPARC ABI.  Thus, it's
23662           intended only for use on targets where the developer acknowledges
23663           that their resulting code is not directly in line with the rules of
23664           the ABI.
23665
23666       -mstd-struct-return
23667       -mno-std-struct-return
23668           With -mstd-struct-return, the compiler generates checking code in
23669           functions returning structures or unions to detect size mismatches
23670           between the two sides of function calls, as per the 32-bit ABI.
23671
23672           The default is -mno-std-struct-return.  This option has no effect
23673           in 64-bit mode.
23674
23675       -mlra
23676       -mno-lra
23677           Enable Local Register Allocation.  This is the default for SPARC
23678           since GCC 7 so -mno-lra needs to be passed to get old Reload.
23679
23680       -mcpu=cpu_type
23681           Set the instruction set, register set, and instruction scheduling
23682           parameters for machine type cpu_type.  Supported values for
23683           cpu_type are v7, cypress, v8, supersparc, hypersparc, leon, leon3,
23684           leon3v7, sparclite, f930, f934, sparclite86x, sparclet, tsc701, v9,
23685           ultrasparc, ultrasparc3, niagara, niagara2, niagara3, niagara4,
23686           niagara7 and m8.
23687
23688           Native Solaris and GNU/Linux toolchains also support the value
23689           native, which selects the best architecture option for the host
23690           processor.  -mcpu=native has no effect if GCC does not recognize
23691           the processor.
23692
23693           Default instruction scheduling parameters are used for values that
23694           select an architecture and not an implementation.  These are v7,
23695           v8, sparclite, sparclet, v9.
23696
23697           Here is a list of each supported architecture and their supported
23698           implementations.
23699
23700           v7  cypress, leon3v7
23701
23702           v8  supersparc, hypersparc, leon, leon3
23703
23704           sparclite
23705               f930, f934, sparclite86x
23706
23707           sparclet
23708               tsc701
23709
23710           v9  ultrasparc, ultrasparc3, niagara, niagara2, niagara3, niagara4,
23711               niagara7, m8
23712
23713           By default (unless configured otherwise), GCC generates code for
23714           the V7 variant of the SPARC architecture.  With -mcpu=cypress, the
23715           compiler additionally optimizes it for the Cypress CY7C602 chip, as
23716           used in the SPARCStation/SPARCServer 3xx series.  This is also
23717           appropriate for the older SPARCStation 1, 2, IPX etc.
23718
23719           With -mcpu=v8, GCC generates code for the V8 variant of the SPARC
23720           architecture.  The only difference from V7 code is that the
23721           compiler emits the integer multiply and integer divide instructions
23722           which exist in SPARC-V8 but not in SPARC-V7.  With
23723           -mcpu=supersparc, the compiler additionally optimizes it for the
23724           SuperSPARC chip, as used in the SPARCStation 10, 1000 and 2000
23725           series.
23726
23727           With -mcpu=sparclite, GCC generates code for the SPARClite variant
23728           of the SPARC architecture.  This adds the integer multiply, integer
23729           divide step and scan ("ffs") instructions which exist in SPARClite
23730           but not in SPARC-V7.  With -mcpu=f930, the compiler additionally
23731           optimizes it for the Fujitsu MB86930 chip, which is the original
23732           SPARClite, with no FPU.  With -mcpu=f934, the compiler additionally
23733           optimizes it for the Fujitsu MB86934 chip, which is the more recent
23734           SPARClite with FPU.
23735
23736           With -mcpu=sparclet, GCC generates code for the SPARClet variant of
23737           the SPARC architecture.  This adds the integer multiply,
23738           multiply/accumulate, integer divide step and scan ("ffs")
23739           instructions which exist in SPARClet but not in SPARC-V7.  With
23740           -mcpu=tsc701, the compiler additionally optimizes it for the TEMIC
23741           SPARClet chip.
23742
23743           With -mcpu=v9, GCC generates code for the V9 variant of the SPARC
23744           architecture.  This adds 64-bit integer and floating-point move
23745           instructions, 3 additional floating-point condition code registers
23746           and conditional move instructions.  With -mcpu=ultrasparc, the
23747           compiler additionally optimizes it for the Sun UltraSPARC I/II/IIi
23748           chips.  With -mcpu=ultrasparc3, the compiler additionally optimizes
23749           it for the Sun UltraSPARC III/III+/IIIi/IIIi+/IV/IV+ chips.  With
23750           -mcpu=niagara, the compiler additionally optimizes it for Sun
23751           UltraSPARC T1 chips.  With -mcpu=niagara2, the compiler
23752           additionally optimizes it for Sun UltraSPARC T2 chips. With
23753           -mcpu=niagara3, the compiler additionally optimizes it for Sun
23754           UltraSPARC T3 chips.  With -mcpu=niagara4, the compiler
23755           additionally optimizes it for Sun UltraSPARC T4 chips.  With
23756           -mcpu=niagara7, the compiler additionally optimizes it for Oracle
23757           SPARC M7 chips.  With -mcpu=m8, the compiler additionally optimizes
23758           it for Oracle M8 chips.
23759
23760       -mtune=cpu_type
23761           Set the instruction scheduling parameters for machine type
23762           cpu_type, but do not set the instruction set or register set that
23763           the option -mcpu=cpu_type does.
23764
23765           The same values for -mcpu=cpu_type can be used for -mtune=cpu_type,
23766           but the only useful values are those that select a particular CPU
23767           implementation.  Those are cypress, supersparc, hypersparc, leon,
23768           leon3, leon3v7, f930, f934, sparclite86x, tsc701, ultrasparc,
23769           ultrasparc3, niagara, niagara2, niagara3, niagara4, niagara7 and
23770           m8.  With native Solaris and GNU/Linux toolchains, native can also
23771           be used.
23772
23773       -mv8plus
23774       -mno-v8plus
23775           With -mv8plus, GCC generates code for the SPARC-V8+ ABI.  The
23776           difference from the V8 ABI is that the global and out registers are
23777           considered 64 bits wide.  This is enabled by default on Solaris in
23778           32-bit mode for all SPARC-V9 processors.
23779
23780       -mvis
23781       -mno-vis
23782           With -mvis, GCC generates code that takes advantage of the
23783           UltraSPARC Visual Instruction Set extensions.  The default is
23784           -mno-vis.
23785
23786       -mvis2
23787       -mno-vis2
23788           With -mvis2, GCC generates code that takes advantage of version 2.0
23789           of the UltraSPARC Visual Instruction Set extensions.  The default
23790           is -mvis2 when targeting a cpu that supports such instructions,
23791           such as UltraSPARC-III and later.  Setting -mvis2 also sets -mvis.
23792
23793       -mvis3
23794       -mno-vis3
23795           With -mvis3, GCC generates code that takes advantage of version 3.0
23796           of the UltraSPARC Visual Instruction Set extensions.  The default
23797           is -mvis3 when targeting a cpu that supports such instructions,
23798           such as niagara-3 and later.  Setting -mvis3 also sets -mvis2 and
23799           -mvis.
23800
23801       -mvis4
23802       -mno-vis4
23803           With -mvis4, GCC generates code that takes advantage of version 4.0
23804           of the UltraSPARC Visual Instruction Set extensions.  The default
23805           is -mvis4 when targeting a cpu that supports such instructions,
23806           such as niagara-7 and later.  Setting -mvis4 also sets -mvis3,
23807           -mvis2 and -mvis.
23808
23809       -mvis4b
23810       -mno-vis4b
23811           With -mvis4b, GCC generates code that takes advantage of version
23812           4.0 of the UltraSPARC Visual Instruction Set extensions, plus the
23813           additional VIS instructions introduced in the Oracle SPARC
23814           Architecture 2017.  The default is -mvis4b when targeting a cpu
23815           that supports such instructions, such as m8 and later.  Setting
23816           -mvis4b also sets -mvis4, -mvis3, -mvis2 and -mvis.
23817
23818       -mcbcond
23819       -mno-cbcond
23820           With -mcbcond, GCC generates code that takes advantage of the
23821           UltraSPARC Compare-and-Branch-on-Condition instructions.  The
23822           default is -mcbcond when targeting a CPU that supports such
23823           instructions, such as Niagara-4 and later.
23824
23825       -mfmaf
23826       -mno-fmaf
23827           With -mfmaf, GCC generates code that takes advantage of the
23828           UltraSPARC Fused Multiply-Add Floating-point instructions.  The
23829           default is -mfmaf when targeting a CPU that supports such
23830           instructions, such as Niagara-3 and later.
23831
23832       -mfsmuld
23833       -mno-fsmuld
23834           With -mfsmuld, GCC generates code that takes advantage of the
23835           Floating-point Multiply Single to Double (FsMULd) instruction.  The
23836           default is -mfsmuld when targeting a CPU supporting the
23837           architecture versions V8 or V9 with FPU except -mcpu=leon.
23838
23839       -mpopc
23840       -mno-popc
23841           With -mpopc, GCC generates code that takes advantage of the
23842           UltraSPARC Population Count instruction.  The default is -mpopc
23843           when targeting a CPU that supports such an instruction, such as
23844           Niagara-2 and later.
23845
23846       -msubxc
23847       -mno-subxc
23848           With -msubxc, GCC generates code that takes advantage of the
23849           UltraSPARC Subtract-Extended-with-Carry instruction.  The default
23850           is -msubxc when targeting a CPU that supports such an instruction,
23851           such as Niagara-7 and later.
23852
23853       -mfix-at697f
23854           Enable the documented workaround for the single erratum of the
23855           Atmel AT697F processor (which corresponds to erratum #13 of the
23856           AT697E processor).
23857
23858       -mfix-ut699
23859           Enable the documented workarounds for the floating-point errata and
23860           the data cache nullify errata of the UT699 processor.
23861
23862       -mfix-ut700
23863           Enable the documented workaround for the back-to-back store errata
23864           of the UT699E/UT700 processor.
23865
23866       -mfix-gr712rc
23867           Enable the documented workaround for the back-to-back store errata
23868           of the GR712RC processor.
23869
23870       These -m options are supported in addition to the above on SPARC-V9
23871       processors in 64-bit environments:
23872
23873       -m32
23874       -m64
23875           Generate code for a 32-bit or 64-bit environment.  The 32-bit
23876           environment sets int, long and pointer to 32 bits.  The 64-bit
23877           environment sets int to 32 bits and long and pointer to 64 bits.
23878
23879       -mcmodel=which
23880           Set the code model to one of
23881
23882           medlow
23883               The Medium/Low code model: 64-bit addresses, programs must be
23884               linked in the low 32 bits of memory.  Programs can be
23885               statically or dynamically linked.
23886
23887           medmid
23888               The Medium/Middle code model: 64-bit addresses, programs must
23889               be linked in the low 44 bits of memory, the text and data
23890               segments must be less than 2GB in size and the data segment
23891               must be located within 2GB of the text segment.
23892
23893           medany
23894               The Medium/Anywhere code model: 64-bit addresses, programs may
23895               be linked anywhere in memory, the text and data segments must
23896               be less than 2GB in size and the data segment must be located
23897               within 2GB of the text segment.
23898
23899           embmedany
23900               The Medium/Anywhere code model for embedded systems: 64-bit
23901               addresses, the text and data segments must be less than 2GB in
23902               size, both starting anywhere in memory (determined at link
23903               time).  The global register %g4 points to the base of the data
23904               segment.  Programs are statically linked and PIC is not
23905               supported.
23906
23907       -mmemory-model=mem-model
23908           Set the memory model in force on the processor to one of
23909
23910           default
23911               The default memory model for the processor and operating
23912               system.
23913
23914           rmo Relaxed Memory Order
23915
23916           pso Partial Store Order
23917
23918           tso Total Store Order
23919
23920           sc  Sequential Consistency
23921
23922           These memory models are formally defined in Appendix D of the
23923           SPARC-V9 architecture manual, as set in the processor's "PSTATE.MM"
23924           field.
23925
23926       -mstack-bias
23927       -mno-stack-bias
23928           With -mstack-bias, GCC assumes that the stack pointer, and frame
23929           pointer if present, are offset by -2047 which must be added back
23930           when making stack frame references.  This is the default in 64-bit
23931           mode.  Otherwise, assume no such offset is present.
23932
23933   Options for System V
23934       These additional options are available on System V Release 4 for
23935       compatibility with other compilers on those systems:
23936
23937       -G  Create a shared object.  It is recommended that -symbolic or
23938           -shared be used instead.
23939
23940       -Qy Identify the versions of each tool used by the compiler, in a
23941           ".ident" assembler directive in the output.
23942
23943       -Qn Refrain from adding ".ident" directives to the output file (this is
23944           the default).
23945
23946       -YP,dirs
23947           Search the directories dirs, and no others, for libraries specified
23948           with -l.
23949
23950       -Ym,dir
23951           Look in the directory dir to find the M4 preprocessor.  The
23952           assembler uses this option.
23953
23954   TILE-Gx Options
23955       These -m options are supported on the TILE-Gx:
23956
23957       -mcmodel=small
23958           Generate code for the small model.  The distance for direct calls
23959           is limited to 500M in either direction.  PC-relative addresses are
23960           32 bits.  Absolute addresses support the full address range.
23961
23962       -mcmodel=large
23963           Generate code for the large model.  There is no limitation on call
23964           distance, pc-relative addresses, or absolute addresses.
23965
23966       -mcpu=name
23967           Selects the type of CPU to be targeted.  Currently the only
23968           supported type is tilegx.
23969
23970       -m32
23971       -m64
23972           Generate code for a 32-bit or 64-bit environment.  The 32-bit
23973           environment sets int, long, and pointer to 32 bits.  The 64-bit
23974           environment sets int to 32 bits and long and pointer to 64 bits.
23975
23976       -mbig-endian
23977       -mlittle-endian
23978           Generate code in big/little endian mode, respectively.
23979
23980   TILEPro Options
23981       These -m options are supported on the TILEPro:
23982
23983       -mcpu=name
23984           Selects the type of CPU to be targeted.  Currently the only
23985           supported type is tilepro.
23986
23987       -m32
23988           Generate code for a 32-bit environment, which sets int, long, and
23989           pointer to 32 bits.  This is the only supported behavior so the
23990           flag is essentially ignored.
23991
23992   V850 Options
23993       These -m options are defined for V850 implementations:
23994
23995       -mlong-calls
23996       -mno-long-calls
23997           Treat all calls as being far away (near).  If calls are assumed to
23998           be far away, the compiler always loads the function's address into
23999           a register, and calls indirect through the pointer.
24000
24001       -mno-ep
24002       -mep
24003           Do not optimize (do optimize) basic blocks that use the same index
24004           pointer 4 or more times to copy pointer into the "ep" register, and
24005           use the shorter "sld" and "sst" instructions.  The -mep option is
24006           on by default if you optimize.
24007
24008       -mno-prolog-function
24009       -mprolog-function
24010           Do not use (do use) external functions to save and restore
24011           registers at the prologue and epilogue of a function.  The external
24012           functions are slower, but use less code space if more than one
24013           function saves the same number of registers.  The -mprolog-function
24014           option is on by default if you optimize.
24015
24016       -mspace
24017           Try to make the code as small as possible.  At present, this just
24018           turns on the -mep and -mprolog-function options.
24019
24020       -mtda=n
24021           Put static or global variables whose size is n bytes or less into
24022           the tiny data area that register "ep" points to.  The tiny data
24023           area can hold up to 256 bytes in total (128 bytes for byte
24024           references).
24025
24026       -msda=n
24027           Put static or global variables whose size is n bytes or less into
24028           the small data area that register "gp" points to.  The small data
24029           area can hold up to 64 kilobytes.
24030
24031       -mzda=n
24032           Put static or global variables whose size is n bytes or less into
24033           the first 32 kilobytes of memory.
24034
24035       -mv850
24036           Specify that the target processor is the V850.
24037
24038       -mv850e3v5
24039           Specify that the target processor is the V850E3V5.  The
24040           preprocessor constant "__v850e3v5__" is defined if this option is
24041           used.
24042
24043       -mv850e2v4
24044           Specify that the target processor is the V850E3V5.  This is an
24045           alias for the -mv850e3v5 option.
24046
24047       -mv850e2v3
24048           Specify that the target processor is the V850E2V3.  The
24049           preprocessor constant "__v850e2v3__" is defined if this option is
24050           used.
24051
24052       -mv850e2
24053           Specify that the target processor is the V850E2.  The preprocessor
24054           constant "__v850e2__" is defined if this option is used.
24055
24056       -mv850e1
24057           Specify that the target processor is the V850E1.  The preprocessor
24058           constants "__v850e1__" and "__v850e__" are defined if this option
24059           is used.
24060
24061       -mv850es
24062           Specify that the target processor is the V850ES.  This is an alias
24063           for the -mv850e1 option.
24064
24065       -mv850e
24066           Specify that the target processor is the V850E.  The preprocessor
24067           constant "__v850e__" is defined if this option is used.
24068
24069           If neither -mv850 nor -mv850e nor -mv850e1 nor -mv850e2 nor
24070           -mv850e2v3 nor -mv850e3v5 are defined then a default target
24071           processor is chosen and the relevant __v850*__ preprocessor
24072           constant is defined.
24073
24074           The preprocessor constants "__v850" and "__v851__" are always
24075           defined, regardless of which processor variant is the target.
24076
24077       -mdisable-callt
24078       -mno-disable-callt
24079           This option suppresses generation of the "CALLT" instruction for
24080           the v850e, v850e1, v850e2, v850e2v3 and v850e3v5 flavors of the
24081           v850 architecture.
24082
24083           This option is enabled by default when the RH850 ABI is in use (see
24084           -mrh850-abi), and disabled by default when the GCC ABI is in use.
24085           If "CALLT" instructions are being generated then the C preprocessor
24086           symbol "__V850_CALLT__" is defined.
24087
24088       -mrelax
24089       -mno-relax
24090           Pass on (or do not pass on) the -mrelax command-line option to the
24091           assembler.
24092
24093       -mlong-jumps
24094       -mno-long-jumps
24095           Disable (or re-enable) the generation of PC-relative jump
24096           instructions.
24097
24098       -msoft-float
24099       -mhard-float
24100           Disable (or re-enable) the generation of hardware floating point
24101           instructions.  This option is only significant when the target
24102           architecture is V850E2V3 or higher.  If hardware floating point
24103           instructions are being generated then the C preprocessor symbol
24104           "__FPU_OK__" is defined, otherwise the symbol "__NO_FPU__" is
24105           defined.
24106
24107       -mloop
24108           Enables the use of the e3v5 LOOP instruction.  The use of this
24109           instruction is not enabled by default when the e3v5 architecture is
24110           selected because its use is still experimental.
24111
24112       -mrh850-abi
24113       -mghs
24114           Enables support for the RH850 version of the V850 ABI.  This is the
24115           default.  With this version of the ABI the following rules apply:
24116
24117           *   Integer sized structures and unions are returned via a memory
24118               pointer rather than a register.
24119
24120           *   Large structures and unions (more than 8 bytes in size) are
24121               passed by value.
24122
24123           *   Functions are aligned to 16-bit boundaries.
24124
24125           *   The -m8byte-align command-line option is supported.
24126
24127           *   The -mdisable-callt command-line option is enabled by default.
24128               The -mno-disable-callt command-line option is not supported.
24129
24130           When this version of the ABI is enabled the C preprocessor symbol
24131           "__V850_RH850_ABI__" is defined.
24132
24133       -mgcc-abi
24134           Enables support for the old GCC version of the V850 ABI.  With this
24135           version of the ABI the following rules apply:
24136
24137           *   Integer sized structures and unions are returned in register
24138               "r10".
24139
24140           *   Large structures and unions (more than 8 bytes in size) are
24141               passed by reference.
24142
24143           *   Functions are aligned to 32-bit boundaries, unless optimizing
24144               for size.
24145
24146           *   The -m8byte-align command-line option is not supported.
24147
24148           *   The -mdisable-callt command-line option is supported but not
24149               enabled by default.
24150
24151           When this version of the ABI is enabled the C preprocessor symbol
24152           "__V850_GCC_ABI__" is defined.
24153
24154       -m8byte-align
24155       -mno-8byte-align
24156           Enables support for "double" and "long long" types to be aligned on
24157           8-byte boundaries.  The default is to restrict the alignment of all
24158           objects to at most 4-bytes.  When -m8byte-align is in effect the C
24159           preprocessor symbol "__V850_8BYTE_ALIGN__" is defined.
24160
24161       -mbig-switch
24162           Generate code suitable for big switch tables.  Use this option only
24163           if the assembler/linker complain about out of range branches within
24164           a switch table.
24165
24166       -mapp-regs
24167           This option causes r2 and r5 to be used in the code generated by
24168           the compiler.  This setting is the default.
24169
24170       -mno-app-regs
24171           This option causes r2 and r5 to be treated as fixed registers.
24172
24173   VAX Options
24174       These -m options are defined for the VAX:
24175
24176       -munix
24177           Do not output certain jump instructions ("aobleq" and so on) that
24178           the Unix assembler for the VAX cannot handle across long ranges.
24179
24180       -mgnu
24181           Do output those jump instructions, on the assumption that the GNU
24182           assembler is being used.
24183
24184       -mg Output code for G-format floating-point numbers instead of
24185           D-format.
24186
24187   Visium Options
24188       -mdebug
24189           A program which performs file I/O and is destined to run on an MCM
24190           target should be linked with this option.  It causes the libraries
24191           libc.a and libdebug.a to be linked.  The program should be run on
24192           the target under the control of the GDB remote debugging stub.
24193
24194       -msim
24195           A program which performs file I/O and is destined to run on the
24196           simulator should be linked with option.  This causes libraries
24197           libc.a and libsim.a to be linked.
24198
24199       -mfpu
24200       -mhard-float
24201           Generate code containing floating-point instructions.  This is the
24202           default.
24203
24204       -mno-fpu
24205       -msoft-float
24206           Generate code containing library calls for floating-point.
24207
24208           -msoft-float changes the calling convention in the output file;
24209           therefore, it is only useful if you compile all of a program with
24210           this option.  In particular, you need to compile libgcc.a, the
24211           library that comes with GCC, with -msoft-float in order for this to
24212           work.
24213
24214       -mcpu=cpu_type
24215           Set the instruction set, register set, and instruction scheduling
24216           parameters for machine type cpu_type.  Supported values for
24217           cpu_type are mcm, gr5 and gr6.
24218
24219           mcm is a synonym of gr5 present for backward compatibility.
24220
24221           By default (unless configured otherwise), GCC generates code for
24222           the GR5 variant of the Visium architecture.
24223
24224           With -mcpu=gr6, GCC generates code for the GR6 variant of the
24225           Visium architecture.  The only difference from GR5 code is that the
24226           compiler will generate block move instructions.
24227
24228       -mtune=cpu_type
24229           Set the instruction scheduling parameters for machine type
24230           cpu_type, but do not set the instruction set or register set that
24231           the option -mcpu=cpu_type would.
24232
24233       -msv-mode
24234           Generate code for the supervisor mode, where there are no
24235           restrictions on the access to general registers.  This is the
24236           default.
24237
24238       -muser-mode
24239           Generate code for the user mode, where the access to some general
24240           registers is forbidden: on the GR5, registers r24 to r31 cannot be
24241           accessed in this mode; on the GR6, only registers r29 to r31 are
24242           affected.
24243
24244   VMS Options
24245       These -m options are defined for the VMS implementations:
24246
24247       -mvms-return-codes
24248           Return VMS condition codes from "main". The default is to return
24249           POSIX-style condition (e.g. error) codes.
24250
24251       -mdebug-main=prefix
24252           Flag the first routine whose name starts with prefix as the main
24253           routine for the debugger.
24254
24255       -mmalloc64
24256           Default to 64-bit memory allocation routines.
24257
24258       -mpointer-size=size
24259           Set the default size of pointers. Possible options for size are 32
24260           or short for 32 bit pointers, 64 or long for 64 bit pointers, and
24261           no for supporting only 32 bit pointers.  The later option disables
24262           "pragma pointer_size".
24263
24264   VxWorks Options
24265       The options in this section are defined for all VxWorks targets.
24266       Options specific to the target hardware are listed with the other
24267       options for that target.
24268
24269       -mrtp
24270           GCC can generate code for both VxWorks kernels and real time
24271           processes (RTPs).  This option switches from the former to the
24272           latter.  It also defines the preprocessor macro "__RTP__".
24273
24274       -non-static
24275           Link an RTP executable against shared libraries rather than static
24276           libraries.  The options -static and -shared can also be used for
24277           RTPs; -static is the default.
24278
24279       -Bstatic
24280       -Bdynamic
24281           These options are passed down to the linker.  They are defined for
24282           compatibility with Diab.
24283
24284       -Xbind-lazy
24285           Enable lazy binding of function calls.  This option is equivalent
24286           to -Wl,-z,now and is defined for compatibility with Diab.
24287
24288       -Xbind-now
24289           Disable lazy binding of function calls.  This option is the default
24290           and is defined for compatibility with Diab.
24291
24292   x86 Options
24293       These -m options are defined for the x86 family of computers.
24294
24295       -march=cpu-type
24296           Generate instructions for the machine type cpu-type.  In contrast
24297           to -mtune=cpu-type, which merely tunes the generated code for the
24298           specified cpu-type, -march=cpu-type allows GCC to generate code
24299           that may not run at all on processors other than the one indicated.
24300           Specifying -march=cpu-type implies -mtune=cpu-type.
24301
24302           The choices for cpu-type are:
24303
24304           native
24305               This selects the CPU to generate code for at compilation time
24306               by determining the processor type of the compiling machine.
24307               Using -march=native enables all instruction subsets supported
24308               by the local machine (hence the result might not run on
24309               different machines).  Using -mtune=native produces code
24310               optimized for the local machine under the constraints of the
24311               selected instruction set.
24312
24313           x86-64
24314               A generic CPU with 64-bit extensions.
24315
24316           i386
24317               Original Intel i386 CPU.
24318
24319           i486
24320               Intel i486 CPU.  (No scheduling is implemented for this chip.)
24321
24322           i586
24323           pentium
24324               Intel Pentium CPU with no MMX support.
24325
24326           lakemont
24327               Intel Lakemont MCU, based on Intel Pentium CPU.
24328
24329           pentium-mmx
24330               Intel Pentium MMX CPU, based on Pentium core with MMX
24331               instruction set support.
24332
24333           pentiumpro
24334               Intel Pentium Pro CPU.
24335
24336           i686
24337               When used with -march, the Pentium Pro instruction set is used,
24338               so the code runs on all i686 family chips.  When used with
24339               -mtune, it has the same meaning as generic.
24340
24341           pentium2
24342               Intel Pentium II CPU, based on Pentium Pro core with MMX
24343               instruction set support.
24344
24345           pentium3
24346           pentium3m
24347               Intel Pentium III CPU, based on Pentium Pro core with MMX and
24348               SSE instruction set support.
24349
24350           pentium-m
24351               Intel Pentium M; low-power version of Intel Pentium III CPU
24352               with MMX, SSE and SSE2 instruction set support.  Used by
24353               Centrino notebooks.
24354
24355           pentium4
24356           pentium4m
24357               Intel Pentium 4 CPU with MMX, SSE and SSE2 instruction set
24358               support.
24359
24360           prescott
24361               Improved version of Intel Pentium 4 CPU with MMX, SSE, SSE2 and
24362               SSE3 instruction set support.
24363
24364           nocona
24365               Improved version of Intel Pentium 4 CPU with 64-bit extensions,
24366               MMX, SSE, SSE2 and SSE3 instruction set support.
24367
24368           core2
24369               Intel Core 2 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3
24370               and SSSE3 instruction set support.
24371
24372           nehalem
24373               Intel Nehalem CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3,
24374               SSSE3, SSE4.1, SSE4.2 and POPCNT instruction set support.
24375
24376           westmere
24377               Intel Westmere CPU with 64-bit extensions, MMX, SSE, SSE2,
24378               SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES and PCLMUL instruction
24379               set support.
24380
24381           sandybridge
24382               Intel Sandy Bridge CPU with 64-bit extensions, MMX, SSE, SSE2,
24383               SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES and PCLMUL
24384               instruction set support.
24385
24386           ivybridge
24387               Intel Ivy Bridge CPU with 64-bit extensions, MMX, SSE, SSE2,
24388               SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES, PCLMUL,
24389               FSGSBASE, RDRND and F16C instruction set support.
24390
24391           haswell
24392               Intel Haswell CPU with 64-bit extensions, MOVBE, MMX, SSE,
24393               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
24394               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2 and F16C instruction
24395               set support.
24396
24397           broadwell
24398               Intel Broadwell CPU with 64-bit extensions, MOVBE, MMX, SSE,
24399               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
24400               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED ADCX and
24401               PREFETCHW instruction set support.
24402
24403           skylake
24404               Intel Skylake CPU with 64-bit extensions, MOVBE, MMX, SSE,
24405               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
24406               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
24407               PREFETCHW, CLFLUSHOPT, XSAVEC and XSAVES instruction set
24408               support.
24409
24410           bonnell
24411               Intel Bonnell CPU with 64-bit extensions, MOVBE, MMX, SSE,
24412               SSE2, SSE3 and SSSE3 instruction set support.
24413
24414           silvermont
24415               Intel Silvermont CPU with 64-bit extensions, MOVBE, MMX, SSE,
24416               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PREFETCHW,
24417               PCLMUL and RDRND instruction set support.
24418
24419           goldmont
24420               Intel Goldmont CPU with 64-bit extensions, MOVBE, MMX, SSE,
24421               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PREFETCHW,
24422               PCLMUL, RDRND, XSAVE, XSAVEC, XSAVES, XSAVEOPT and FSGSBASE
24423               instruction set support.
24424
24425           goldmont-plus
24426               Intel Goldmont Plus CPU with 64-bit extensions, MOVBE, MMX,
24427               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PREFETCHW,
24428               PCLMUL, RDRND, XSAVE, XSAVEC, XSAVES, XSAVEOPT, FSGSBASE,
24429               PTWRITE, RDPID, SGX and UMIP instruction set support.
24430
24431           tremont
24432               Intel Tremont CPU with 64-bit extensions, MOVBE, MMX, SSE,
24433               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PREFETCHW,
24434               PCLMUL, RDRND, XSAVE, XSAVEC, XSAVES, XSAVEOPT, FSGSBASE,
24435               PTWRITE, RDPID, SGX, UMIP, GFNI-SSE, CLWB, MOVDIRI, MOVDIR64B,
24436               CLDEMOTE and WAITPKG instruction set support.
24437
24438           knl Intel Knight's Landing CPU with 64-bit extensions, MOVBE, MMX,
24439               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
24440               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
24441               PREFETCHW, PREFETCHWT1, AVX512F, AVX512PF, AVX512ER and
24442               AVX512CD instruction set support.
24443
24444           knm Intel Knights Mill CPU with 64-bit extensions, MOVBE, MMX, SSE,
24445               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
24446               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
24447               PREFETCHW, PREFETCHWT1, AVX512F, AVX512PF, AVX512ER, AVX512CD,
24448               AVX5124VNNIW, AVX5124FMAPS and AVX512VPOPCNTDQ instruction set
24449               support.
24450
24451           skylake-avx512
24452               Intel Skylake Server CPU with 64-bit extensions, MOVBE, MMX,
24453               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
24454               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
24455               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, CLWB,
24456               AVX512VL, AVX512BW, AVX512DQ and AVX512CD instruction set
24457               support.
24458
24459           cannonlake
24460               Intel Cannonlake Server CPU with 64-bit extensions, MOVBE, MMX,
24461               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
24462               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
24463               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
24464               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA and
24465               UMIP instruction set support.
24466
24467           icelake-client
24468               Intel Icelake Client CPU with 64-bit extensions, MOVBE, MMX,
24469               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
24470               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
24471               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
24472               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA,
24473               CLWB, UMIP, RDPID, GFNI, AVX512VBMI2, AVX512VPOPCNTDQ,
24474               AVX512BITALG, AVX512VNNI, VPCLMULQDQ, VAES instruction set
24475               support.
24476
24477           icelake-server
24478               Intel Icelake Server CPU with 64-bit extensions, MOVBE, MMX,
24479               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
24480               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
24481               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
24482               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA,
24483               CLWB, UMIP, RDPID, GFNI, AVX512VBMI2, AVX512VPOPCNTDQ,
24484               AVX512BITALG, AVX512VNNI, VPCLMULQDQ, VAES, PCONFIG and
24485               WBNOINVD instruction set support.
24486
24487           cascadelake
24488               Intel Cascadelake CPU with 64-bit extensions, MOVBE, MMX, SSE,
24489               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2, AES,
24490               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
24491               PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, CLWB, AVX512VL,
24492               AVX512BW, AVX512DQ, AVX512CD and AVX512VNNI instruction set
24493               support.
24494
24495           cooperlake
24496               Intel cooperlake CPU with 64-bit extensions, MOVBE, MMX, SSE,
24497               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2, AES,
24498               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
24499               PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, CLWB, AVX512VL,
24500               AVX512BW, AVX512DQ, AVX512CD, AVX512VNNI and AVX512BF16
24501               instruction set support.
24502
24503           tigerlake
24504               Intel Tigerlake CPU with 64-bit extensions, MOVBE, MMX, SSE,
24505               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2, AES,
24506               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
24507               PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
24508               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA,
24509               CLWB, UMIP, RDPID, GFNI, AVX512VBMI2, AVX512VPOPCNTDQ,
24510               AVX512BITALG, AVX512VNNI, VPCLMULQDQ, VAES, PCONFIG, WBNOINVD,
24511               MOVDIRI, MOVDIR64B, AVX512VP2INTERSECT and KEYLOCKER
24512               instruction set support.
24513
24514           k6  AMD K6 CPU with MMX instruction set support.
24515
24516           k6-2
24517           k6-3
24518               Improved versions of AMD K6 CPU with MMX and 3DNow! instruction
24519               set support.
24520
24521           athlon
24522           athlon-tbird
24523               AMD Athlon CPU with MMX, 3dNOW!, enhanced 3DNow! and SSE
24524               prefetch instructions support.
24525
24526           athlon-4
24527           athlon-xp
24528           athlon-mp
24529               Improved AMD Athlon CPU with MMX, 3DNow!, enhanced 3DNow! and
24530               full SSE instruction set support.
24531
24532           k8
24533           opteron
24534           athlon64
24535           athlon-fx
24536               Processors based on the AMD K8 core with x86-64 instruction set
24537               support, including the AMD Opteron, Athlon 64, and Athlon 64 FX
24538               processors.  (This supersets MMX, SSE, SSE2, 3DNow!, enhanced
24539               3DNow! and 64-bit instruction set extensions.)
24540
24541           k8-sse3
24542           opteron-sse3
24543           athlon64-sse3
24544               Improved versions of AMD K8 cores with SSE3 instruction set
24545               support.
24546
24547           amdfam10
24548           barcelona
24549               CPUs based on AMD Family 10h cores with x86-64 instruction set
24550               support.  (This supersets MMX, SSE, SSE2, SSE3, SSE4A, 3DNow!,
24551               enhanced 3DNow!, ABM and 64-bit instruction set extensions.)
24552
24553           bdver1
24554               CPUs based on AMD Family 15h cores with x86-64 instruction set
24555               support.  (This supersets FMA4, AVX, XOP, LWP, AES, PCLMUL,
24556               CX16, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1, SSE4.2, ABM
24557               and 64-bit instruction set extensions.)
24558
24559           bdver2
24560               AMD Family 15h core based CPUs with x86-64 instruction set
24561               support.  (This supersets BMI, TBM, F16C, FMA, FMA4, AVX, XOP,
24562               LWP, AES, PCLMUL, CX16, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3,
24563               SSE4.1, SSE4.2, ABM and 64-bit instruction set extensions.)
24564
24565           bdver3
24566               AMD Family 15h core based CPUs with x86-64 instruction set
24567               support.  (This supersets BMI, TBM, F16C, FMA, FMA4, FSGSBASE,
24568               AVX, XOP, LWP, AES, PCLMUL, CX16, MMX, SSE, SSE2, SSE3, SSE4A,
24569               SSSE3, SSE4.1, SSE4.2, ABM and 64-bit instruction set
24570               extensions.)
24571
24572           bdver4
24573               AMD Family 15h core based CPUs with x86-64 instruction set
24574               support.  (This supersets BMI, BMI2, TBM, F16C, FMA, FMA4,
24575               FSGSBASE, AVX, AVX2, XOP, LWP, AES, PCLMUL, CX16, MOVBE, MMX,
24576               SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1, SSE4.2, ABM and 64-bit
24577               instruction set extensions.)
24578
24579           znver1
24580               AMD Family 17h core based CPUs with x86-64 instruction set
24581               support.  (This supersets BMI, BMI2, F16C, FMA, FSGSBASE, AVX,
24582               AVX2, ADCX, RDSEED, MWAITX, SHA, CLZERO, AES, PCLMUL, CX16,
24583               MOVBE, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1, SSE4.2, ABM,
24584               XSAVEC, XSAVES, CLFLUSHOPT, POPCNT, and 64-bit instruction set
24585               extensions.)
24586
24587           znver2
24588               AMD Family 17h core based CPUs with x86-64 instruction set
24589               support. (This supersets BMI, BMI2, CLWB, F16C, FMA, FSGSBASE,
24590               AVX, AVX2, ADCX, RDSEED, MWAITX, SHA, CLZERO, AES, PCLMUL,
24591               CX16, MOVBE, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1,
24592               SSE4.2, ABM, XSAVEC, XSAVES, CLFLUSHOPT, POPCNT, RDPID,
24593               WBNOINVD, and 64-bit instruction set extensions.)
24594
24595           btver1
24596               CPUs based on AMD Family 14h cores with x86-64 instruction set
24597               support.  (This supersets MMX, SSE, SSE2, SSE3, SSSE3, SSE4A,
24598               CX16, ABM and 64-bit instruction set extensions.)
24599
24600           btver2
24601               CPUs based on AMD Family 16h cores with x86-64 instruction set
24602               support. This includes MOVBE, F16C, BMI, AVX, PCLMUL, AES,
24603               SSE4.2, SSE4.1, CX16, ABM, SSE4A, SSSE3, SSE3, SSE2, SSE, MMX
24604               and 64-bit instruction set extensions.
24605
24606           winchip-c6
24607               IDT WinChip C6 CPU, dealt in same way as i486 with additional
24608               MMX instruction set support.
24609
24610           winchip2
24611               IDT WinChip 2 CPU, dealt in same way as i486 with additional
24612               MMX and 3DNow!  instruction set support.
24613
24614           c3  VIA C3 CPU with MMX and 3DNow! instruction set support.  (No
24615               scheduling is implemented for this chip.)
24616
24617           c3-2
24618               VIA C3-2 (Nehemiah/C5XL) CPU with MMX and SSE instruction set
24619               support.  (No scheduling is implemented for this chip.)
24620
24621           c7  VIA C7 (Esther) CPU with MMX, SSE, SSE2 and SSE3 instruction
24622               set support.  (No scheduling is implemented for this chip.)
24623
24624           samuel-2
24625               VIA Eden Samuel 2 CPU with MMX and 3DNow! instruction set
24626               support.  (No scheduling is implemented for this chip.)
24627
24628           nehemiah
24629               VIA Eden Nehemiah CPU with MMX and SSE instruction set support.
24630               (No scheduling is implemented for this chip.)
24631
24632           esther
24633               VIA Eden Esther CPU with MMX, SSE, SSE2 and SSE3 instruction
24634               set support.  (No scheduling is implemented for this chip.)
24635
24636           eden-x2
24637               VIA Eden X2 CPU with x86-64, MMX, SSE, SSE2 and SSE3
24638               instruction set support.  (No scheduling is implemented for
24639               this chip.)
24640
24641           eden-x4
24642               VIA Eden X4 CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3,
24643               SSE4.1, SSE4.2, AVX and AVX2 instruction set support.  (No
24644               scheduling is implemented for this chip.)
24645
24646           nano
24647               Generic VIA Nano CPU with x86-64, MMX, SSE, SSE2, SSE3 and
24648               SSSE3 instruction set support.  (No scheduling is implemented
24649               for this chip.)
24650
24651           nano-1000
24652               VIA Nano 1xxx CPU with x86-64, MMX, SSE, SSE2, SSE3 and SSSE3
24653               instruction set support.  (No scheduling is implemented for
24654               this chip.)
24655
24656           nano-2000
24657               VIA Nano 2xxx CPU with x86-64, MMX, SSE, SSE2, SSE3 and SSSE3
24658               instruction set support.  (No scheduling is implemented for
24659               this chip.)
24660
24661           nano-3000
24662               VIA Nano 3xxx CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3 and
24663               SSE4.1 instruction set support.  (No scheduling is implemented
24664               for this chip.)
24665
24666           nano-x2
24667               VIA Nano Dual Core CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3
24668               and SSE4.1 instruction set support.  (No scheduling is
24669               implemented for this chip.)
24670
24671           nano-x4
24672               VIA Nano Quad Core CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3
24673               and SSE4.1 instruction set support.  (No scheduling is
24674               implemented for this chip.)
24675
24676           geode
24677               AMD Geode embedded processor with MMX and 3DNow! instruction
24678               set support.
24679
24680       -mtune=cpu-type
24681           Tune to cpu-type everything applicable about the generated code,
24682           except for the ABI and the set of available instructions.  While
24683           picking a specific cpu-type schedules things appropriately for that
24684           particular chip, the compiler does not generate any code that
24685           cannot run on the default machine type unless you use a -march=cpu-
24686           type option.  For example, if GCC is configured for
24687           i686-pc-linux-gnu then -mtune=pentium4 generates code that is tuned
24688           for Pentium 4 but still runs on i686 machines.
24689
24690           The choices for cpu-type are the same as for -march.  In addition,
24691           -mtune supports 2 extra choices for cpu-type:
24692
24693           generic
24694               Produce code optimized for the most common IA32/AMD64/EM64T
24695               processors.  If you know the CPU on which your code will run,
24696               then you should use the corresponding -mtune or -march option
24697               instead of -mtune=generic.  But, if you do not know exactly
24698               what CPU users of your application will have, then you should
24699               use this option.
24700
24701               As new processors are deployed in the marketplace, the behavior
24702               of this option will change.  Therefore, if you upgrade to a
24703               newer version of GCC, code generation controlled by this option
24704               will change to reflect the processors that are most common at
24705               the time that version of GCC is released.
24706
24707               There is no -march=generic option because -march indicates the
24708               instruction set the compiler can use, and there is no generic
24709               instruction set applicable to all processors.  In contrast,
24710               -mtune indicates the processor (or, in this case, collection of
24711               processors) for which the code is optimized.
24712
24713           intel
24714               Produce code optimized for the most current Intel processors,
24715               which are Haswell and Silvermont for this version of GCC.  If
24716               you know the CPU on which your code will run, then you should
24717               use the corresponding -mtune or -march option instead of
24718               -mtune=intel.  But, if you want your application performs
24719               better on both Haswell and Silvermont, then you should use this
24720               option.
24721
24722               As new Intel processors are deployed in the marketplace, the
24723               behavior of this option will change.  Therefore, if you upgrade
24724               to a newer version of GCC, code generation controlled by this
24725               option will change to reflect the most current Intel processors
24726               at the time that version of GCC is released.
24727
24728               There is no -march=intel option because -march indicates the
24729               instruction set the compiler can use, and there is no common
24730               instruction set applicable to all processors.  In contrast,
24731               -mtune indicates the processor (or, in this case, collection of
24732               processors) for which the code is optimized.
24733
24734       -mcpu=cpu-type
24735           A deprecated synonym for -mtune.
24736
24737       -mfpmath=unit
24738           Generate floating-point arithmetic for selected unit unit.  The
24739           choices for unit are:
24740
24741           387 Use the standard 387 floating-point coprocessor present on the
24742               majority of chips and emulated otherwise.  Code compiled with
24743               this option runs almost everywhere.  The temporary results are
24744               computed in 80-bit precision instead of the precision specified
24745               by the type, resulting in slightly different results compared
24746               to most of other chips.  See -ffloat-store for more detailed
24747               description.
24748
24749               This is the default choice for non-Darwin x86-32 targets.
24750
24751           sse Use scalar floating-point instructions present in the SSE
24752               instruction set.  This instruction set is supported by Pentium
24753               III and newer chips, and in the AMD line by Athlon-4, Athlon XP
24754               and Athlon MP chips.  The earlier version of the SSE
24755               instruction set supports only single-precision arithmetic, thus
24756               the double and extended-precision arithmetic are still done
24757               using 387.  A later version, present only in Pentium 4 and AMD
24758               x86-64 chips, supports double-precision arithmetic too.
24759
24760               For the x86-32 compiler, you must use -march=cpu-type, -msse or
24761               -msse2 switches to enable SSE extensions and make this option
24762               effective.  For the x86-64 compiler, these extensions are
24763               enabled by default.
24764
24765               The resulting code should be considerably faster in the
24766               majority of cases and avoid the numerical instability problems
24767               of 387 code, but may break some existing code that expects
24768               temporaries to be 80 bits.
24769
24770               This is the default choice for the x86-64 compiler, Darwin
24771               x86-32 targets, and the default choice for x86-32 targets with
24772               the SSE2 instruction set when -ffast-math is enabled.
24773
24774           sse,387
24775           sse+387
24776           both
24777               Attempt to utilize both instruction sets at once.  This
24778               effectively doubles the amount of available registers, and on
24779               chips with separate execution units for 387 and SSE the
24780               execution resources too.  Use this option with care, as it is
24781               still experimental, because the GCC register allocator does not
24782               model separate functional units well, resulting in unstable
24783               performance.
24784
24785       -masm=dialect
24786           Output assembly instructions using selected dialect.  Also affects
24787           which dialect is used for basic "asm" and extended "asm". Supported
24788           choices (in dialect order) are att or intel. The default is att.
24789           Darwin does not support intel.
24790
24791       -mieee-fp
24792       -mno-ieee-fp
24793           Control whether or not the compiler uses IEEE floating-point
24794           comparisons.  These correctly handle the case where the result of a
24795           comparison is unordered.
24796
24797       -m80387
24798       -mhard-float
24799           Generate output containing 80387 instructions for floating point.
24800
24801       -mno-80387
24802       -msoft-float
24803           Generate output containing library calls for floating point.
24804
24805           Warning: the requisite libraries are not part of GCC.  Normally the
24806           facilities of the machine's usual C compiler are used, but this
24807           cannot be done directly in cross-compilation.  You must make your
24808           own arrangements to provide suitable library functions for cross-
24809           compilation.
24810
24811           On machines where a function returns floating-point results in the
24812           80387 register stack, some floating-point opcodes may be emitted
24813           even if -msoft-float is used.
24814
24815       -mno-fp-ret-in-387
24816           Do not use the FPU registers for return values of functions.
24817
24818           The usual calling convention has functions return values of types
24819           "float" and "double" in an FPU register, even if there is no FPU.
24820           The idea is that the operating system should emulate an FPU.
24821
24822           The option -mno-fp-ret-in-387 causes such values to be returned in
24823           ordinary CPU registers instead.
24824
24825       -mno-fancy-math-387
24826           Some 387 emulators do not support the "sin", "cos" and "sqrt"
24827           instructions for the 387.  Specify this option to avoid generating
24828           those instructions.  This option is overridden when -march
24829           indicates that the target CPU always has an FPU and so the
24830           instruction does not need emulation.  These instructions are not
24831           generated unless you also use the -funsafe-math-optimizations
24832           switch.
24833
24834       -malign-double
24835       -mno-align-double
24836           Control whether GCC aligns "double", "long double", and "long long"
24837           variables on a two-word boundary or a one-word boundary.  Aligning
24838           "double" variables on a two-word boundary produces code that runs
24839           somewhat faster on a Pentium at the expense of more memory.
24840
24841           On x86-64, -malign-double is enabled by default.
24842
24843           Warning: if you use the -malign-double switch, structures
24844           containing the above types are aligned differently than the
24845           published application binary interface specifications for the
24846           x86-32 and are not binary compatible with structures in code
24847           compiled without that switch.
24848
24849       -m96bit-long-double
24850       -m128bit-long-double
24851           These switches control the size of "long double" type.  The x86-32
24852           application binary interface specifies the size to be 96 bits, so
24853           -m96bit-long-double is the default in 32-bit mode.
24854
24855           Modern architectures (Pentium and newer) prefer "long double" to be
24856           aligned to an 8- or 16-byte boundary.  In arrays or structures
24857           conforming to the ABI, this is not possible.  So specifying
24858           -m128bit-long-double aligns "long double" to a 16-byte boundary by
24859           padding the "long double" with an additional 32-bit zero.
24860
24861           In the x86-64 compiler, -m128bit-long-double is the default choice
24862           as its ABI specifies that "long double" is aligned on 16-byte
24863           boundary.
24864
24865           Notice that neither of these options enable any extra precision
24866           over the x87 standard of 80 bits for a "long double".
24867
24868           Warning: if you override the default value for your target ABI,
24869           this changes the size of structures and arrays containing "long
24870           double" variables, as well as modifying the function calling
24871           convention for functions taking "long double".  Hence they are not
24872           binary-compatible with code compiled without that switch.
24873
24874       -mlong-double-64
24875       -mlong-double-80
24876       -mlong-double-128
24877           These switches control the size of "long double" type. A size of 64
24878           bits makes the "long double" type equivalent to the "double" type.
24879           This is the default for 32-bit Bionic C library.  A size of 128
24880           bits makes the "long double" type equivalent to the "__float128"
24881           type. This is the default for 64-bit Bionic C library.
24882
24883           Warning: if you override the default value for your target ABI,
24884           this changes the size of structures and arrays containing "long
24885           double" variables, as well as modifying the function calling
24886           convention for functions taking "long double".  Hence they are not
24887           binary-compatible with code compiled without that switch.
24888
24889       -malign-data=type
24890           Control how GCC aligns variables.  Supported values for type are
24891           compat uses increased alignment value compatible uses GCC 4.8 and
24892           earlier, abi uses alignment value as specified by the psABI, and
24893           cacheline uses increased alignment value to match the cache line
24894           size.  compat is the default.
24895
24896       -mlarge-data-threshold=threshold
24897           When -mcmodel=medium is specified, data objects larger than
24898           threshold are placed in the large data section.  This value must be
24899           the same across all objects linked into the binary, and defaults to
24900           65535.
24901
24902       -mrtd
24903           Use a different function-calling convention, in which functions
24904           that take a fixed number of arguments return with the "ret num"
24905           instruction, which pops their arguments while returning.  This
24906           saves one instruction in the caller since there is no need to pop
24907           the arguments there.
24908
24909           You can specify that an individual function is called with this
24910           calling sequence with the function attribute "stdcall".  You can
24911           also override the -mrtd option by using the function attribute
24912           "cdecl".
24913
24914           Warning: this calling convention is incompatible with the one
24915           normally used on Unix, so you cannot use it if you need to call
24916           libraries compiled with the Unix compiler.
24917
24918           Also, you must provide function prototypes for all functions that
24919           take variable numbers of arguments (including "printf"); otherwise
24920           incorrect code is generated for calls to those functions.
24921
24922           In addition, seriously incorrect code results if you call a
24923           function with too many arguments.  (Normally, extra arguments are
24924           harmlessly ignored.)
24925
24926       -mregparm=num
24927           Control how many registers are used to pass integer arguments.  By
24928           default, no registers are used to pass arguments, and at most 3
24929           registers can be used.  You can control this behavior for a
24930           specific function by using the function attribute "regparm".
24931
24932           Warning: if you use this switch, and num is nonzero, then you must
24933           build all modules with the same value, including any libraries.
24934           This includes the system libraries and startup modules.
24935
24936       -msseregparm
24937           Use SSE register passing conventions for float and double arguments
24938           and return values.  You can control this behavior for a specific
24939           function by using the function attribute "sseregparm".
24940
24941           Warning: if you use this switch then you must build all modules
24942           with the same value, including any libraries.  This includes the
24943           system libraries and startup modules.
24944
24945       -mvect8-ret-in-mem
24946           Return 8-byte vectors in memory instead of MMX registers.  This is
24947           the default on VxWorks to match the ABI of the Sun Studio compilers
24948           until version 12.  Only use this option if you need to remain
24949           compatible with existing code produced by those previous compiler
24950           versions or older versions of GCC.
24951
24952       -mpc32
24953       -mpc64
24954       -mpc80
24955           Set 80387 floating-point precision to 32, 64 or 80 bits.  When
24956           -mpc32 is specified, the significands of results of floating-point
24957           operations are rounded to 24 bits (single precision); -mpc64 rounds
24958           the significands of results of floating-point operations to 53 bits
24959           (double precision) and -mpc80 rounds the significands of results of
24960           floating-point operations to 64 bits (extended double precision),
24961           which is the default.  When this option is used, floating-point
24962           operations in higher precisions are not available to the programmer
24963           without setting the FPU control word explicitly.
24964
24965           Setting the rounding of floating-point operations to less than the
24966           default 80 bits can speed some programs by 2% or more.  Note that
24967           some mathematical libraries assume that extended-precision (80-bit)
24968           floating-point operations are enabled by default; routines in such
24969           libraries could suffer significant loss of accuracy, typically
24970           through so-called "catastrophic cancellation", when this option is
24971           used to set the precision to less than extended precision.
24972
24973       -mstackrealign
24974           Realign the stack at entry.  On the x86, the -mstackrealign option
24975           generates an alternate prologue and epilogue that realigns the run-
24976           time stack if necessary.  This supports mixing legacy codes that
24977           keep 4-byte stack alignment with modern codes that keep 16-byte
24978           stack alignment for SSE compatibility.  See also the attribute
24979           "force_align_arg_pointer", applicable to individual functions.
24980
24981       -mpreferred-stack-boundary=num
24982           Attempt to keep the stack boundary aligned to a 2 raised to num
24983           byte boundary.  If -mpreferred-stack-boundary is not specified, the
24984           default is 4 (16 bytes or 128 bits).
24985
24986           Warning: When generating code for the x86-64 architecture with SSE
24987           extensions disabled, -mpreferred-stack-boundary=3 can be used to
24988           keep the stack boundary aligned to 8 byte boundary.  Since x86-64
24989           ABI require 16 byte stack alignment, this is ABI incompatible and
24990           intended to be used in controlled environment where stack space is
24991           important limitation.  This option leads to wrong code when
24992           functions compiled with 16 byte stack alignment (such as functions
24993           from a standard library) are called with misaligned stack.  In this
24994           case, SSE instructions may lead to misaligned memory access traps.
24995           In addition, variable arguments are handled incorrectly for 16 byte
24996           aligned objects (including x87 long double and __int128), leading
24997           to wrong results.  You must build all modules with
24998           -mpreferred-stack-boundary=3, including any libraries.  This
24999           includes the system libraries and startup modules.
25000
25001       -mincoming-stack-boundary=num
25002           Assume the incoming stack is aligned to a 2 raised to num byte
25003           boundary.  If -mincoming-stack-boundary is not specified, the one
25004           specified by -mpreferred-stack-boundary is used.
25005
25006           On Pentium and Pentium Pro, "double" and "long double" values
25007           should be aligned to an 8-byte boundary (see -malign-double) or
25008           suffer significant run time performance penalties.  On Pentium III,
25009           the Streaming SIMD Extension (SSE) data type "__m128" may not work
25010           properly if it is not 16-byte aligned.
25011
25012           To ensure proper alignment of this values on the stack, the stack
25013           boundary must be as aligned as that required by any value stored on
25014           the stack.  Further, every function must be generated such that it
25015           keeps the stack aligned.  Thus calling a function compiled with a
25016           higher preferred stack boundary from a function compiled with a
25017           lower preferred stack boundary most likely misaligns the stack.  It
25018           is recommended that libraries that use callbacks always use the
25019           default setting.
25020
25021           This extra alignment does consume extra stack space, and generally
25022           increases code size.  Code that is sensitive to stack space usage,
25023           such as embedded systems and operating system kernels, may want to
25024           reduce the preferred alignment to -mpreferred-stack-boundary=2.
25025
25026       -mmmx
25027       -msse
25028       -msse2
25029       -msse3
25030       -mssse3
25031       -msse4
25032       -msse4a
25033       -msse4.1
25034       -msse4.2
25035       -mavx
25036       -mavx2
25037       -mavx512f
25038       -mavx512pf
25039       -mavx512er
25040       -mavx512cd
25041       -mavx512vl
25042       -mavx512bw
25043       -mavx512dq
25044       -mavx512ifma
25045       -mavx512vbmi
25046       -msha
25047       -maes
25048       -mpclmul
25049       -mclflushopt
25050       -mclwb
25051       -mfsgsbase
25052       -mptwrite
25053       -mrdrnd
25054       -mf16c
25055       -mfma
25056       -mpconfig
25057       -mwbnoinvd
25058       -mfma4
25059       -mprfchw
25060       -mrdpid
25061       -mprefetchwt1
25062       -mrdseed
25063       -msgx
25064       -mxop
25065       -mlwp
25066       -m3dnow
25067       -m3dnowa
25068       -mpopcnt
25069       -mabm
25070       -madx
25071       -mbmi
25072       -mbmi2
25073       -mlzcnt
25074       -mfxsr
25075       -mxsave
25076       -mxsaveopt
25077       -mxsavec
25078       -mxsaves
25079       -mrtm
25080       -mhle
25081       -mtbm
25082       -mmwaitx
25083       -mclzero
25084       -mpku
25085       -mavx512vbmi2
25086       -mavx512bf16
25087       -mgfni
25088       -mvaes
25089       -mwaitpkg
25090       -mvpclmulqdq
25091       -mavx512bitalg
25092       -mmovdiri
25093       -mmovdir64b
25094       -menqcmd
25095       -mavx512vpopcntdq
25096       -mavx512vp2intersect
25097       -mavx5124fmaps
25098       -mavx512vnni
25099       -mavx5124vnniw
25100       -mcldemote
25101           These switches enable the use of instructions in the MMX, SSE,
25102           SSE2, SSE3, SSSE3, SSE4, SSE4A, SSE4.1, SSE4.2, AVX, AVX2, AVX512F,
25103           AVX512PF, AVX512ER, AVX512CD, AVX512VL, AVX512BW, AVX512DQ,
25104           AVX512IFMA, AVX512VBMI, SHA, AES, PCLMUL, CLFLUSHOPT, CLWB,
25105           FSGSBASE, PTWRITE, RDRND, F16C, FMA, PCONFIG, WBNOINVD, FMA4,
25106           PREFETCHW, RDPID, PREFETCHWT1, RDSEED, SGX, XOP, LWP, 3DNow!,
25107           enhanced 3DNow!, POPCNT, ABM, ADX, BMI, BMI2, LZCNT, FXSR, XSAVE,
25108           XSAVEOPT, XSAVEC, XSAVES, RTM, HLE, TBM, MWAITX, CLZERO, PKU,
25109           AVX512VBMI2, GFNI, VAES, WAITPKG, VPCLMULQDQ, AVX512BITALG,
25110           MOVDIRI, MOVDIR64B, AVX512BF16, ENQCMD, AVX512VPOPCNTDQ,
25111           AVX5124FMAPS, AVX512VNNI, AVX5124VNNIW, or CLDEMOTE extended
25112           instruction sets.  Each has a corresponding -mno- option to disable
25113           use of these instructions.
25114
25115           These extensions are also available as built-in functions: see x86
25116           Built-in Functions, for details of the functions enabled and
25117           disabled by these switches.
25118
25119           To generate SSE/SSE2 instructions automatically from floating-point
25120           code (as opposed to 387 instructions), see -mfpmath=sse.
25121
25122           GCC depresses SSEx instructions when -mavx is used. Instead, it
25123           generates new AVX instructions or AVX equivalence for all SSEx
25124           instructions when needed.
25125
25126           These options enable GCC to use these extended instructions in
25127           generated code, even without -mfpmath=sse.  Applications that
25128           perform run-time CPU detection must compile separate files for each
25129           supported architecture, using the appropriate flags.  In
25130           particular, the file containing the CPU detection code should be
25131           compiled without these options.
25132
25133       -mdump-tune-features
25134           This option instructs GCC to dump the names of the x86 performance
25135           tuning features and default settings. The names can be used in
25136           -mtune-ctrl=feature-list.
25137
25138       -mtune-ctrl=feature-list
25139           This option is used to do fine grain control of x86 code generation
25140           features.  feature-list is a comma separated list of feature names.
25141           See also -mdump-tune-features. When specified, the feature is
25142           turned on if it is not preceded with ^, otherwise, it is turned
25143           off.  -mtune-ctrl=feature-list is intended to be used by GCC
25144           developers. Using it may lead to code paths not covered by testing
25145           and can potentially result in compiler ICEs or runtime errors.
25146
25147       -mno-default
25148           This option instructs GCC to turn off all tunable features. See
25149           also -mtune-ctrl=feature-list and -mdump-tune-features.
25150
25151       -mcld
25152           This option instructs GCC to emit a "cld" instruction in the
25153           prologue of functions that use string instructions.  String
25154           instructions depend on the DF flag to select between autoincrement
25155           or autodecrement mode.  While the ABI specifies the DF flag to be
25156           cleared on function entry, some operating systems violate this
25157           specification by not clearing the DF flag in their exception
25158           dispatchers.  The exception handler can be invoked with the DF flag
25159           set, which leads to wrong direction mode when string instructions
25160           are used.  This option can be enabled by default on 32-bit x86
25161           targets by configuring GCC with the --enable-cld configure option.
25162           Generation of "cld" instructions can be suppressed with the
25163           -mno-cld compiler option in this case.
25164
25165       -mvzeroupper
25166           This option instructs GCC to emit a "vzeroupper" instruction before
25167           a transfer of control flow out of the function to minimize the AVX
25168           to SSE transition penalty as well as remove unnecessary "zeroupper"
25169           intrinsics.
25170
25171       -mprefer-avx128
25172           This option instructs GCC to use 128-bit AVX instructions instead
25173           of 256-bit AVX instructions in the auto-vectorizer.
25174
25175       -mprefer-vector-width=opt
25176           This option instructs GCC to use opt-bit vector width in
25177           instructions instead of default on the selected platform.
25178
25179           none
25180               No extra limitations applied to GCC other than defined by the
25181               selected platform.
25182
25183           128 Prefer 128-bit vector width for instructions.
25184
25185           256 Prefer 256-bit vector width for instructions.
25186
25187           512 Prefer 512-bit vector width for instructions.
25188
25189       -mcx16
25190           This option enables GCC to generate "CMPXCHG16B" instructions in
25191           64-bit code to implement compare-and-exchange operations on 16-byte
25192           aligned 128-bit objects.  This is useful for atomic updates of data
25193           structures exceeding one machine word in size.  The compiler uses
25194           this instruction to implement __sync Builtins.  However, for
25195           __atomic Builtins operating on 128-bit integers, a library call is
25196           always used.
25197
25198       -msahf
25199           This option enables generation of "SAHF" instructions in 64-bit
25200           code.  Early Intel Pentium 4 CPUs with Intel 64 support, prior to
25201           the introduction of Pentium 4 G1 step in December 2005, lacked the
25202           "LAHF" and "SAHF" instructions which are supported by AMD64.  These
25203           are load and store instructions, respectively, for certain status
25204           flags.  In 64-bit mode, the "SAHF" instruction is used to optimize
25205           "fmod", "drem", and "remainder" built-in functions; see Other
25206           Builtins for details.
25207
25208       -mmovbe
25209           This option enables use of the "movbe" instruction to implement
25210           "__builtin_bswap32" and "__builtin_bswap64".
25211
25212       -mshstk
25213           The -mshstk option enables shadow stack built-in functions from x86
25214           Control-flow Enforcement Technology (CET).
25215
25216       -mcrc32
25217           This option enables built-in functions "__builtin_ia32_crc32qi",
25218           "__builtin_ia32_crc32hi", "__builtin_ia32_crc32si" and
25219           "__builtin_ia32_crc32di" to generate the "crc32" machine
25220           instruction.
25221
25222       -mrecip
25223           This option enables use of "RCPSS" and "RSQRTSS" instructions (and
25224           their vectorized variants "RCPPS" and "RSQRTPS") with an additional
25225           Newton-Raphson step to increase precision instead of "DIVSS" and
25226           "SQRTSS" (and their vectorized variants) for single-precision
25227           floating-point arguments.  These instructions are generated only
25228           when -funsafe-math-optimizations is enabled together with
25229           -ffinite-math-only and -fno-trapping-math.  Note that while the
25230           throughput of the sequence is higher than the throughput of the
25231           non-reciprocal instruction, the precision of the sequence can be
25232           decreased by up to 2 ulp (i.e. the inverse of 1.0 equals
25233           0.99999994).
25234
25235           Note that GCC implements "1.0f/sqrtf(x)" in terms of "RSQRTSS" (or
25236           "RSQRTPS") already with -ffast-math (or the above option
25237           combination), and doesn't need -mrecip.
25238
25239           Also note that GCC emits the above sequence with additional Newton-
25240           Raphson step for vectorized single-float division and vectorized
25241           "sqrtf(x)" already with -ffast-math (or the above option
25242           combination), and doesn't need -mrecip.
25243
25244       -mrecip=opt
25245           This option controls which reciprocal estimate instructions may be
25246           used.  opt is a comma-separated list of options, which may be
25247           preceded by a ! to invert the option:
25248
25249           all Enable all estimate instructions.
25250
25251           default
25252               Enable the default instructions, equivalent to -mrecip.
25253
25254           none
25255               Disable all estimate instructions, equivalent to -mno-recip.
25256
25257           div Enable the approximation for scalar division.
25258
25259           vec-div
25260               Enable the approximation for vectorized division.
25261
25262           sqrt
25263               Enable the approximation for scalar square root.
25264
25265           vec-sqrt
25266               Enable the approximation for vectorized square root.
25267
25268           So, for example, -mrecip=all,!sqrt enables all of the reciprocal
25269           approximations, except for square root.
25270
25271       -mveclibabi=type
25272           Specifies the ABI type to use for vectorizing intrinsics using an
25273           external library.  Supported values for type are svml for the Intel
25274           short vector math library and acml for the AMD math core library.
25275           To use this option, both -ftree-vectorize and
25276           -funsafe-math-optimizations have to be enabled, and an SVML or ACML
25277           ABI-compatible library must be specified at link time.
25278
25279           GCC currently emits calls to "vmldExp2", "vmldLn2", "vmldLog102",
25280           "vmldPow2", "vmldTanh2", "vmldTan2", "vmldAtan2", "vmldAtanh2",
25281           "vmldCbrt2", "vmldSinh2", "vmldSin2", "vmldAsinh2", "vmldAsin2",
25282           "vmldCosh2", "vmldCos2", "vmldAcosh2", "vmldAcos2", "vmlsExp4",
25283           "vmlsLn4", "vmlsLog104", "vmlsPow4", "vmlsTanh4", "vmlsTan4",
25284           "vmlsAtan4", "vmlsAtanh4", "vmlsCbrt4", "vmlsSinh4", "vmlsSin4",
25285           "vmlsAsinh4", "vmlsAsin4", "vmlsCosh4", "vmlsCos4", "vmlsAcosh4"
25286           and "vmlsAcos4" for corresponding function type when
25287           -mveclibabi=svml is used, and "__vrd2_sin", "__vrd2_cos",
25288           "__vrd2_exp", "__vrd2_log", "__vrd2_log2", "__vrd2_log10",
25289           "__vrs4_sinf", "__vrs4_cosf", "__vrs4_expf", "__vrs4_logf",
25290           "__vrs4_log2f", "__vrs4_log10f" and "__vrs4_powf" for the
25291           corresponding function type when -mveclibabi=acml is used.
25292
25293       -mabi=name
25294           Generate code for the specified calling convention.  Permissible
25295           values are sysv for the ABI used on GNU/Linux and other systems,
25296           and ms for the Microsoft ABI.  The default is to use the Microsoft
25297           ABI when targeting Microsoft Windows and the SysV ABI on all other
25298           systems.  You can control this behavior for specific functions by
25299           using the function attributes "ms_abi" and "sysv_abi".
25300
25301       -mforce-indirect-call
25302           Force all calls to functions to be indirect. This is useful when
25303           using Intel Processor Trace where it generates more precise timing
25304           information for function calls.
25305
25306       -mmanual-endbr
25307           Insert ENDBR instruction at function entry only via the "cf_check"
25308           function attribute. This is useful when used with the option
25309           -fcf-protection=branch to control ENDBR insertion at the function
25310           entry.
25311
25312       -mcall-ms2sysv-xlogues
25313           Due to differences in 64-bit ABIs, any Microsoft ABI function that
25314           calls a System V ABI function must consider RSI, RDI and XMM6-15 as
25315           clobbered.  By default, the code for saving and restoring these
25316           registers is emitted inline, resulting in fairly lengthy prologues
25317           and epilogues.  Using -mcall-ms2sysv-xlogues emits prologues and
25318           epilogues that use stubs in the static portion of libgcc to perform
25319           these saves and restores, thus reducing function size at the cost
25320           of a few extra instructions.
25321
25322       -mtls-dialect=type
25323           Generate code to access thread-local storage using the gnu or gnu2
25324           conventions.  gnu is the conservative default; gnu2 is more
25325           efficient, but it may add compile- and run-time requirements that
25326           cannot be satisfied on all systems.
25327
25328       -mpush-args
25329       -mno-push-args
25330           Use PUSH operations to store outgoing parameters.  This method is
25331           shorter and usually equally fast as method using SUB/MOV operations
25332           and is enabled by default.  In some cases disabling it may improve
25333           performance because of improved scheduling and reduced
25334           dependencies.
25335
25336       -maccumulate-outgoing-args
25337           If enabled, the maximum amount of space required for outgoing
25338           arguments is computed in the function prologue.  This is faster on
25339           most modern CPUs because of reduced dependencies, improved
25340           scheduling and reduced stack usage when the preferred stack
25341           boundary is not equal to 2.  The drawback is a notable increase in
25342           code size.  This switch implies -mno-push-args.
25343
25344       -mthreads
25345           Support thread-safe exception handling on MinGW.  Programs that
25346           rely on thread-safe exception handling must compile and link all
25347           code with the -mthreads option.  When compiling, -mthreads defines
25348           -D_MT; when linking, it links in a special thread helper library
25349           -lmingwthrd which cleans up per-thread exception-handling data.
25350
25351       -mms-bitfields
25352       -mno-ms-bitfields
25353           Enable/disable bit-field layout compatible with the native
25354           Microsoft Windows compiler.
25355
25356           If "packed" is used on a structure, or if bit-fields are used, it
25357           may be that the Microsoft ABI lays out the structure differently
25358           than the way GCC normally does.  Particularly when moving packed
25359           data between functions compiled with GCC and the native Microsoft
25360           compiler (either via function call or as data in a file), it may be
25361           necessary to access either format.
25362
25363           This option is enabled by default for Microsoft Windows targets.
25364           This behavior can also be controlled locally by use of variable or
25365           type attributes.  For more information, see x86 Variable Attributes
25366           and x86 Type Attributes.
25367
25368           The Microsoft structure layout algorithm is fairly simple with the
25369           exception of the bit-field packing.  The padding and alignment of
25370           members of structures and whether a bit-field can straddle a
25371           storage-unit boundary are determine by these rules:
25372
25373           1. Structure members are stored sequentially in the order in which
25374           they are
25375               declared: the first member has the lowest memory address and
25376               the last member the highest.
25377
25378           2. Every data object has an alignment requirement.  The alignment
25379           requirement
25380               for all data except structures, unions, and arrays is either
25381               the size of the object or the current packing size (specified
25382               with either the "aligned" attribute or the "pack" pragma),
25383               whichever is less.  For structures, unions, and arrays, the
25384               alignment requirement is the largest alignment requirement of
25385               its members.  Every object is allocated an offset so that:
25386
25387                       offset % alignment_requirement == 0
25388
25389           3. Adjacent bit-fields are packed into the same 1-, 2-, or 4-byte
25390           allocation
25391               unit if the integral types are the same size and if the next
25392               bit-field fits into the current allocation unit without
25393               crossing the boundary imposed by the common alignment
25394               requirements of the bit-fields.
25395
25396           MSVC interprets zero-length bit-fields in the following ways:
25397
25398           1. If a zero-length bit-field is inserted between two bit-fields
25399           that
25400               are normally coalesced, the bit-fields are not coalesced.
25401
25402               For example:
25403
25404                       struct
25405                        {
25406                          unsigned long bf_1 : 12;
25407                          unsigned long : 0;
25408                          unsigned long bf_2 : 12;
25409                        } t1;
25410
25411               The size of "t1" is 8 bytes with the zero-length bit-field.  If
25412               the zero-length bit-field were removed, "t1"'s size would be 4
25413               bytes.
25414
25415           2. If a zero-length bit-field is inserted after a bit-field, "foo",
25416           and the
25417               alignment of the zero-length bit-field is greater than the
25418               member that follows it, "bar", "bar" is aligned as the type of
25419               the zero-length bit-field.
25420
25421               For example:
25422
25423                       struct
25424                        {
25425                          char foo : 4;
25426                          short : 0;
25427                          char bar;
25428                        } t2;
25429
25430                       struct
25431                        {
25432                          char foo : 4;
25433                          short : 0;
25434                          double bar;
25435                        } t3;
25436
25437               For "t2", "bar" is placed at offset 2, rather than offset 1.
25438               Accordingly, the size of "t2" is 4.  For "t3", the zero-length
25439               bit-field does not affect the alignment of "bar" or, as a
25440               result, the size of the structure.
25441
25442               Taking this into account, it is important to note the
25443               following:
25444
25445               1. If a zero-length bit-field follows a normal bit-field, the
25446               type of the
25447                   zero-length bit-field may affect the alignment of the
25448                   structure as whole. For example, "t2" has a size of 4
25449                   bytes, since the zero-length bit-field follows a normal
25450                   bit-field, and is of type short.
25451
25452               2. Even if a zero-length bit-field is not followed by a normal
25453               bit-field, it may
25454                   still affect the alignment of the structure:
25455
25456                           struct
25457                            {
25458                              char foo : 6;
25459                              long : 0;
25460                            } t4;
25461
25462                   Here, "t4" takes up 4 bytes.
25463
25464           3. Zero-length bit-fields following non-bit-field members are
25465           ignored:
25466                       struct
25467                        {
25468                          char foo;
25469                          long : 0;
25470                          char bar;
25471                        } t5;
25472
25473               Here, "t5" takes up 2 bytes.
25474
25475       -mno-align-stringops
25476           Do not align the destination of inlined string operations.  This
25477           switch reduces code size and improves performance in case the
25478           destination is already aligned, but GCC doesn't know about it.
25479
25480       -minline-all-stringops
25481           By default GCC inlines string operations only when the destination
25482           is known to be aligned to least a 4-byte boundary.  This enables
25483           more inlining and increases code size, but may improve performance
25484           of code that depends on fast "memcpy" and "memset" for short
25485           lengths.  The option enables inline expansion of "strlen" for all
25486           pointer alignments.
25487
25488       -minline-stringops-dynamically
25489           For string operations of unknown size, use run-time checks with
25490           inline code for small blocks and a library call for large blocks.
25491
25492       -mstringop-strategy=alg
25493           Override the internal decision heuristic for the particular
25494           algorithm to use for inlining string operations.  The allowed
25495           values for alg are:
25496
25497           rep_byte
25498           rep_4byte
25499           rep_8byte
25500               Expand using i386 "rep" prefix of the specified size.
25501
25502           byte_loop
25503           loop
25504           unrolled_loop
25505               Expand into an inline loop.
25506
25507           libcall
25508               Always use a library call.
25509
25510       -mmemcpy-strategy=strategy
25511           Override the internal decision heuristic to decide if
25512           "__builtin_memcpy" should be inlined and what inline algorithm to
25513           use when the expected size of the copy operation is known. strategy
25514           is a comma-separated list of alg:max_size:dest_align triplets.  alg
25515           is specified in -mstringop-strategy, max_size specifies the max
25516           byte size with which inline algorithm alg is allowed.  For the last
25517           triplet, the max_size must be "-1". The max_size of the triplets in
25518           the list must be specified in increasing order.  The minimal byte
25519           size for alg is 0 for the first triplet and "max_size + 1" of the
25520           preceding range.
25521
25522       -mmemset-strategy=strategy
25523           The option is similar to -mmemcpy-strategy= except that it is to
25524           control "__builtin_memset" expansion.
25525
25526       -momit-leaf-frame-pointer
25527           Don't keep the frame pointer in a register for leaf functions.
25528           This avoids the instructions to save, set up, and restore frame
25529           pointers and makes an extra register available in leaf functions.
25530           The option -fomit-leaf-frame-pointer removes the frame pointer for
25531           leaf functions, which might make debugging harder.
25532
25533       -mtls-direct-seg-refs
25534       -mno-tls-direct-seg-refs
25535           Controls whether TLS variables may be accessed with offsets from
25536           the TLS segment register (%gs for 32-bit, %fs for 64-bit), or
25537           whether the thread base pointer must be added.  Whether or not this
25538           is valid depends on the operating system, and whether it maps the
25539           segment to cover the entire TLS area.
25540
25541           For systems that use the GNU C Library, the default is on.
25542
25543       -msse2avx
25544       -mno-sse2avx
25545           Specify that the assembler should encode SSE instructions with VEX
25546           prefix.  The option -mavx turns this on by default.
25547
25548       -mfentry
25549       -mno-fentry
25550           If profiling is active (-pg), put the profiling counter call before
25551           the prologue.  Note: On x86 architectures the attribute
25552           "ms_hook_prologue" isn't possible at the moment for -mfentry and
25553           -pg.
25554
25555       -mrecord-mcount
25556       -mno-record-mcount
25557           If profiling is active (-pg), generate a __mcount_loc section that
25558           contains pointers to each profiling call. This is useful for
25559           automatically patching and out calls.
25560
25561       -mnop-mcount
25562       -mno-nop-mcount
25563           If profiling is active (-pg), generate the calls to the profiling
25564           functions as NOPs. This is useful when they should be patched in
25565           later dynamically. This is likely only useful together with
25566           -mrecord-mcount.
25567
25568       -minstrument-return=type
25569           Instrument function exit in -pg -mfentry instrumented functions
25570           with call to specified function. This only instruments true returns
25571           ending with ret, but not sibling calls ending with jump. Valid
25572           types are none to not instrument, call to generate a call to
25573           __return__, or nop5 to generate a 5 byte nop.
25574
25575       -mrecord-return
25576       -mno-record-return
25577           Generate a __return_loc section pointing to all return
25578           instrumentation code.
25579
25580       -mfentry-name=name
25581           Set name of __fentry__ symbol called at function entry for -pg
25582           -mfentry functions.
25583
25584       -mfentry-section=name
25585           Set name of section to record -mrecord-mcount calls (default
25586           __mcount_loc).
25587
25588       -mskip-rax-setup
25589       -mno-skip-rax-setup
25590           When generating code for the x86-64 architecture with SSE
25591           extensions disabled, -mskip-rax-setup can be used to skip setting
25592           up RAX register when there are no variable arguments passed in
25593           vector registers.
25594
25595           Warning: Since RAX register is used to avoid unnecessarily saving
25596           vector registers on stack when passing variable arguments, the
25597           impacts of this option are callees may waste some stack space,
25598           misbehave or jump to a random location.  GCC 4.4 or newer don't
25599           have those issues, regardless the RAX register value.
25600
25601       -m8bit-idiv
25602       -mno-8bit-idiv
25603           On some processors, like Intel Atom, 8-bit unsigned integer divide
25604           is much faster than 32-bit/64-bit integer divide.  This option
25605           generates a run-time check.  If both dividend and divisor are
25606           within range of 0 to 255, 8-bit unsigned integer divide is used
25607           instead of 32-bit/64-bit integer divide.
25608
25609       -mavx256-split-unaligned-load
25610       -mavx256-split-unaligned-store
25611           Split 32-byte AVX unaligned load and store.
25612
25613       -mstack-protector-guard=guard
25614       -mstack-protector-guard-reg=reg
25615       -mstack-protector-guard-offset=offset
25616           Generate stack protection code using canary at guard.  Supported
25617           locations are global for global canary or tls for per-thread canary
25618           in the TLS block (the default).  This option has effect only when
25619           -fstack-protector or -fstack-protector-all is specified.
25620
25621           With the latter choice the options -mstack-protector-guard-reg=reg
25622           and -mstack-protector-guard-offset=offset furthermore specify which
25623           segment register (%fs or %gs) to use as base register for reading
25624           the canary, and from what offset from that base register.  The
25625           default for those is as specified in the relevant ABI.
25626
25627       -mgeneral-regs-only
25628           Generate code that uses only the general-purpose registers.  This
25629           prevents the compiler from using floating-point, vector, mask and
25630           bound registers.
25631
25632       -mindirect-branch=choice
25633           Convert indirect call and jump with choice.  The default is keep,
25634           which keeps indirect call and jump unmodified.  thunk converts
25635           indirect call and jump to call and return thunk.  thunk-inline
25636           converts indirect call and jump to inlined call and return thunk.
25637           thunk-extern converts indirect call and jump to external call and
25638           return thunk provided in a separate object file.  You can control
25639           this behavior for a specific function by using the function
25640           attribute "indirect_branch".
25641
25642           Note that -mcmodel=large is incompatible with
25643           -mindirect-branch=thunk and -mindirect-branch=thunk-extern since
25644           the thunk function may not be reachable in the large code model.
25645
25646           Note that -mindirect-branch=thunk-extern is compatible with
25647           -fcf-protection=branch since the external thunk can be made to
25648           enable control-flow check.
25649
25650       -mfunction-return=choice
25651           Convert function return with choice.  The default is keep, which
25652           keeps function return unmodified.  thunk converts function return
25653           to call and return thunk.  thunk-inline converts function return to
25654           inlined call and return thunk.  thunk-extern converts function
25655           return to external call and return thunk provided in a separate
25656           object file.  You can control this behavior for a specific function
25657           by using the function attribute "function_return".
25658
25659           Note that -mindirect-return=thunk-extern is compatible with
25660           -fcf-protection=branch since the external thunk can be made to
25661           enable control-flow check.
25662
25663           Note that -mcmodel=large is incompatible with
25664           -mfunction-return=thunk and -mfunction-return=thunk-extern since
25665           the thunk function may not be reachable in the large code model.
25666
25667       -mindirect-branch-register
25668           Force indirect call and jump via register.
25669
25670       These -m switches are supported in addition to the above on x86-64
25671       processors in 64-bit environments.
25672
25673       -m32
25674       -m64
25675       -mx32
25676       -m16
25677       -miamcu
25678           Generate code for a 16-bit, 32-bit or 64-bit environment.  The -m32
25679           option sets "int", "long", and pointer types to 32 bits, and
25680           generates code that runs on any i386 system.
25681
25682           The -m64 option sets "int" to 32 bits and "long" and pointer types
25683           to 64 bits, and generates code for the x86-64 architecture.  For
25684           Darwin only the -m64 option also turns off the -fno-pic and
25685           -mdynamic-no-pic options.
25686
25687           The -mx32 option sets "int", "long", and pointer types to 32 bits,
25688           and generates code for the x86-64 architecture.
25689
25690           The -m16 option is the same as -m32, except for that it outputs the
25691           ".code16gcc" assembly directive at the beginning of the assembly
25692           output so that the binary can run in 16-bit mode.
25693
25694           The -miamcu option generates code which conforms to Intel MCU
25695           psABI.  It requires the -m32 option to be turned on.
25696
25697       -mno-red-zone
25698           Do not use a so-called "red zone" for x86-64 code.  The red zone is
25699           mandated by the x86-64 ABI; it is a 128-byte area beyond the
25700           location of the stack pointer that is not modified by signal or
25701           interrupt handlers and therefore can be used for temporary data
25702           without adjusting the stack pointer.  The flag -mno-red-zone
25703           disables this red zone.
25704
25705       -mcmodel=small
25706           Generate code for the small code model: the program and its symbols
25707           must be linked in the lower 2 GB of the address space.  Pointers
25708           are 64 bits.  Programs can be statically or dynamically linked.
25709           This is the default code model.
25710
25711       -mcmodel=kernel
25712           Generate code for the kernel code model.  The kernel runs in the
25713           negative 2 GB of the address space.  This model has to be used for
25714           Linux kernel code.
25715
25716       -mcmodel=medium
25717           Generate code for the medium model: the program is linked in the
25718           lower 2 GB of the address space.  Small symbols are also placed
25719           there.  Symbols with sizes larger than -mlarge-data-threshold are
25720           put into large data or BSS sections and can be located above 2GB.
25721           Programs can be statically or dynamically linked.
25722
25723       -mcmodel=large
25724           Generate code for the large model.  This model makes no assumptions
25725           about addresses and sizes of sections.
25726
25727       -maddress-mode=long
25728           Generate code for long address mode.  This is only supported for
25729           64-bit and x32 environments.  It is the default address mode for
25730           64-bit environments.
25731
25732       -maddress-mode=short
25733           Generate code for short address mode.  This is only supported for
25734           32-bit and x32 environments.  It is the default address mode for
25735           32-bit and x32 environments.
25736
25737   x86 Windows Options
25738       These additional options are available for Microsoft Windows targets:
25739
25740       -mconsole
25741           This option specifies that a console application is to be
25742           generated, by instructing the linker to set the PE header subsystem
25743           type required for console applications.  This option is available
25744           for Cygwin and MinGW targets and is enabled by default on those
25745           targets.
25746
25747       -mdll
25748           This option is available for Cygwin and MinGW targets.  It
25749           specifies that a DLL---a dynamic link library---is to be generated,
25750           enabling the selection of the required runtime startup object and
25751           entry point.
25752
25753       -mnop-fun-dllimport
25754           This option is available for Cygwin and MinGW targets.  It
25755           specifies that the "dllimport" attribute should be ignored.
25756
25757       -mthread
25758           This option is available for MinGW targets. It specifies that
25759           MinGW-specific thread support is to be used.
25760
25761       -municode
25762           This option is available for MinGW-w64 targets.  It causes the
25763           "UNICODE" preprocessor macro to be predefined, and chooses Unicode-
25764           capable runtime startup code.
25765
25766       -mwin32
25767           This option is available for Cygwin and MinGW targets.  It
25768           specifies that the typical Microsoft Windows predefined macros are
25769           to be set in the pre-processor, but does not influence the choice
25770           of runtime library/startup code.
25771
25772       -mwindows
25773           This option is available for Cygwin and MinGW targets.  It
25774           specifies that a GUI application is to be generated by instructing
25775           the linker to set the PE header subsystem type appropriately.
25776
25777       -fno-set-stack-executable
25778           This option is available for MinGW targets. It specifies that the
25779           executable flag for the stack used by nested functions isn't set.
25780           This is necessary for binaries running in kernel mode of Microsoft
25781           Windows, as there the User32 API, which is used to set executable
25782           privileges, isn't available.
25783
25784       -fwritable-relocated-rdata
25785           This option is available for MinGW and Cygwin targets.  It
25786           specifies that relocated-data in read-only section is put into the
25787           ".data" section.  This is a necessary for older runtimes not
25788           supporting modification of ".rdata" sections for pseudo-relocation.
25789
25790       -mpe-aligned-commons
25791           This option is available for Cygwin and MinGW targets.  It
25792           specifies that the GNU extension to the PE file format that permits
25793           the correct alignment of COMMON variables should be used when
25794           generating code.  It is enabled by default if GCC detects that the
25795           target assembler found during configuration supports the feature.
25796
25797       See also under x86 Options for standard options.
25798
25799   Xstormy16 Options
25800       These options are defined for Xstormy16:
25801
25802       -msim
25803           Choose startup files and linker script suitable for the simulator.
25804
25805   Xtensa Options
25806       These options are supported for Xtensa targets:
25807
25808       -mconst16
25809       -mno-const16
25810           Enable or disable use of "CONST16" instructions for loading
25811           constant values.  The "CONST16" instruction is currently not a
25812           standard option from Tensilica.  When enabled, "CONST16"
25813           instructions are always used in place of the standard "L32R"
25814           instructions.  The use of "CONST16" is enabled by default only if
25815           the "L32R" instruction is not available.
25816
25817       -mfused-madd
25818       -mno-fused-madd
25819           Enable or disable use of fused multiply/add and multiply/subtract
25820           instructions in the floating-point option.  This has no effect if
25821           the floating-point option is not also enabled.  Disabling fused
25822           multiply/add and multiply/subtract instructions forces the compiler
25823           to use separate instructions for the multiply and add/subtract
25824           operations.  This may be desirable in some cases where strict IEEE
25825           754-compliant results are required: the fused multiply add/subtract
25826           instructions do not round the intermediate result, thereby
25827           producing results with more bits of precision than specified by the
25828           IEEE standard.  Disabling fused multiply add/subtract instructions
25829           also ensures that the program output is not sensitive to the
25830           compiler's ability to combine multiply and add/subtract operations.
25831
25832       -mserialize-volatile
25833       -mno-serialize-volatile
25834           When this option is enabled, GCC inserts "MEMW" instructions before
25835           "volatile" memory references to guarantee sequential consistency.
25836           The default is -mserialize-volatile.  Use -mno-serialize-volatile
25837           to omit the "MEMW" instructions.
25838
25839       -mforce-no-pic
25840           For targets, like GNU/Linux, where all user-mode Xtensa code must
25841           be position-independent code (PIC), this option disables PIC for
25842           compiling kernel code.
25843
25844       -mtext-section-literals
25845       -mno-text-section-literals
25846           These options control the treatment of literal pools.  The default
25847           is -mno-text-section-literals, which places literals in a separate
25848           section in the output file.  This allows the literal pool to be
25849           placed in a data RAM/ROM, and it also allows the linker to combine
25850           literal pools from separate object files to remove redundant
25851           literals and improve code size.  With -mtext-section-literals, the
25852           literals are interspersed in the text section in order to keep them
25853           as close as possible to their references.  This may be necessary
25854           for large assembly files.  Literals for each function are placed
25855           right before that function.
25856
25857       -mauto-litpools
25858       -mno-auto-litpools
25859           These options control the treatment of literal pools.  The default
25860           is -mno-auto-litpools, which places literals in a separate section
25861           in the output file unless -mtext-section-literals is used.  With
25862           -mauto-litpools the literals are interspersed in the text section
25863           by the assembler.  Compiler does not produce explicit ".literal"
25864           directives and loads literals into registers with "MOVI"
25865           instructions instead of "L32R" to let the assembler do relaxation
25866           and place literals as necessary.  This option allows assembler to
25867           create several literal pools per function and assemble very big
25868           functions, which may not be possible with -mtext-section-literals.
25869
25870       -mtarget-align
25871       -mno-target-align
25872           When this option is enabled, GCC instructs the assembler to
25873           automatically align instructions to reduce branch penalties at the
25874           expense of some code density.  The assembler attempts to widen
25875           density instructions to align branch targets and the instructions
25876           following call instructions.  If there are not enough preceding
25877           safe density instructions to align a target, no widening is
25878           performed.  The default is -mtarget-align.  These options do not
25879           affect the treatment of auto-aligned instructions like "LOOP",
25880           which the assembler always aligns, either by widening density
25881           instructions or by inserting NOP instructions.
25882
25883       -mlongcalls
25884       -mno-longcalls
25885           When this option is enabled, GCC instructs the assembler to
25886           translate direct calls to indirect calls unless it can determine
25887           that the target of a direct call is in the range allowed by the
25888           call instruction.  This translation typically occurs for calls to
25889           functions in other source files.  Specifically, the assembler
25890           translates a direct "CALL" instruction into an "L32R" followed by a
25891           "CALLX" instruction.  The default is -mno-longcalls.  This option
25892           should be used in programs where the call target can potentially be
25893           out of range.  This option is implemented in the assembler, not the
25894           compiler, so the assembly code generated by GCC still shows direct
25895           call instructions---look at the disassembled object code to see the
25896           actual instructions.  Note that the assembler uses an indirect call
25897           for every cross-file call, not just those that really are out of
25898           range.
25899
25900   zSeries Options
25901       These are listed under
25902

ENVIRONMENT

25904       This section describes several environment variables that affect how
25905       GCC operates.  Some of them work by specifying directories or prefixes
25906       to use when searching for various kinds of files.  Some are used to
25907       specify other aspects of the compilation environment.
25908
25909       Note that you can also specify places to search using options such as
25910       -B, -I and -L.  These take precedence over places specified using
25911       environment variables, which in turn take precedence over those
25912       specified by the configuration of GCC.
25913
25914       LANG
25915       LC_CTYPE
25916       LC_MESSAGES
25917       LC_ALL
25918           These environment variables control the way that GCC uses
25919           localization information which allows GCC to work with different
25920           national conventions.  GCC inspects the locale categories LC_CTYPE
25921           and LC_MESSAGES if it has been configured to do so.  These locale
25922           categories can be set to any value supported by your installation.
25923           A typical value is en_GB.UTF-8 for English in the United Kingdom
25924           encoded in UTF-8.
25925
25926           The LC_CTYPE environment variable specifies character
25927           classification.  GCC uses it to determine the character boundaries
25928           in a string; this is needed for some multibyte encodings that
25929           contain quote and escape characters that are otherwise interpreted
25930           as a string end or escape.
25931
25932           The LC_MESSAGES environment variable specifies the language to use
25933           in diagnostic messages.
25934
25935           If the LC_ALL environment variable is set, it overrides the value
25936           of LC_CTYPE and LC_MESSAGES; otherwise, LC_CTYPE and LC_MESSAGES
25937           default to the value of the LANG environment variable.  If none of
25938           these variables are set, GCC defaults to traditional C English
25939           behavior.
25940
25941       TMPDIR
25942           If TMPDIR is set, it specifies the directory to use for temporary
25943           files.  GCC uses temporary files to hold the output of one stage of
25944           compilation which is to be used as input to the next stage: for
25945           example, the output of the preprocessor, which is the input to the
25946           compiler proper.
25947
25948       GCC_COMPARE_DEBUG
25949           Setting GCC_COMPARE_DEBUG is nearly equivalent to passing
25950           -fcompare-debug to the compiler driver.  See the documentation of
25951           this option for more details.
25952
25953       GCC_EXEC_PREFIX
25954           If GCC_EXEC_PREFIX is set, it specifies a prefix to use in the
25955           names of the subprograms executed by the compiler.  No slash is
25956           added when this prefix is combined with the name of a subprogram,
25957           but you can specify a prefix that ends with a slash if you wish.
25958
25959           If GCC_EXEC_PREFIX is not set, GCC attempts to figure out an
25960           appropriate prefix to use based on the pathname it is invoked with.
25961
25962           If GCC cannot find the subprogram using the specified prefix, it
25963           tries looking in the usual places for the subprogram.
25964
25965           The default value of GCC_EXEC_PREFIX is prefix/lib/gcc/ where
25966           prefix is the prefix to the installed compiler. In many cases
25967           prefix is the value of "prefix" when you ran the configure script.
25968
25969           Other prefixes specified with -B take precedence over this prefix.
25970
25971           This prefix is also used for finding files such as crt0.o that are
25972           used for linking.
25973
25974           In addition, the prefix is used in an unusual way in finding the
25975           directories to search for header files.  For each of the standard
25976           directories whose name normally begins with /usr/local/lib/gcc
25977           (more precisely, with the value of GCC_INCLUDE_DIR), GCC tries
25978           replacing that beginning with the specified prefix to produce an
25979           alternate directory name.  Thus, with -Bfoo/, GCC searches foo/bar
25980           just before it searches the standard directory /usr/local/lib/bar.
25981           If a standard directory begins with the configured prefix then the
25982           value of prefix is replaced by GCC_EXEC_PREFIX when looking for
25983           header files.
25984
25985       COMPILER_PATH
25986           The value of COMPILER_PATH is a colon-separated list of
25987           directories, much like PATH.  GCC tries the directories thus
25988           specified when searching for subprograms, if it cannot find the
25989           subprograms using GCC_EXEC_PREFIX.
25990
25991       LIBRARY_PATH
25992           The value of LIBRARY_PATH is a colon-separated list of directories,
25993           much like PATH.  When configured as a native compiler, GCC tries
25994           the directories thus specified when searching for special linker
25995           files, if it cannot find them using GCC_EXEC_PREFIX.  Linking using
25996           GCC also uses these directories when searching for ordinary
25997           libraries for the -l option (but directories specified with -L come
25998           first).
25999
26000       LANG
26001           This variable is used to pass locale information to the compiler.
26002           One way in which this information is used is to determine the
26003           character set to be used when character literals, string literals
26004           and comments are parsed in C and C++.  When the compiler is
26005           configured to allow multibyte characters, the following values for
26006           LANG are recognized:
26007
26008           C-JIS
26009               Recognize JIS characters.
26010
26011           C-SJIS
26012               Recognize SJIS characters.
26013
26014           C-EUCJP
26015               Recognize EUCJP characters.
26016
26017           If LANG is not defined, or if it has some other value, then the
26018           compiler uses "mblen" and "mbtowc" as defined by the default locale
26019           to recognize and translate multibyte characters.
26020
26021       Some additional environment variables affect the behavior of the
26022       preprocessor.
26023
26024       CPATH
26025       C_INCLUDE_PATH
26026       CPLUS_INCLUDE_PATH
26027       OBJC_INCLUDE_PATH
26028           Each variable's value is a list of directories separated by a
26029           special character, much like PATH, in which to look for header
26030           files.  The special character, "PATH_SEPARATOR", is target-
26031           dependent and determined at GCC build time.  For Microsoft Windows-
26032           based targets it is a semicolon, and for almost all other targets
26033           it is a colon.
26034
26035           CPATH specifies a list of directories to be searched as if
26036           specified with -I, but after any paths given with -I options on the
26037           command line.  This environment variable is used regardless of
26038           which language is being preprocessed.
26039
26040           The remaining environment variables apply only when preprocessing
26041           the particular language indicated.  Each specifies a list of
26042           directories to be searched as if specified with -isystem, but after
26043           any paths given with -isystem options on the command line.
26044
26045           In all these variables, an empty element instructs the compiler to
26046           search its current working directory.  Empty elements can appear at
26047           the beginning or end of a path.  For instance, if the value of
26048           CPATH is ":/special/include", that has the same effect as
26049           -I. -I/special/include.
26050
26051       DEPENDENCIES_OUTPUT
26052           If this variable is set, its value specifies how to output
26053           dependencies for Make based on the non-system header files
26054           processed by the compiler.  System header files are ignored in the
26055           dependency output.
26056
26057           The value of DEPENDENCIES_OUTPUT can be just a file name, in which
26058           case the Make rules are written to that file, guessing the target
26059           name from the source file name.  Or the value can have the form
26060           file target, in which case the rules are written to file file using
26061           target as the target name.
26062
26063           In other words, this environment variable is equivalent to
26064           combining the options -MM and -MF, with an optional -MT switch too.
26065
26066       SUNPRO_DEPENDENCIES
26067           This variable is the same as DEPENDENCIES_OUTPUT (see above),
26068           except that system header files are not ignored, so it implies -M
26069           rather than -MM.  However, the dependence on the main input file is
26070           omitted.
26071
26072       SOURCE_DATE_EPOCH
26073           If this variable is set, its value specifies a UNIX timestamp to be
26074           used in replacement of the current date and time in the "__DATE__"
26075           and "__TIME__" macros, so that the embedded timestamps become
26076           reproducible.
26077
26078           The value of SOURCE_DATE_EPOCH must be a UNIX timestamp, defined as
26079           the number of seconds (excluding leap seconds) since 01 Jan 1970
26080           00:00:00 represented in ASCII; identical to the output of
26081           @command{date +%s} on GNU/Linux and other systems that support the
26082           %s extension in the "date" command.
26083
26084           The value should be a known timestamp such as the last modification
26085           time of the source or package and it should be set by the build
26086           process.
26087

BUGS

26089       For instructions on reporting bugs, see
26090       <http://bugzilla.redhat.com/bugzilla>.
26091

FOOTNOTES

26093       1.  On some systems, gcc -shared needs to build supplementary stub code
26094           for constructors to work.  On multi-libbed systems, gcc -shared
26095           must select the correct support libraries to link against.  Failing
26096           to supply the correct flags may lead to subtle defects.  Supplying
26097           them in cases where they are not necessary is innocuous.
26098

SEE ALSO

26100       gpl(7), gfdl(7), fsf-funding(7), cpp(1), gcov(1), as(1), ld(1), gdb(1),
26101       dbx(1) and the Info entries for gcc, cpp, as, ld, binutils and gdb.
26102

AUTHOR

26104       See the Info entry for gcc, or
26105       <http://gcc.gnu.org/onlinedocs/gcc/Contributors.html>, for contributors
26106       to GCC.
26107
26109       Copyright (c) 1988-2020 Free Software Foundation, Inc.
26110
26111       Permission is granted to copy, distribute and/or modify this document
26112       under the terms of the GNU Free Documentation License, Version 1.3 or
26113       any later version published by the Free Software Foundation; with the
26114       Invariant Sections being "GNU General Public License" and "Funding Free
26115       Software", the Front-Cover texts being (a) (see below), and with the
26116       Back-Cover Texts being (b) (see below).  A copy of the license is
26117       included in the gfdl(7) man page.
26118
26119       (a) The FSF's Front-Cover Text is:
26120
26121            A GNU Manual
26122
26123       (b) The FSF's Back-Cover Text is:
26124
26125            You have freedom to copy and modify this GNU Manual, like GNU
26126            software.  Copies published by the Free Software Foundation raise
26127            funds for GNU development.
26128
26129
26130
26131gcc-10                            2020-11-25                            GCC(1)
Impressum