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-loop-limit=n -fconstexpr-ops-limit=n
101           -fno-elide-constructors -fno-enforce-eh-specs -fno-gnu-keywords
102           -fno-implicit-templates -fno-implicit-inline-templates
103           -fno-implement-inlines  -fms-extensions -fnew-inheriting-ctors
104           -fnew-ttp-matching -fno-nonansi-builtins  -fnothrow-opt
105           -fno-operator-names -fno-optional-diags  -fpermissive
106           -fno-pretty-templates -frepo  -fno-rtti  -fsized-deallocation
107           -ftemplate-backtrace-limit=n -ftemplate-depth=n
108           -fno-threadsafe-statics  -fuse-cxa-atexit -fno-weak  -nostdinc++
109           -fvisibility-inlines-hidden -fvisibility-ms-compat
110           -fext-numeric-literals -Wabi=n  -Wabi-tag  -Wconversion-null
111           -Wctor-dtor-privacy -Wdelete-non-virtual-dtor  -Wdeprecated-copy
112           -Wdeprecated-copy-dtor -Wliteral-suffix -Wmultiple-inheritance
113           -Wno-init-list-lifetime -Wnamespaces  -Wnarrowing
114           -Wpessimizing-move  -Wredundant-move -Wnoexcept  -Wnoexcept-type
115           -Wclass-memaccess -Wnon-virtual-dtor  -Wreorder  -Wregister
116           -Weffc++  -Wstrict-null-sentinel  -Wtemplates
117           -Wno-non-template-friend  -Wold-style-cast -Woverloaded-virtual
118           -Wno-pmf-conversions -Wno-class-conversion  -Wno-terminate
119           -Wsign-promo  -Wvirtual-inheritance
120
121       Objective-C and Objective-C++ Language Options
122           -fconstant-string-class=class-name -fgnu-runtime  -fnext-runtime
123           -fno-nil-receivers -fobjc-abi-version=n -fobjc-call-cxx-cdtors
124           -fobjc-direct-dispatch -fobjc-exceptions -fobjc-gc -fobjc-nilcheck
125           -fobjc-std=objc1 -fno-local-ivars
126           -fivar-visibility=[public|protected|private|package]
127           -freplace-objc-classes -fzero-link -gen-decls -Wassign-intercept
128           -Wno-protocol  -Wselector -Wstrict-selector-match
129           -Wundeclared-selector
130
131       Diagnostic Message Formatting Options
132           -fmessage-length=n -fdiagnostics-show-location=[once|every-line]
133           -fdiagnostics-color=[auto|never|always]
134           -fdiagnostics-format=[text|json] -fno-diagnostics-show-option
135           -fno-diagnostics-show-caret -fno-diagnostics-show-labels
136           -fno-diagnostics-show-line-numbers
137           -fdiagnostics-minimum-margin-width=width
138           -fdiagnostics-parseable-fixits  -fdiagnostics-generate-patch
139           -fdiagnostics-show-template-tree  -fno-elide-type -fno-show-column
140
141       Warning Options
142           -fsyntax-only  -fmax-errors=n  -Wpedantic -pedantic-errors -w
143           -Wextra  -Wall  -Waddress  -Waddress-of-packed-member
144           -Waggregate-return  -Waligned-new -Walloc-zero
145           -Walloc-size-larger-than=byte-size -Walloca
146           -Walloca-larger-than=byte-size -Wno-aggressive-loop-optimizations
147           -Warray-bounds  -Warray-bounds=n -Wno-attributes
148           -Wattribute-alias=n -Wbool-compare  -Wbool-operation
149           -Wno-builtin-declaration-mismatch -Wno-builtin-macro-redefined
150           -Wc90-c99-compat  -Wc99-c11-compat -Wc++-compat  -Wc++11-compat
151           -Wc++14-compat  -Wc++17-compat -Wcast-align  -Wcast-align=strict
152           -Wcast-function-type  -Wcast-qual -Wchar-subscripts  -Wcatch-value
153           -Wcatch-value=n -Wclobbered  -Wcomment  -Wconditionally-supported
154           -Wconversion  -Wcoverage-mismatch  -Wno-cpp  -Wdangling-else
155           -Wdate-time -Wdelete-incomplete -Wno-attribute-warning
156           -Wno-deprecated  -Wno-deprecated-declarations  -Wno-designated-init
157           -Wdisabled-optimization -Wno-discarded-qualifiers
158           -Wno-discarded-array-qualifiers -Wno-div-by-zero
159           -Wdouble-promotion -Wduplicated-branches  -Wduplicated-cond
160           -Wempty-body  -Wenum-compare  -Wno-endif-labels
161           -Wexpansion-to-defined -Werror  -Werror=*  -Wextra-semi
162           -Wfatal-errors -Wfloat-equal  -Wformat  -Wformat=2
163           -Wno-format-contains-nul  -Wno-format-extra-args
164           -Wformat-nonliteral  -Wformat-overflow=n -Wformat-security
165           -Wformat-signedness  -Wformat-truncation=n -Wformat-y2k
166           -Wframe-address -Wframe-larger-than=byte-size
167           -Wno-free-nonheap-object -Wjump-misses-init -Whsa  -Wif-not-aligned
168           -Wignored-qualifiers  -Wignored-attributes
169           -Wincompatible-pointer-types -Wimplicit  -Wimplicit-fallthrough
170           -Wimplicit-fallthrough=n -Wimplicit-function-declaration
171           -Wimplicit-int -Winit-self  -Winline  -Wno-int-conversion
172           -Wint-in-bool-context -Wno-int-to-pointer-cast
173           -Winvalid-memory-model  -Wno-invalid-offsetof -Winvalid-pch
174           -Wlarger-than=byte-size -Wlogical-op  -Wlogical-not-parentheses
175           -Wlong-long -Wmain  -Wmaybe-uninitialized  -Wmemset-elt-size
176           -Wmemset-transposed-args -Wmisleading-indentation
177           -Wmissing-attributes  -Wmissing-braces -Wmissing-field-initializers
178           -Wmissing-format-attribute -Wmissing-include-dirs
179           -Wmissing-noreturn  -Wmissing-profile -Wno-multichar
180           -Wmultistatement-macros  -Wnonnull  -Wnonnull-compare
181           -Wnormalized=[none|id|nfc|nfkc] -Wnull-dereference  -Wodr
182           -Wno-overflow  -Wopenmp-simd -Woverride-init-side-effects
183           -Woverlength-strings -Wpacked  -Wpacked-bitfield-compat
184           -Wpacked-not-aligned  -Wpadded -Wparentheses
185           -Wno-pedantic-ms-format -Wplacement-new  -Wplacement-new=n
186           -Wpointer-arith  -Wpointer-compare  -Wno-pointer-to-int-cast
187           -Wno-pragmas  -Wno-prio-ctor-dtor  -Wredundant-decls -Wrestrict
188           -Wno-return-local-addr -Wreturn-type  -Wsequence-point  -Wshadow
189           -Wno-shadow-ivar -Wshadow=global,  -Wshadow=local,
190           -Wshadow=compatible-local -Wshift-overflow  -Wshift-overflow=n
191           -Wshift-count-negative  -Wshift-count-overflow
192           -Wshift-negative-value -Wsign-compare  -Wsign-conversion
193           -Wfloat-conversion -Wno-scalar-storage-order  -Wsizeof-pointer-div
194           -Wsizeof-pointer-memaccess  -Wsizeof-array-argument
195           -Wstack-protector  -Wstack-usage=byte-size  -Wstrict-aliasing
196           -Wstrict-aliasing=n  -Wstrict-overflow  -Wstrict-overflow=n
197           -Wstringop-overflow=n  -Wstringop-truncation  -Wsubobject-linkage
198           -Wsuggest-attribute=[pure|const|noreturn|format|malloc]
199           -Wsuggest-final-types   -Wsuggest-final-methods  -Wsuggest-override
200           -Wswitch  -Wswitch-bool  -Wswitch-default  -Wswitch-enum
201           -Wswitch-unreachable  -Wsync-nand -Wsystem-headers
202           -Wtautological-compare  -Wtrampolines  -Wtrigraphs -Wtype-limits
203           -Wundef -Wuninitialized  -Wunknown-pragmas
204           -Wunsuffixed-float-constants  -Wunused  -Wunused-function
205           -Wunused-label  -Wunused-local-typedefs  -Wunused-macros
206           -Wunused-parameter  -Wno-unused-result -Wunused-value
207           -Wunused-variable -Wunused-const-variable
208           -Wunused-const-variable=n -Wunused-but-set-parameter
209           -Wunused-but-set-variable -Wuseless-cast  -Wvariadic-macros
210           -Wvector-operation-performance -Wvla  -Wvla-larger-than=byte-size
211           -Wvolatile-register-var -Wwrite-strings
212           -Wzero-as-null-pointer-constant
213
214       C and Objective-C-only Warning Options
215           -Wbad-function-cast  -Wmissing-declarations
216           -Wmissing-parameter-type  -Wmissing-prototypes  -Wnested-externs
217           -Wold-style-declaration  -Wold-style-definition -Wstrict-prototypes
218           -Wtraditional  -Wtraditional-conversion
219           -Wdeclaration-after-statement  -Wpointer-sign
220
221       Debugging Options
222           -g  -glevel  -gdwarf  -gdwarf-version -ggdb  -grecord-gcc-switches
223           -gno-record-gcc-switches -gstabs  -gstabs+  -gstrict-dwarf
224           -gno-strict-dwarf -gas-loc-support  -gno-as-loc-support
225           -gas-locview-support  -gno-as-locview-support -gcolumn-info
226           -gno-column-info -gstatement-frontiers  -gno-statement-frontiers
227           -gvariable-location-views  -gno-variable-location-views
228           -ginternal-reset-location-views  -gno-internal-reset-location-views
229           -ginline-points  -gno-inline-points -gvms  -gxcoff  -gxcoff+
230           -gz[=type] -gsplit-dwarf  -gdescribe-dies  -gno-describe-dies
231           -fdebug-prefix-map=old=new  -fdebug-types-section
232           -fno-eliminate-unused-debug-types -femit-struct-debug-baseonly
233           -femit-struct-debug-reduced -femit-struct-debug-detailed[=spec-
234           list] -feliminate-unused-debug-symbols  -femit-class-debug-always
235           -fno-merge-debug-strings  -fno-dwarf2-cfi-asm -fvar-tracking
236           -fvar-tracking-assignments
237
238       Optimization Options
239           -faggressive-loop-optimizations -falign-functions[=n[:m:[n2[:m2]]]]
240           -falign-jumps[=n[:m:[n2[:m2]]]] -falign-labels[=n[:m:[n2[:m2]]]]
241           -falign-loops[=n[:m:[n2[:m2]]]] -fassociative-math  -fauto-profile
242           -fauto-profile[=path] -fauto-inc-dec  -fbranch-probabilities
243           -fbranch-target-load-optimize  -fbranch-target-load-optimize2
244           -fbtr-bb-exclusive  -fcaller-saves -fcombine-stack-adjustments
245           -fconserve-stack -fcompare-elim  -fcprop-registers  -fcrossjumping
246           -fcse-follow-jumps  -fcse-skip-blocks  -fcx-fortran-rules
247           -fcx-limited-range -fdata-sections  -fdce  -fdelayed-branch
248           -fdelete-null-pointer-checks  -fdevirtualize
249           -fdevirtualize-speculatively -fdevirtualize-at-ltrans  -fdse
250           -fearly-inlining  -fipa-sra  -fexpensive-optimizations
251           -ffat-lto-objects -ffast-math  -ffinite-math-only  -ffloat-store
252           -fexcess-precision=style -fforward-propagate  -ffp-contract=style
253           -ffunction-sections -fgcse  -fgcse-after-reload  -fgcse-las
254           -fgcse-lm  -fgraphite-identity -fgcse-sm  -fhoist-adjacent-loads
255           -fif-conversion -fif-conversion2  -findirect-inlining
256           -finline-functions  -finline-functions-called-once
257           -finline-limit=n -finline-small-functions  -fipa-cp  -fipa-cp-clone
258           -fipa-bit-cp  -fipa-vrp  -fipa-pta  -fipa-profile  -fipa-pure-const
259           -fipa-reference  -fipa-reference-addressable -fipa-stack-alignment
260           -fipa-icf  -fira-algorithm=algorithm -flive-patching=level
261           -fira-region=region  -fira-hoist-pressure -fira-loop-pressure
262           -fno-ira-share-save-slots -fno-ira-share-spill-slots
263           -fisolate-erroneous-paths-dereference
264           -fisolate-erroneous-paths-attribute -fivopts
265           -fkeep-inline-functions  -fkeep-static-functions
266           -fkeep-static-consts  -flimit-function-alignment
267           -flive-range-shrinkage -floop-block  -floop-interchange
268           -floop-strip-mine -floop-unroll-and-jam  -floop-nest-optimize
269           -floop-parallelize-all  -flra-remat  -flto  -flto-compression-level
270           -flto-partition=alg  -fmerge-all-constants -fmerge-constants
271           -fmodulo-sched  -fmodulo-sched-allow-regmoves
272           -fmove-loop-invariants  -fno-branch-count-reg -fno-defer-pop
273           -fno-fp-int-builtin-inexact  -fno-function-cse
274           -fno-guess-branch-probability  -fno-inline  -fno-math-errno
275           -fno-peephole -fno-peephole2  -fno-printf-return-value
276           -fno-sched-interblock -fno-sched-spec  -fno-signed-zeros
277           -fno-toplevel-reorder  -fno-trapping-math
278           -fno-zero-initialized-in-bss -fomit-frame-pointer
279           -foptimize-sibling-calls -fpartial-inlining  -fpeel-loops
280           -fpredictive-commoning -fprefetch-loop-arrays -fprofile-correction
281           -fprofile-use  -fprofile-use=path  -fprofile-values
282           -fprofile-reorder-functions -freciprocal-math  -free
283           -frename-registers  -freorder-blocks
284           -freorder-blocks-algorithm=algorithm -freorder-blocks-and-partition
285           -freorder-functions -frerun-cse-after-loop
286           -freschedule-modulo-scheduled-loops -frounding-math
287           -fsave-optimization-record -fsched2-use-superblocks
288           -fsched-pressure -fsched-spec-load  -fsched-spec-load-dangerous
289           -fsched-stalled-insns-dep[=n]  -fsched-stalled-insns[=n]
290           -fsched-group-heuristic  -fsched-critical-path-heuristic
291           -fsched-spec-insn-heuristic  -fsched-rank-heuristic
292           -fsched-last-insn-heuristic  -fsched-dep-count-heuristic
293           -fschedule-fusion -fschedule-insns  -fschedule-insns2
294           -fsection-anchors -fselective-scheduling  -fselective-scheduling2
295           -fsel-sched-pipelining  -fsel-sched-pipelining-outer-loops
296           -fsemantic-interposition  -fshrink-wrap  -fshrink-wrap-separate
297           -fsignaling-nans -fsingle-precision-constant
298           -fsplit-ivs-in-unroller  -fsplit-loops -fsplit-paths
299           -fsplit-wide-types  -fssa-backprop  -fssa-phiopt -fstdarg-opt
300           -fstore-merging  -fstrict-aliasing -fthread-jumps  -ftracer
301           -ftree-bit-ccp -ftree-builtin-call-dce  -ftree-ccp  -ftree-ch
302           -ftree-coalesce-vars  -ftree-copy-prop  -ftree-dce
303           -ftree-dominator-opts -ftree-dse  -ftree-forwprop  -ftree-fre
304           -fcode-hoisting -ftree-loop-if-convert  -ftree-loop-im
305           -ftree-phiprop  -ftree-loop-distribution
306           -ftree-loop-distribute-patterns -ftree-loop-ivcanon
307           -ftree-loop-linear  -ftree-loop-optimize -ftree-loop-vectorize
308           -ftree-parallelize-loops=n  -ftree-pre  -ftree-partial-pre
309           -ftree-pta -ftree-reassoc  -ftree-scev-cprop  -ftree-sink
310           -ftree-slsr  -ftree-sra -ftree-switch-conversion  -ftree-tail-merge
311           -ftree-ter  -ftree-vectorize  -ftree-vrp  -funconstrained-commons
312           -funit-at-a-time  -funroll-all-loops  -funroll-loops
313           -funsafe-math-optimizations  -funswitch-loops -fipa-ra
314           -fvariable-expansion-in-unroller  -fvect-cost-model  -fvpt -fweb
315           -fwhole-program  -fwpa  -fuse-linker-plugin --param name=value -O
316           -O0  -O1  -O2  -O3  -Os  -Ofast  -Og
317
318       Program Instrumentation Options
319           -p  -pg  -fprofile-arcs  --coverage  -ftest-coverage
320           -fprofile-abs-path -fprofile-dir=path  -fprofile-generate
321           -fprofile-generate=path -fprofile-update=method
322           -fprofile-filter-files=regex -fprofile-exclude-files=regex
323           -fsanitize=style  -fsanitize-recover  -fsanitize-recover=style
324           -fasan-shadow-offset=number  -fsanitize-sections=s1,s2,...
325           -fsanitize-undefined-trap-on-error  -fbounds-check
326           -fcf-protection=[full|branch|return|none] -fstack-protector
327           -fstack-protector-all  -fstack-protector-strong
328           -fstack-protector-explicit  -fstack-check
329           -fstack-limit-register=reg  -fstack-limit-symbol=sym
330           -fno-stack-limit  -fsplit-stack -fvtable-verify=[std|preinit|none]
331           -fvtv-counts  -fvtv-debug -finstrument-functions
332           -finstrument-functions-exclude-function-list=sym,sym,...
333           -finstrument-functions-exclude-file-list=file,file,...
334
335       Preprocessor Options
336           -Aquestion=answer -A-question[=answer] -C  -CC  -Dmacro[=defn] -dD
337           -dI  -dM  -dN  -dU -fdebug-cpp  -fdirectives-only
338           -fdollars-in-identifiers -fexec-charset=charset
339           -fextended-identifiers -finput-charset=charset
340           -fmacro-prefix-map=old=new -fno-canonical-system-headers
341           -fpch-deps  -fpch-preprocess -fpreprocessed  -ftabstop=width
342           -ftrack-macro-expansion -fwide-exec-charset=charset
343           -fworking-directory -H  -imacros file  -include file -M  -MD  -MF
344           -MG  -MM  -MMD  -MP  -MQ  -MT -no-integrated-cpp  -P  -pthread
345           -remap -traditional  -traditional-cpp  -trigraphs -Umacro  -undef
346           -Wp,option  -Xpreprocessor option
347
348       Assembler Options
349           -Wa,option  -Xassembler option
350
351       Linker Options
352           object-file-name  -fuse-ld=linker  -llibrary -nostartfiles
353           -nodefaultlibs  -nolibc  -nostdlib -e entry  --entry=entry -pie
354           -pthread  -r  -rdynamic -s  -static  -static-pie  -static-libgcc
355           -static-libstdc++ -static-libasan  -static-libtsan  -static-liblsan
356           -static-libubsan -shared  -shared-libgcc  -symbolic -T script
357           -Wl,option  -Xlinker option -u symbol  -z keyword
358
359       Directory Options
360           -Bprefix  -Idir  -I- -idirafter dir -imacros file  -imultilib dir
361           -iplugindir=dir  -iprefix file -iquote dir  -isysroot dir  -isystem
362           dir -iwithprefix dir  -iwithprefixbefore dir -Ldir
363           -no-canonical-prefixes  --no-sysroot-suffix -nostdinc  -nostdinc++
364           --sysroot=dir
365
366       Code Generation Options
367           -fcall-saved-reg  -fcall-used-reg -ffixed-reg  -fexceptions
368           -fnon-call-exceptions  -fdelete-dead-exceptions  -funwind-tables
369           -fasynchronous-unwind-tables -fno-gnu-unique
370           -finhibit-size-directive  -fno-common  -fno-ident
371           -fpcc-struct-return  -fpic  -fPIC  -fpie  -fPIE  -fno-plt
372           -fno-jump-tables -frecord-gcc-switches -freg-struct-return
373           -fshort-enums  -fshort-wchar -fverbose-asm  -fpack-struct[=n]
374           -fleading-underscore  -ftls-model=model -fstack-reuse=reuse_level
375           -ftrampolines  -ftrapv  -fwrapv
376           -fvisibility=[default|internal|hidden|protected]
377           -fstrict-volatile-bitfields  -fsync-libcalls
378
379       Developer Options
380           -dletters  -dumpspecs  -dumpmachine  -dumpversion -dumpfullversion
381           -fchecking  -fchecking=n  -fdbg-cnt-list -fdbg-cnt=counter-value-
382           list -fdisable-ipa-pass_name -fdisable-rtl-pass_name
383           -fdisable-rtl-pass-name=range-list -fdisable-tree-pass_name
384           -fdisable-tree-pass-name=range-list -fdump-debug  -fdump-earlydebug
385           -fdump-noaddr  -fdump-unnumbered  -fdump-unnumbered-links
386           -fdump-final-insns[=file] -fdump-ipa-all  -fdump-ipa-cgraph
387           -fdump-ipa-inline -fdump-lang-all -fdump-lang-switch
388           -fdump-lang-switch-options -fdump-lang-switch-options=filename
389           -fdump-passes -fdump-rtl-pass  -fdump-rtl-pass=filename
390           -fdump-statistics -fdump-tree-all -fdump-tree-switch
391           -fdump-tree-switch-options -fdump-tree-switch-options=filename
392           -fcompare-debug[=opts]  -fcompare-debug-second -fenable-kind-pass
393           -fenable-kind-pass=range-list -fira-verbose=n -flto-report
394           -flto-report-wpa  -fmem-report-wpa -fmem-report
395           -fpre-ipa-mem-report  -fpost-ipa-mem-report -fopt-info
396           -fopt-info-options[=file] -fprofile-report -frandom-seed=string
397           -fsched-verbose=n -fsel-sched-verbose  -fsel-sched-dump-cfg
398           -fsel-sched-pipelining-verbose -fstats  -fstack-usage
399           -ftime-report  -ftime-report-details
400           -fvar-tracking-assignments-toggle  -gtoggle
401           -print-file-name=library  -print-libgcc-file-name
402           -print-multi-directory  -print-multi-lib  -print-multi-os-directory
403           -print-prog-name=program  -print-search-dirs  -Q -print-sysroot
404           -print-sysroot-headers-suffix -save-temps  -save-temps=cwd
405           -save-temps=obj  -time[=file]
406
407       Machine-Dependent Options
408           AArch64 Options -mabi=name  -mbig-endian  -mlittle-endian
409           -mgeneral-regs-only -mcmodel=tiny  -mcmodel=small  -mcmodel=large
410           -mstrict-align  -mno-strict-align -momit-leaf-frame-pointer
411           -mtls-dialect=desc  -mtls-dialect=traditional -mtls-size=size
412           -mfix-cortex-a53-835769  -mfix-cortex-a53-843419
413           -mlow-precision-recip-sqrt  -mlow-precision-sqrt
414           -mlow-precision-div -mpc-relative-literal-loads
415           -msign-return-address=scope -mbranch-protection=none|standard|pac-
416           ret[+leaf]|bti -march=name  -mcpu=name  -mtune=name
417           -moverride=string  -mverbose-cost-dump
418           -mstack-protector-guard=guard -mstack-protector-guard-reg=sysreg
419           -mstack-protector-guard-offset=offset -mtrack-speculation
420
421           Adapteva Epiphany Options -mhalf-reg-file  -mprefer-short-insn-regs
422           -mbranch-cost=num  -mcmove  -mnops=num  -msoft-cmpsf -msplit-lohi
423           -mpost-inc  -mpost-modify  -mstack-offset=num -mround-nearest
424           -mlong-calls  -mshort-calls  -msmall16 -mfp-mode=mode
425           -mvect-double  -max-vect-align=num -msplit-vecmove-early
426           -m1reg-reg
427
428           AMD GCN Options -march=gpu -mtune=gpu -mstack-size=bytes
429
430           ARC Options -mbarrel-shifter  -mjli-always -mcpu=cpu  -mA6
431           -mARC600  -mA7  -mARC700 -mdpfp  -mdpfp-compact  -mdpfp-fast
432           -mno-dpfp-lrsr -mea  -mno-mpy  -mmul32x16  -mmul64  -matomic -mnorm
433           -mspfp  -mspfp-compact  -mspfp-fast  -msimd  -msoft-float  -mswap
434           -mcrc  -mdsp-packa  -mdvbf  -mlock  -mmac-d16  -mmac-24  -mrtsc
435           -mswape -mtelephony  -mxy  -misize  -mannotate-align  -marclinux
436           -marclinux_prof -mlong-calls  -mmedium-calls  -msdata
437           -mirq-ctrl-saved -mrgf-banked-regs  -mlpc-width=width  -G num
438           -mvolatile-cache  -mtp-regno=regno -malign-call  -mauto-modify-reg
439           -mbbit-peephole  -mno-brcc -mcase-vector-pcrel  -mcompact-casesi
440           -mno-cond-exec  -mearly-cbranchsi -mexpand-adddi  -mindexed-loads
441           -mlra  -mlra-priority-none -mlra-priority-compact mlra-priority-
442           noncompact  -mmillicode -mmixed-code  -mq-class  -mRcq  -mRcw
443           -msize-level=level -mtune=cpu  -mmultcost=num  -mcode-density-frame
444           -munalign-prob-threshold=probability  -mmpy-option=multo -mdiv-rem
445           -mcode-density  -mll64  -mfpu=fpu  -mrf16  -mbranch-index
446
447           ARM Options -mapcs-frame  -mno-apcs-frame -mabi=name
448           -mapcs-stack-check  -mno-apcs-stack-check -mapcs-reentrant
449           -mno-apcs-reentrant -mgeneral-regs-only -msched-prolog
450           -mno-sched-prolog -mlittle-endian  -mbig-endian -mbe8  -mbe32
451           -mfloat-abi=name -mfp16-format=name -mthumb-interwork
452           -mno-thumb-interwork -mcpu=name  -march=name  -mfpu=name
453           -mtune=name  -mprint-tune-info -mstructure-size-boundary=n
454           -mabort-on-noreturn -mlong-calls  -mno-long-calls -msingle-pic-base
455           -mno-single-pic-base -mpic-register=reg -mnop-fun-dllimport
456           -mpoke-function-name -mthumb  -marm  -mflip-thumb -mtpcs-frame
457           -mtpcs-leaf-frame -mcaller-super-interworking
458           -mcallee-super-interworking -mtp=name  -mtls-dialect=dialect
459           -mword-relocations -mfix-cortex-m3-ldrd -munaligned-access
460           -mneon-for-64bits -mslow-flash-data -masm-syntax-unified
461           -mrestrict-it -mverbose-cost-dump -mpure-code -mcmse
462
463           AVR Options -mmcu=mcu  -mabsdata  -maccumulate-args
464           -mbranch-cost=cost -mcall-prologues  -mgas-isr-prologues  -mint8
465           -mn_flash=size  -mno-interrupts -mmain-is-OS_task  -mrelax  -mrmw
466           -mstrict-X  -mtiny-stack -mfract-convert-truncate -mshort-calls
467           -nodevicelib -Waddr-space-convert  -Wmisspelled-isr
468
469           Blackfin Options -mcpu=cpu[-sirevision] -msim
470           -momit-leaf-frame-pointer  -mno-omit-leaf-frame-pointer
471           -mspecld-anomaly  -mno-specld-anomaly  -mcsync-anomaly
472           -mno-csync-anomaly -mlow-64k  -mno-low64k  -mstack-check-l1
473           -mid-shared-library -mno-id-shared-library  -mshared-library-id=n
474           -mleaf-id-shared-library  -mno-leaf-id-shared-library -msep-data
475           -mno-sep-data  -mlong-calls  -mno-long-calls -mfast-fp
476           -minline-plt  -mmulticore  -mcorea  -mcoreb  -msdram -micplb
477
478           C6X Options -mbig-endian  -mlittle-endian  -march=cpu -msim
479           -msdata=sdata-type
480
481           CRIS Options -mcpu=cpu  -march=cpu  -mtune=cpu -mmax-stack-frame=n
482           -melinux-stacksize=n -metrax4  -metrax100  -mpdebug  -mcc-init
483           -mno-side-effects -mstack-align  -mdata-align  -mconst-align
484           -m32-bit  -m16-bit  -m8-bit  -mno-prologue-epilogue  -mno-gotplt
485           -melf  -maout  -melinux  -mlinux  -sim  -sim2 -mmul-bug-workaround
486           -mno-mul-bug-workaround
487
488           CR16 Options -mmac -mcr16cplus  -mcr16c -msim  -mint32  -mbit-ops
489           -mdata-model=model
490
491           C-SKY Options -march=arch  -mcpu=cpu -mbig-endian  -EB
492           -mlittle-endian  -EL -mhard-float  -msoft-float  -mfpu=fpu
493           -mdouble-float  -mfdivdu -melrw  -mistack  -mmp  -mcp  -mcache
494           -msecurity  -mtrust -mdsp  -medsp  -mvdsp -mdiv  -msmart
495           -mhigh-registers  -manchor -mpushpop  -mmultiple-stld  -mconstpool
496           -mstack-size  -mccrt -mbranch-cost=n  -mcse-cc  -msched-prolog
497
498           Darwin Options -all_load  -allowable_client  -arch
499           -arch_errors_fatal -arch_only  -bind_at_load  -bundle
500           -bundle_loader -client_name  -compatibility_version
501           -current_version -dead_strip -dependency-file  -dylib_file
502           -dylinker_install_name -dynamic  -dynamiclib
503           -exported_symbols_list -filelist  -flat_namespace
504           -force_cpusubtype_ALL -force_flat_namespace
505           -headerpad_max_install_names -iframework -image_base  -init
506           -install_name  -keep_private_externs -multi_module
507           -multiply_defined  -multiply_defined_unused -noall_load
508           -no_dead_strip_inits_and_terms -nofixprebinding  -nomultidefs
509           -noprebind  -noseglinkedit -pagezero_size  -prebind
510           -prebind_all_twolevel_modules -private_bundle  -read_only_relocs
511           -sectalign -sectobjectsymbols  -whyload  -seg1addr -sectcreate
512           -sectobjectsymbols  -sectorder -segaddr  -segs_read_only_addr
513           -segs_read_write_addr -seg_addr_table  -seg_addr_table_filename
514           -seglinkedit -segprot  -segs_read_only_addr  -segs_read_write_addr
515           -single_module  -static  -sub_library  -sub_umbrella
516           -twolevel_namespace  -umbrella  -undefined -unexported_symbols_list
517           -weak_reference_mismatches -whatsloaded  -F  -gused  -gfull
518           -mmacosx-version-min=version -mkernel  -mone-byte-bool
519
520           DEC Alpha Options -mno-fp-regs  -msoft-float -mieee
521           -mieee-with-inexact  -mieee-conformant -mfp-trap-mode=mode
522           -mfp-rounding-mode=mode -mtrap-precision=mode  -mbuild-constants
523           -mcpu=cpu-type  -mtune=cpu-type -mbwx  -mmax  -mfix  -mcix
524           -mfloat-vax  -mfloat-ieee -mexplicit-relocs  -msmall-data
525           -mlarge-data -msmall-text  -mlarge-text -mmemory-latency=time
526
527           FR30 Options -msmall-model  -mno-lsim
528
529           FT32 Options -msim  -mlra  -mnodiv  -mft32b  -mcompress  -mnopm
530
531           FRV Options -mgpr-32  -mgpr-64  -mfpr-32  -mfpr-64 -mhard-float
532           -msoft-float -malloc-cc  -mfixed-cc  -mdword  -mno-dword -mdouble
533           -mno-double -mmedia  -mno-media  -mmuladd  -mno-muladd -mfdpic
534           -minline-plt  -mgprel-ro  -multilib-library-pic -mlinked-fp
535           -mlong-calls  -malign-labels -mlibrary-pic  -macc-4  -macc-8 -mpack
536           -mno-pack  -mno-eflags  -mcond-move  -mno-cond-move
537           -moptimize-membar  -mno-optimize-membar -mscc  -mno-scc
538           -mcond-exec  -mno-cond-exec -mvliw-branch  -mno-vliw-branch
539           -mmulti-cond-exec  -mno-multi-cond-exec  -mnested-cond-exec
540           -mno-nested-cond-exec  -mtomcat-stats -mTLS  -mtls -mcpu=cpu
541
542           GNU/Linux Options -mglibc  -muclibc  -mmusl  -mbionic  -mandroid
543           -tno-android-cc  -tno-android-ld
544
545           H8/300 Options -mrelax  -mh  -ms  -mn  -mexr  -mno-exr  -mint32
546           -malign-300
547
548           HPPA Options -march=architecture-type -mcaller-copies
549           -mdisable-fpregs  -mdisable-indexing -mfast-indirect-calls  -mgas
550           -mgnu-ld   -mhp-ld -mfixed-range=register-range -mjump-in-delay
551           -mlinker-opt  -mlong-calls -mlong-load-store  -mno-disable-fpregs
552           -mno-disable-indexing  -mno-fast-indirect-calls  -mno-gas
553           -mno-jump-in-delay  -mno-long-load-store -mno-portable-runtime
554           -mno-soft-float -mno-space-regs  -msoft-float  -mpa-risc-1-0
555           -mpa-risc-1-1  -mpa-risc-2-0  -mportable-runtime -mschedule=cpu-
556           type  -mspace-regs  -msio  -mwsio -munix=unix-std  -nolibdld
557           -static  -threads
558
559           IA-64 Options -mbig-endian  -mlittle-endian  -mgnu-as  -mgnu-ld
560           -mno-pic -mvolatile-asm-stop  -mregister-names  -msdata  -mno-sdata
561           -mconstant-gp  -mauto-pic  -mfused-madd
562           -minline-float-divide-min-latency
563           -minline-float-divide-max-throughput -mno-inline-float-divide
564           -minline-int-divide-min-latency -minline-int-divide-max-throughput
565           -mno-inline-int-divide -minline-sqrt-min-latency
566           -minline-sqrt-max-throughput -mno-inline-sqrt -mdwarf2-asm
567           -mearly-stop-bits -mfixed-range=register-range  -mtls-size=tls-size
568           -mtune=cpu-type  -milp32  -mlp64 -msched-br-data-spec
569           -msched-ar-data-spec  -msched-control-spec -msched-br-in-data-spec
570           -msched-ar-in-data-spec  -msched-in-control-spec -msched-spec-ldc
571           -msched-spec-control-ldc -msched-prefer-non-data-spec-insns
572           -msched-prefer-non-control-spec-insns
573           -msched-stop-bits-after-every-cycle
574           -msched-count-spec-in-critical-path
575           -msel-sched-dont-check-control-spec  -msched-fp-mem-deps-zero-cost
576           -msched-max-memory-insns-hard-limit  -msched-max-memory-insns=max-
577           insns
578
579           LM32 Options -mbarrel-shift-enabled  -mdivide-enabled
580           -mmultiply-enabled -msign-extend-enabled  -muser-enabled
581
582           M32R/D Options -m32r2  -m32rx  -m32r -mdebug -malign-loops
583           -mno-align-loops -missue-rate=number -mbranch-cost=number
584           -mmodel=code-size-model-type -msdata=sdata-type -mno-flush-func
585           -mflush-func=name -mno-flush-trap  -mflush-trap=number -G num
586
587           M32C Options -mcpu=cpu  -msim  -memregs=number
588
589           M680x0 Options -march=arch  -mcpu=cpu  -mtune=tune -m68000  -m68020
590           -m68020-40  -m68020-60  -m68030  -m68040 -m68060  -mcpu32  -m5200
591           -m5206e  -m528x  -m5307  -m5407 -mcfv4e  -mbitfield  -mno-bitfield
592           -mc68000  -mc68020 -mnobitfield  -mrtd  -mno-rtd  -mdiv  -mno-div
593           -mshort -mno-short  -mhard-float  -m68881  -msoft-float  -mpcrel
594           -malign-int  -mstrict-align  -msep-data  -mno-sep-data
595           -mshared-library-id=n  -mid-shared-library  -mno-id-shared-library
596           -mxgot  -mno-xgot  -mlong-jump-table-offsets
597
598           MCore Options -mhardlit  -mno-hardlit  -mdiv  -mno-div
599           -mrelax-immediates -mno-relax-immediates  -mwide-bitfields
600           -mno-wide-bitfields -m4byte-functions  -mno-4byte-functions
601           -mcallgraph-data -mno-callgraph-data  -mslow-bytes  -mno-slow-bytes
602           -mno-lsim -mlittle-endian  -mbig-endian  -m210  -m340
603           -mstack-increment
604
605           MeP Options -mabsdiff  -mall-opts  -maverage  -mbased=n  -mbitops
606           -mc=n  -mclip  -mconfig=name  -mcop  -mcop32  -mcop64  -mivc2 -mdc
607           -mdiv  -meb  -mel  -mio-volatile  -ml  -mleadz  -mm  -mminmax
608           -mmult  -mno-opts  -mrepeat  -ms  -msatur  -msdram  -msim
609           -msimnovec  -mtf -mtiny=n
610
611           MicroBlaze Options -msoft-float  -mhard-float  -msmall-divides
612           -mcpu=cpu -mmemcpy  -mxl-soft-mul  -mxl-soft-div  -mxl-barrel-shift
613           -mxl-pattern-compare  -mxl-stack-check  -mxl-gp-opt  -mno-clearbss
614           -mxl-multiply-high  -mxl-float-convert  -mxl-float-sqrt
615           -mbig-endian  -mlittle-endian  -mxl-reorder  -mxl-mode-app-model
616           -mpic-data-is-text-relative
617
618           MIPS Options -EL  -EB  -march=arch  -mtune=arch -mips1  -mips2
619           -mips3  -mips4  -mips32  -mips32r2  -mips32r3  -mips32r5 -mips32r6
620           -mips64  -mips64r2  -mips64r3  -mips64r5  -mips64r6 -mips16
621           -mno-mips16  -mflip-mips16 -minterlink-compressed
622           -mno-interlink-compressed -minterlink-mips16  -mno-interlink-mips16
623           -mabi=abi  -mabicalls  -mno-abicalls -mshared  -mno-shared  -mplt
624           -mno-plt  -mxgot  -mno-xgot -mgp32  -mgp64  -mfp32  -mfpxx  -mfp64
625           -mhard-float  -msoft-float -mno-float  -msingle-float
626           -mdouble-float -modd-spreg  -mno-odd-spreg -mabs=mode
627           -mnan=encoding -mdsp  -mno-dsp  -mdspr2  -mno-dspr2 -mmcu
628           -mmno-mcu -meva  -mno-eva -mvirt  -mno-virt -mxpa  -mno-xpa -mcrc
629           -mno-crc -mginv  -mno-ginv -mmicromips  -mno-micromips -mmsa
630           -mno-msa -mloongson-mmi  -mno-loongson-mmi -mloongson-ext
631           -mno-loongson-ext -mloongson-ext2  -mno-loongson-ext2 -mfpu=fpu-
632           type -msmartmips  -mno-smartmips -mpaired-single
633           -mno-paired-single  -mdmx  -mno-mdmx -mips3d  -mno-mips3d  -mmt
634           -mno-mt  -mllsc  -mno-llsc -mlong64  -mlong32  -msym32  -mno-sym32
635           -Gnum  -mlocal-sdata  -mno-local-sdata -mextern-sdata
636           -mno-extern-sdata  -mgpopt  -mno-gopt -membedded-data
637           -mno-embedded-data -muninit-const-in-rodata
638           -mno-uninit-const-in-rodata -mcode-readable=setting
639           -msplit-addresses  -mno-split-addresses -mexplicit-relocs
640           -mno-explicit-relocs -mcheck-zero-division
641           -mno-check-zero-division -mdivide-traps  -mdivide-breaks
642           -mload-store-pairs  -mno-load-store-pairs -mmemcpy  -mno-memcpy
643           -mlong-calls  -mno-long-calls -mmad  -mno-mad  -mimadd  -mno-imadd
644           -mfused-madd  -mno-fused-madd  -nocpp -mfix-24k  -mno-fix-24k
645           -mfix-r4000  -mno-fix-r4000  -mfix-r4400  -mno-fix-r4400
646           -mfix-r5900  -mno-fix-r5900 -mfix-r10000  -mno-fix-r10000
647           -mfix-rm7000  -mno-fix-rm7000 -mfix-vr4120  -mno-fix-vr4120
648           -mfix-vr4130  -mno-fix-vr4130  -mfix-sb1  -mno-fix-sb1
649           -mflush-func=func  -mno-flush-func -mbranch-cost=num
650           -mbranch-likely  -mno-branch-likely -mcompact-branches=policy
651           -mfp-exceptions  -mno-fp-exceptions -mvr4130-align
652           -mno-vr4130-align  -msynci  -mno-synci -mlxc1-sxc1  -mno-lxc1-sxc1
653           -mmadd4  -mno-madd4 -mrelax-pic-calls  -mno-relax-pic-calls
654           -mmcount-ra-address -mframe-header-opt  -mno-frame-header-opt
655
656           MMIX Options -mlibfuncs  -mno-libfuncs  -mepsilon  -mno-epsilon
657           -mabi=gnu -mabi=mmixware  -mzero-extend  -mknuthdiv
658           -mtoplevel-symbols -melf  -mbranch-predict  -mno-branch-predict
659           -mbase-addresses -mno-base-addresses  -msingle-exit
660           -mno-single-exit
661
662           MN10300 Options -mmult-bug  -mno-mult-bug -mno-am33  -mam33
663           -mam33-2  -mam34 -mtune=cpu-type -mreturn-pointer-on-d0 -mno-crt0
664           -mrelax  -mliw  -msetlb
665
666           Moxie Options -meb  -mel  -mmul.x  -mno-crt0
667
668           MSP430 Options -msim  -masm-hex  -mmcu=  -mcpu=  -mlarge  -msmall
669           -mrelax -mwarn-mcu -mcode-region=  -mdata-region= -msilicon-errata=
670           -msilicon-errata-warn= -mhwmult=  -minrt
671
672           NDS32 Options -mbig-endian  -mlittle-endian -mreduced-regs
673           -mfull-regs -mcmov  -mno-cmov -mext-perf  -mno-ext-perf -mext-perf2
674           -mno-ext-perf2 -mext-string  -mno-ext-string -mv3push  -mno-v3push
675           -m16bit  -mno-16bit -misr-vector-size=num -mcache-block-size=num
676           -march=arch -mcmodel=code-model -mctor-dtor  -mrelax
677
678           Nios II Options -G num  -mgpopt=option  -mgpopt  -mno-gpopt
679           -mgprel-sec=regexp  -mr0rel-sec=regexp -mel  -meb -mno-bypass-cache
680           -mbypass-cache -mno-cache-volatile  -mcache-volatile
681           -mno-fast-sw-div  -mfast-sw-div -mhw-mul  -mno-hw-mul  -mhw-mulx
682           -mno-hw-mulx  -mno-hw-div  -mhw-div -mcustom-insn=N
683           -mno-custom-insn -mcustom-fpu-cfg=name -mhal  -msmallc
684           -msys-crt0=name  -msys-lib=name -march=arch  -mbmx  -mno-bmx  -mcdx
685           -mno-cdx
686
687           Nvidia PTX Options -m32  -m64  -mmainkernel  -moptimize
688
689           OpenRISC Options -mboard=name  -mnewlib  -mhard-mul  -mhard-div
690           -msoft-mul  -msoft-div -mcmov  -mror  -msext  -msfimm  -mshftimm
691
692           PDP-11 Options -mfpu  -msoft-float  -mac0  -mno-ac0  -m40  -m45
693           -m10 -mint32  -mno-int16  -mint16  -mno-int32 -msplit  -munix-asm
694           -mdec-asm  -mgnu-asm  -mlra
695
696           picoChip Options -mae=ae_type  -mvliw-lookahead=N
697           -msymbol-as-address  -mno-inefficient-warnings
698
699           PowerPC Options See RS/6000 and PowerPC Options.
700
701           RISC-V Options -mbranch-cost=N-instruction -mplt  -mno-plt
702           -mabi=ABI-string -mfdiv  -mno-fdiv -mdiv  -mno-div -march=ISA-
703           string -mtune=processor-string -mpreferred-stack-boundary=num
704           -msmall-data-limit=N-bytes -msave-restore  -mno-save-restore
705           -mstrict-align  -mno-strict-align -mcmodel=medlow  -mcmodel=medany
706           -mexplicit-relocs  -mno-explicit-relocs -mrelax  -mno-relax
707           -mriscv-attribute  -mmo-riscv-attribute
708
709           RL78 Options -msim  -mmul=none  -mmul=g13  -mmul=g14  -mallregs
710           -mcpu=g10  -mcpu=g13  -mcpu=g14  -mg10  -mg13  -mg14
711           -m64bit-doubles  -m32bit-doubles  -msave-mduc-in-interrupts
712
713           RS/6000 and PowerPC Options -mcpu=cpu-type -mtune=cpu-type
714           -mcmodel=code-model -mpowerpc64 -maltivec  -mno-altivec
715           -mpowerpc-gpopt  -mno-powerpc-gpopt -mpowerpc-gfxopt
716           -mno-powerpc-gfxopt -mmfcrf  -mno-mfcrf  -mpopcntb  -mno-popcntb
717           -mpopcntd  -mno-popcntd -mfprnd  -mno-fprnd -mcmpb  -mno-cmpb
718           -mmfpgpr  -mno-mfpgpr  -mhard-dfp  -mno-hard-dfp -mfull-toc
719           -mminimal-toc  -mno-fp-in-toc  -mno-sum-in-toc -m64  -m32
720           -mxl-compat  -mno-xl-compat  -mpe -malign-power  -malign-natural
721           -msoft-float  -mhard-float  -mmultiple  -mno-multiple -mupdate
722           -mno-update -mavoid-indexed-addresses  -mno-avoid-indexed-addresses
723           -mfused-madd  -mno-fused-madd  -mbit-align  -mno-bit-align
724           -mstrict-align  -mno-strict-align  -mrelocatable -mno-relocatable
725           -mrelocatable-lib  -mno-relocatable-lib -mtoc  -mno-toc  -mlittle
726           -mlittle-endian  -mbig  -mbig-endian -mdynamic-no-pic  -mswdiv
727           -msingle-pic-base -mprioritize-restricted-insns=priority
728           -msched-costly-dep=dependence_type -minsert-sched-nops=scheme
729           -mcall-aixdesc  -mcall-eabi  -mcall-freebsd -mcall-linux
730           -mcall-netbsd  -mcall-openbsd -mcall-sysv  -mcall-sysv-eabi
731           -mcall-sysv-noeabi -mtraceback=traceback_type -maix-struct-return
732           -msvr4-struct-return -mabi=abi-type  -msecure-plt  -mbss-plt
733           -mlongcall  -mno-longcall  -mpltseq  -mno-pltseq
734           -mblock-move-inline-limit=num -mblock-compare-inline-limit=num
735           -mblock-compare-inline-loop-limit=num
736           -mstring-compare-inline-limit=num -misel  -mno-isel -mvrsave
737           -mno-vrsave -mmulhw  -mno-mulhw -mdlmzb  -mno-dlmzb -mprototype
738           -mno-prototype -msim  -mmvme  -mads  -myellowknife  -memb  -msdata
739           -msdata=opt  -mreadonly-in-sdata  -mvxworks  -G num -mrecip
740           -mrecip=opt  -mno-recip  -mrecip-precision -mno-recip-precision
741           -mveclibabi=type  -mfriz  -mno-friz -mpointers-to-nested-functions
742           -mno-pointers-to-nested-functions -msave-toc-indirect
743           -mno-save-toc-indirect -mpower8-fusion  -mno-mpower8-fusion
744           -mpower8-vector  -mno-power8-vector -mcrypto  -mno-crypto  -mhtm
745           -mno-htm -mquad-memory  -mno-quad-memory -mquad-memory-atomic
746           -mno-quad-memory-atomic -mcompat-align-parm  -mno-compat-align-parm
747           -mfloat128  -mno-float128  -mfloat128-hardware
748           -mno-float128-hardware -mgnu-attribute  -mno-gnu-attribute
749           -mstack-protector-guard=guard -mstack-protector-guard-reg=reg
750           -mstack-protector-guard-offset=offset
751
752           RX Options -m64bit-doubles  -m32bit-doubles  -fpu  -nofpu -mcpu=
753           -mbig-endian-data  -mlittle-endian-data -msmall-data -msim
754           -mno-sim -mas100-syntax  -mno-as100-syntax -mrelax
755           -mmax-constant-size= -mint-register= -mpid -mallow-string-insns
756           -mno-allow-string-insns -mjsr -mno-warn-multiple-fast-interrupts
757           -msave-acc-in-interrupts
758
759           S/390 and zSeries Options -mtune=cpu-type  -march=cpu-type
760           -mhard-float  -msoft-float  -mhard-dfp  -mno-hard-dfp
761           -mlong-double-64  -mlong-double-128 -mbackchain  -mno-backchain
762           -mpacked-stack  -mno-packed-stack -msmall-exec  -mno-small-exec
763           -mmvcle  -mno-mvcle -m64  -m31  -mdebug  -mno-debug  -mesa  -mzarch
764           -mhtm  -mvx  -mzvector -mtpf-trace  -mno-tpf-trace  -mfused-madd
765           -mno-fused-madd -mwarn-framesize  -mwarn-dynamicstack  -mstack-size
766           -mstack-guard -mhotpatch=halfwords,halfwords
767
768           Score Options -meb  -mel -mnhwloop -muls -mmac -mscore5  -mscore5u
769           -mscore7  -mscore7d
770
771           SH Options -m1  -m2  -m2e -m2a-nofpu  -m2a-single-only  -m2a-single
772           -m2a -m3  -m3e -m4-nofpu  -m4-single-only  -m4-single  -m4
773           -m4a-nofpu  -m4a-single-only  -m4a-single  -m4a  -m4al -mb  -ml
774           -mdalign  -mrelax -mbigtable  -mfmovd  -mrenesas  -mno-renesas
775           -mnomacsave -mieee  -mno-ieee  -mbitops  -misize
776           -minline-ic_invalidate  -mpadstruct -mprefergot  -musermode
777           -multcost=number  -mdiv=strategy -mdivsi3_libfunc=name
778           -mfixed-range=register-range -maccumulate-outgoing-args
779           -matomic-model=atomic-model -mbranch-cost=num  -mzdcbranch
780           -mno-zdcbranch -mcbranch-force-delay-slot -mfused-madd
781           -mno-fused-madd  -mfsca  -mno-fsca  -mfsrra  -mno-fsrra
782           -mpretend-cmove  -mtas
783
784           Solaris 2 Options -mclear-hwcap  -mno-clear-hwcap  -mimpure-text
785           -mno-impure-text -pthreads
786
787           SPARC Options -mcpu=cpu-type -mtune=cpu-type -mcmodel=code-model
788           -mmemory-model=mem-model -m32  -m64  -mapp-regs  -mno-app-regs
789           -mfaster-structs  -mno-faster-structs  -mflat  -mno-flat -mfpu
790           -mno-fpu  -mhard-float  -msoft-float -mhard-quad-float
791           -msoft-quad-float -mstack-bias  -mno-stack-bias -mstd-struct-return
792           -mno-std-struct-return -munaligned-doubles  -mno-unaligned-doubles
793           -muser-mode  -mno-user-mode -mv8plus  -mno-v8plus  -mvis  -mno-vis
794           -mvis2  -mno-vis2  -mvis3  -mno-vis3 -mvis4  -mno-vis4  -mvis4b
795           -mno-vis4b -mcbcond  -mno-cbcond  -mfmaf  -mno-fmaf  -mfsmuld
796           -mno-fsmuld -mpopc  -mno-popc  -msubxc  -mno-subxc -mfix-at697f
797           -mfix-ut699  -mfix-ut700  -mfix-gr712rc -mlra  -mno-lra
798
799           SPU Options -mwarn-reloc  -merror-reloc -msafe-dma  -munsafe-dma
800           -mbranch-hints -msmall-mem  -mlarge-mem  -mstdmain
801           -mfixed-range=register-range -mea32  -mea64
802           -maddress-space-conversion  -mno-address-space-conversion
803           -mcache-size=cache-size -matomic-updates  -mno-atomic-updates
804
805           System V Options -Qy  -Qn  -YP,paths  -Ym,dir
806
807           TILE-Gx Options -mcpu=CPU  -m32  -m64  -mbig-endian
808           -mlittle-endian -mcmodel=code-model
809
810           TILEPro Options -mcpu=cpu  -m32
811
812           V850 Options -mlong-calls  -mno-long-calls  -mep  -mno-ep
813           -mprolog-function  -mno-prolog-function  -mspace -mtda=n  -msda=n
814           -mzda=n -mapp-regs  -mno-app-regs -mdisable-callt
815           -mno-disable-callt -mv850e2v3  -mv850e2  -mv850e1  -mv850es -mv850e
816           -mv850  -mv850e3v5 -mloop -mrelax -mlong-jumps -msoft-float
817           -mhard-float -mgcc-abi -mrh850-abi -mbig-switch
818
819           VAX Options -mg  -mgnu  -munix
820
821           Visium Options -mdebug  -msim  -mfpu  -mno-fpu  -mhard-float
822           -msoft-float -mcpu=cpu-type  -mtune=cpu-type  -msv-mode
823           -muser-mode
824
825           VMS Options -mvms-return-codes  -mdebug-main=prefix  -mmalloc64
826           -mpointer-size=size
827
828           VxWorks Options -mrtp  -non-static  -Bstatic  -Bdynamic -Xbind-lazy
829           -Xbind-now
830
831           x86 Options -mtune=cpu-type  -march=cpu-type -mtune-ctrl=feature-
832           list  -mdump-tune-features  -mno-default -mfpmath=unit
833           -masm=dialect  -mno-fancy-math-387 -mno-fp-ret-in-387  -m80387
834           -mhard-float  -msoft-float -mno-wide-multiply  -mrtd
835           -malign-double -mpreferred-stack-boundary=num
836           -mincoming-stack-boundary=num -mcld  -mcx16  -msahf  -mmovbe
837           -mcrc32 -mrecip  -mrecip=opt -mvzeroupper  -mprefer-avx128
838           -mprefer-vector-width=opt -mmmx  -msse  -msse2  -msse3  -mssse3
839           -msse4.1  -msse4.2  -msse4  -mavx -mavx2  -mavx512f  -mavx512pf
840           -mavx512er  -mavx512cd  -mavx512vl -mavx512bw  -mavx512dq
841           -mavx512ifma  -mavx512vbmi  -msha  -maes -mpclmul  -mfsgsbase
842           -mrdrnd  -mf16c  -mfma  -mpconfig  -mwbnoinvd -mptwrite
843           -mprefetchwt1  -mclflushopt  -mclwb  -mxsavec  -mxsaves -msse4a
844           -m3dnow  -m3dnowa  -mpopcnt  -mabm  -mbmi  -mtbm  -mfma4  -mxop
845           -madx  -mlzcnt  -mbmi2  -mfxsr  -mxsave  -mxsaveopt  -mrtm  -mhle
846           -mlwp -mmwaitx  -mclzero  -mpku  -mthreads  -mgfni  -mvaes
847           -mwaitpkg -mshstk -mmanual-endbr -mforce-indirect-call
848           -mavx512vbmi2 -mvpclmulqdq  -mavx512bitalg  -mmovdiri  -mmovdir64b
849           -mavx512vpopcntdq -mavx5124fmaps  -mavx512vnni  -mavx5124vnniw
850           -mprfchw  -mrdpid -mrdseed  -msgx -mcldemote  -mms-bitfields
851           -mno-align-stringops  -minline-all-stringops
852           -minline-stringops-dynamically  -mstringop-strategy=alg
853           -mmemcpy-strategy=strategy  -mmemset-strategy=strategy -mpush-args
854           -maccumulate-outgoing-args  -m128bit-long-double
855           -m96bit-long-double  -mlong-double-64  -mlong-double-80
856           -mlong-double-128 -mregparm=num  -msseregparm -mveclibabi=type
857           -mvect8-ret-in-mem -mpc32  -mpc64  -mpc80  -mstackrealign
858           -momit-leaf-frame-pointer  -mno-red-zone  -mno-tls-direct-seg-refs
859           -mcmodel=code-model  -mabi=name  -maddress-mode=mode -m32  -m64
860           -mx32  -m16  -miamcu  -mlarge-data-threshold=num -msse2avx
861           -mfentry  -mrecord-mcount  -mnop-mcount  -m8bit-idiv
862           -minstrument-return=type -mfentry-name=name -mfentry-section=name
863           -mavx256-split-unaligned-load  -mavx256-split-unaligned-store
864           -malign-data=type  -mstack-protector-guard=guard
865           -mstack-protector-guard-reg=reg
866           -mstack-protector-guard-offset=offset
867           -mstack-protector-guard-symbol=symbol -mgeneral-regs-only
868           -mcall-ms2sysv-xlogues -mindirect-branch=choice
869           -mfunction-return=choice -mindirect-branch-register
870
871           x86 Windows Options -mconsole  -mcygwin  -mno-cygwin  -mdll
872           -mnop-fun-dllimport  -mthread -municode  -mwin32  -mwindows
873           -fno-set-stack-executable
874
875           Xstormy16 Options -msim
876
877           Xtensa Options -mconst16  -mno-const16 -mfused-madd
878           -mno-fused-madd -mforce-no-pic -mserialize-volatile
879           -mno-serialize-volatile -mtext-section-literals
880           -mno-text-section-literals -mauto-litpools  -mno-auto-litpools
881           -mtarget-align  -mno-target-align -mlongcalls  -mno-longcalls
882
883           zSeries Options See S/390 and zSeries Options.
884
885   Options Controlling the Kind of Output
886       Compilation can involve up to four stages: preprocessing, compilation
887       proper, assembly and linking, always in that order.  GCC is capable of
888       preprocessing and compiling several files either into several assembler
889       input files, or into one assembler input file; then each assembler
890       input file produces an object file, and linking combines all the object
891       files (those newly compiled, and those specified as input) into an
892       executable file.
893
894       For any given input file, the file name suffix determines what kind of
895       compilation is done:
896
897       file.c
898           C source code that must be preprocessed.
899
900       file.i
901           C source code that should not be preprocessed.
902
903       file.ii
904           C++ source code that should not be preprocessed.
905
906       file.m
907           Objective-C source code.  Note that you must link with the libobjc
908           library to make an Objective-C program work.
909
910       file.mi
911           Objective-C source code that should not be preprocessed.
912
913       file.mm
914       file.M
915           Objective-C++ source code.  Note that you must link with the
916           libobjc library to make an Objective-C++ program work.  Note that
917           .M refers to a literal capital M.
918
919       file.mii
920           Objective-C++ source code that should not be preprocessed.
921
922       file.h
923           C, C++, Objective-C or Objective-C++ header file to be turned into
924           a precompiled header (default), or C, C++ header file to be turned
925           into an Ada spec (via the -fdump-ada-spec switch).
926
927       file.cc
928       file.cp
929       file.cxx
930       file.cpp
931       file.CPP
932       file.c++
933       file.C
934           C++ source code that must be preprocessed.  Note that in .cxx, the
935           last two letters must both be literally x.  Likewise, .C refers to
936           a literal capital C.
937
938       file.mm
939       file.M
940           Objective-C++ source code that must be preprocessed.
941
942       file.mii
943           Objective-C++ source code that should not be preprocessed.
944
945       file.hh
946       file.H
947       file.hp
948       file.hxx
949       file.hpp
950       file.HPP
951       file.h++
952       file.tcc
953           C++ header file to be turned into a precompiled header or Ada spec.
954
955       file.f
956       file.for
957       file.ftn
958           Fixed form Fortran source code that should not be preprocessed.
959
960       file.F
961       file.FOR
962       file.fpp
963       file.FPP
964       file.FTN
965           Fixed form Fortran source code that must be preprocessed (with the
966           traditional preprocessor).
967
968       file.f90
969       file.f95
970       file.f03
971       file.f08
972           Free form Fortran source code that should not be preprocessed.
973
974       file.F90
975       file.F95
976       file.F03
977       file.F08
978           Free form Fortran source code that must be preprocessed (with the
979           traditional preprocessor).
980
981       file.go
982           Go source code.
983
984       file.brig
985           BRIG files (binary representation of HSAIL).
986
987       file.d
988           D source code.
989
990       file.di
991           D interface file.
992
993       file.dd
994           D documentation code (Ddoc).
995
996       file.ads
997           Ada source code file that contains a library unit declaration (a
998           declaration of a package, subprogram, or generic, or a generic
999           instantiation), or a library unit renaming declaration (a package,
1000           generic, or subprogram renaming declaration).  Such files are also
1001           called specs.
1002
1003       file.adb
1004           Ada source code file containing a library unit body (a subprogram
1005           or package body).  Such files are also called bodies.
1006
1007       file.s
1008           Assembler code.
1009
1010       file.S
1011       file.sx
1012           Assembler code that must be preprocessed.
1013
1014       other
1015           An object file to be fed straight into linking.  Any file name with
1016           no recognized suffix is treated this way.
1017
1018       You can specify the input language explicitly with the -x option:
1019
1020       -x language
1021           Specify explicitly the language for the following input files
1022           (rather than letting the compiler choose a default based on the
1023           file name suffix).  This option applies to all following input
1024           files until the next -x option.  Possible values for language are:
1025
1026                   c  c-header  cpp-output
1027                   c++  c++-header  c++-cpp-output
1028                   objective-c  objective-c-header  objective-c-cpp-output
1029                   objective-c++ objective-c++-header objective-c++-cpp-output
1030                   assembler  assembler-with-cpp
1031                   ada
1032                   d
1033                   f77  f77-cpp-input f95  f95-cpp-input
1034                   go
1035                   brig
1036
1037       -x none
1038           Turn off any specification of a language, so that subsequent files
1039           are handled according to their file name suffixes (as they are if
1040           -x has not been used at all).
1041
1042       If you only want some of the stages of compilation, you can use -x (or
1043       filename suffixes) to tell gcc where to start, and one of the options
1044       -c, -S, or -E to say where gcc is to stop.  Note that some combinations
1045       (for example, -x cpp-output -E) instruct gcc to do nothing at all.
1046
1047       -c  Compile or assemble the source files, but do not link.  The linking
1048           stage simply is not done.  The ultimate output is in the form of an
1049           object file for each source file.
1050
1051           By default, the object file name for a source file is made by
1052           replacing the suffix .c, .i, .s, etc., with .o.
1053
1054           Unrecognized input files, not requiring compilation or assembly,
1055           are ignored.
1056
1057       -S  Stop after the stage of compilation proper; do not assemble.  The
1058           output is in the form of an assembler code file for each non-
1059           assembler input file specified.
1060
1061           By default, the assembler file name for a source file is made by
1062           replacing the suffix .c, .i, etc., with .s.
1063
1064           Input files that don't require compilation are ignored.
1065
1066       -E  Stop after the preprocessing stage; do not run the compiler proper.
1067           The output is in the form of preprocessed source code, which is
1068           sent to the standard output.
1069
1070           Input files that don't require preprocessing are ignored.
1071
1072       -o file
1073           Place output in file file.  This applies to whatever sort of output
1074           is being produced, whether it be an executable file, an object
1075           file, an assembler file or preprocessed C code.
1076
1077           If -o is not specified, the default is to put an executable file in
1078           a.out, the object file for source.suffix in source.o, its assembler
1079           file in source.s, a precompiled header file in source.suffix.gch,
1080           and all preprocessed C source on standard output.
1081
1082       -v  Print (on standard error output) the commands executed to run the
1083           stages of compilation.  Also print the version number of the
1084           compiler driver program and of the preprocessor and the compiler
1085           proper.
1086
1087       -###
1088           Like -v except the commands are not executed and arguments are
1089           quoted unless they contain only alphanumeric characters or "./-_".
1090           This is useful for shell scripts to capture the driver-generated
1091           command lines.
1092
1093       --help
1094           Print (on the standard output) a description of the command-line
1095           options understood by gcc.  If the -v option is also specified then
1096           --help is also passed on to the various processes invoked by gcc,
1097           so that they can display the command-line options they accept.  If
1098           the -Wextra option has also been specified (prior to the --help
1099           option), then command-line options that have no documentation
1100           associated with them are also displayed.
1101
1102       --target-help
1103           Print (on the standard output) a description of target-specific
1104           command-line options for each tool.  For some targets extra target-
1105           specific information may also be printed.
1106
1107       --help={class|[^]qualifier}[,...]
1108           Print (on the standard output) a description of the command-line
1109           options understood by the compiler that fit into all specified
1110           classes and qualifiers.  These are the supported classes:
1111
1112           optimizers
1113               Display all of the optimization options supported by the
1114               compiler.
1115
1116           warnings
1117               Display all of the options controlling warning messages
1118               produced by the compiler.
1119
1120           target
1121               Display target-specific options.  Unlike the --target-help
1122               option however, target-specific options of the linker and
1123               assembler are not displayed.  This is because those tools do
1124               not currently support the extended --help= syntax.
1125
1126           params
1127               Display the values recognized by the --param option.
1128
1129           language
1130               Display the options supported for language, where language is
1131               the name of one of the languages supported in this version of
1132               GCC.
1133
1134           common
1135               Display the options that are common to all languages.
1136
1137           These are the supported qualifiers:
1138
1139           undocumented
1140               Display only those options that are undocumented.
1141
1142           joined
1143               Display options taking an argument that appears after an equal
1144               sign in the same continuous piece of text, such as:
1145               --help=target.
1146
1147           separate
1148               Display options taking an argument that appears as a separate
1149               word following the original option, such as: -o output-file.
1150
1151           Thus for example to display all the undocumented target-specific
1152           switches supported by the compiler, use:
1153
1154                   --help=target,undocumented
1155
1156           The sense of a qualifier can be inverted by prefixing it with the ^
1157           character, so for example to display all binary warning options
1158           (i.e., ones that are either on or off and that do not take an
1159           argument) that have a description, use:
1160
1161                   --help=warnings,^joined,^undocumented
1162
1163           The argument to --help= should not consist solely of inverted
1164           qualifiers.
1165
1166           Combining several classes is possible, although this usually
1167           restricts the output so much that there is nothing to display.  One
1168           case where it does work, however, is when one of the classes is
1169           target.  For example, to display all the target-specific
1170           optimization options, use:
1171
1172                   --help=target,optimizers
1173
1174           The --help= option can be repeated on the command line.  Each
1175           successive use displays its requested class of options, skipping
1176           those that have already been displayed.  If --help is also
1177           specified anywhere on the command line then this takes precedence
1178           over any --help= option.
1179
1180           If the -Q option appears on the command line before the --help=
1181           option, then the descriptive text displayed by --help= is changed.
1182           Instead of describing the displayed options, an indication is given
1183           as to whether the option is enabled, disabled or set to a specific
1184           value (assuming that the compiler knows this at the point where the
1185           --help= option is used).
1186
1187           Here is a truncated example from the ARM port of gcc:
1188
1189                     % gcc -Q -mabi=2 --help=target -c
1190                     The following options are target specific:
1191                     -mabi=                                2
1192                     -mabort-on-noreturn                   [disabled]
1193                     -mapcs                                [disabled]
1194
1195           The output is sensitive to the effects of previous command-line
1196           options, so for example it is possible to find out which
1197           optimizations are enabled at -O2 by using:
1198
1199                   -Q -O2 --help=optimizers
1200
1201           Alternatively you can discover which binary optimizations are
1202           enabled by -O3 by using:
1203
1204                   gcc -c -Q -O3 --help=optimizers > /tmp/O3-opts
1205                   gcc -c -Q -O2 --help=optimizers > /tmp/O2-opts
1206                   diff /tmp/O2-opts /tmp/O3-opts | grep enabled
1207
1208       --version
1209           Display the version number and copyrights of the invoked GCC.
1210
1211       -pass-exit-codes
1212           Normally the gcc program exits with the code of 1 if any phase of
1213           the compiler returns a non-success return code.  If you specify
1214           -pass-exit-codes, the gcc program instead returns with the
1215           numerically highest error produced by any phase returning an error
1216           indication.  The C, C++, and Fortran front ends return 4 if an
1217           internal compiler error is encountered.
1218
1219       -pipe
1220           Use pipes rather than temporary files for communication between the
1221           various stages of compilation.  This fails to work on some systems
1222           where the assembler is unable to read from a pipe; but the GNU
1223           assembler has no trouble.
1224
1225       -specs=file
1226           Process file after the compiler reads in the standard specs file,
1227           in order to override the defaults which the gcc driver program uses
1228           when determining what switches to pass to cc1, cc1plus, as, ld,
1229           etc.  More than one -specs=file can be specified on the command
1230           line, and they are processed in order, from left to right.
1231
1232       -wrapper
1233           Invoke all subcommands under a wrapper program.  The name of the
1234           wrapper program and its parameters are passed as a comma separated
1235           list.
1236
1237                   gcc -c t.c -wrapper gdb,--args
1238
1239           This invokes all subprograms of gcc under gdb --args, thus the
1240           invocation of cc1 is gdb --args cc1 ....
1241
1242       -ffile-prefix-map=old=new
1243           When compiling files residing in directory old, record any
1244           references to them in the result of the compilation as if the files
1245           resided in directory new instead.  Specifying this option is
1246           equivalent to specifying all the individual -f*-prefix-map options.
1247           This can be used to make reproducible builds that are location
1248           independent.  See also -fmacro-prefix-map and -fdebug-prefix-map.
1249
1250       -fplugin=name.so
1251           Load the plugin code in file name.so, assumed to be a shared object
1252           to be dlopen'd by the compiler.  The base name of the shared object
1253           file is used to identify the plugin for the purposes of argument
1254           parsing (See -fplugin-arg-name-key=value below).  Each plugin
1255           should define the callback functions specified in the Plugins API.
1256
1257       -fplugin-arg-name-key=value
1258           Define an argument called key with a value of value for the plugin
1259           called name.
1260
1261       -fdump-ada-spec[-slim]
1262           For C and C++ source and include files, generate corresponding Ada
1263           specs.
1264
1265       -fada-spec-parent=unit
1266           In conjunction with -fdump-ada-spec[-slim] above, generate Ada
1267           specs as child units of parent unit.
1268
1269       -fdump-go-spec=file
1270           For input files in any language, generate corresponding Go
1271           declarations in file.  This generates Go "const", "type", "var",
1272           and "func" declarations which may be a useful way to start writing
1273           a Go interface to code written in some other language.
1274
1275       @file
1276           Read command-line options from file.  The options read are inserted
1277           in place of the original @file option.  If file does not exist, or
1278           cannot be read, then the option will be treated literally, and not
1279           removed.
1280
1281           Options in file are separated by whitespace.  A whitespace
1282           character may be included in an option by surrounding the entire
1283           option in either single or double quotes.  Any character (including
1284           a backslash) may be included by prefixing the character to be
1285           included with a backslash.  The file may itself contain additional
1286           @file options; any such options will be processed recursively.
1287
1288   Compiling C++ Programs
1289       C++ source files conventionally use one of the suffixes .C, .cc, .cpp,
1290       .CPP, .c++, .cp, or .cxx; C++ header files often use .hh, .hpp, .H, or
1291       (for shared template code) .tcc; and preprocessed C++ files use the
1292       suffix .ii.  GCC recognizes files with these names and compiles them as
1293       C++ programs even if you call the compiler the same way as for
1294       compiling C programs (usually with the name gcc).
1295
1296       However, the use of gcc does not add the C++ library.  g++ is a program
1297       that calls GCC and automatically specifies linking against the C++
1298       library.  It treats .c, .h and .i files as C++ source files instead of
1299       C source files unless -x is used.  This program is also useful when
1300       precompiling a C header file with a .h extension for use in C++
1301       compilations.  On many systems, g++ is also installed with the name
1302       c++.
1303
1304       When you compile C++ programs, you may specify many of the same
1305       command-line options that you use for compiling programs in any
1306       language; or command-line options meaningful for C and related
1307       languages; or options that are meaningful only for C++ programs.
1308
1309   Options Controlling C Dialect
1310       The following options control the dialect of C (or languages derived
1311       from C, such as C++, Objective-C and Objective-C++) that the compiler
1312       accepts:
1313
1314       -ansi
1315           In C mode, this is equivalent to -std=c90. In C++ mode, it is
1316           equivalent to -std=c++98.
1317
1318           This turns off certain features of GCC that are incompatible with
1319           ISO C90 (when compiling C code), or of standard C++ (when compiling
1320           C++ code), such as the "asm" and "typeof" keywords, and predefined
1321           macros such as "unix" and "vax" that identify the type of system
1322           you are using.  It also enables the undesirable and rarely used ISO
1323           trigraph feature.  For the C compiler, it disables recognition of
1324           C++ style // comments as well as the "inline" keyword.
1325
1326           The alternate keywords "__asm__", "__extension__", "__inline__" and
1327           "__typeof__" continue to work despite -ansi.  You would not want to
1328           use them in an ISO C program, of course, but it is useful to put
1329           them in header files that might be included in compilations done
1330           with -ansi.  Alternate predefined macros such as "__unix__" and
1331           "__vax__" are also available, with or without -ansi.
1332
1333           The -ansi option does not cause non-ISO programs to be rejected
1334           gratuitously.  For that, -Wpedantic is required in addition to
1335           -ansi.
1336
1337           The macro "__STRICT_ANSI__" is predefined when the -ansi option is
1338           used.  Some header files may notice this macro and refrain from
1339           declaring certain functions or defining certain macros that the ISO
1340           standard doesn't call for; this is to avoid interfering with any
1341           programs that might use these names for other things.
1342
1343           Functions that are normally built in but do not have semantics
1344           defined by ISO C (such as "alloca" and "ffs") are not built-in
1345           functions when -ansi is used.
1346
1347       -std=
1348           Determine the language standard.   This option is currently only
1349           supported when compiling C or C++.
1350
1351           The compiler can accept several base standards, such as c90 or
1352           c++98, and GNU dialects of those standards, such as gnu90 or
1353           gnu++98.  When a base standard is specified, the compiler accepts
1354           all programs following that standard plus those using GNU
1355           extensions that do not contradict it.  For example, -std=c90 turns
1356           off certain features of GCC that are incompatible with ISO C90,
1357           such as the "asm" and "typeof" keywords, but not other GNU
1358           extensions that do not have a meaning in ISO C90, such as omitting
1359           the middle term of a "?:" expression. On the other hand, when a GNU
1360           dialect of a standard is specified, all features supported by the
1361           compiler are enabled, even when those features change the meaning
1362           of the base standard.  As a result, some strict-conforming programs
1363           may be rejected.  The particular standard is used by -Wpedantic to
1364           identify which features are GNU extensions given that version of
1365           the standard. For example -std=gnu90 -Wpedantic warns about C++
1366           style // comments, while -std=gnu99 -Wpedantic does not.
1367
1368           A value for this option must be provided; possible values are
1369
1370           c90
1371           c89
1372           iso9899:1990
1373               Support all ISO C90 programs (certain GNU extensions that
1374               conflict with ISO C90 are disabled). Same as -ansi for C code.
1375
1376           iso9899:199409
1377               ISO C90 as modified in amendment 1.
1378
1379           c99
1380           c9x
1381           iso9899:1999
1382           iso9899:199x
1383               ISO C99.  This standard is substantially completely supported,
1384               modulo bugs and floating-point issues (mainly but not entirely
1385               relating to optional C99 features from Annexes F and G).  See
1386               <http://gcc.gnu.org/c99status.html> for more information.  The
1387               names c9x and iso9899:199x are deprecated.
1388
1389           c11
1390           c1x
1391           iso9899:2011
1392               ISO C11, the 2011 revision of the ISO C standard.  This
1393               standard is substantially completely supported, modulo bugs,
1394               floating-point issues (mainly but not entirely relating to
1395               optional C11 features from Annexes F and G) and the optional
1396               Annexes K (Bounds-checking interfaces) and L (Analyzability).
1397               The name c1x is deprecated.
1398
1399           c17
1400           c18
1401           iso9899:2017
1402           iso9899:2018
1403               ISO C17, the 2017 revision of the ISO C standard (published in
1404               2018).  This standard is same as C11 except for corrections of
1405               defects (all of which are also applied with -std=c11) and a new
1406               value of "__STDC_VERSION__", and so is supported to the same
1407               extent as C11.
1408
1409           c2x The next version of the ISO C standard, still under
1410               development.  The support for this version is experimental and
1411               incomplete.
1412
1413           gnu90
1414           gnu89
1415               GNU dialect of ISO C90 (including some C99 features).
1416
1417           gnu99
1418           gnu9x
1419               GNU dialect of ISO C99.  The name gnu9x is deprecated.
1420
1421           gnu11
1422           gnu1x
1423               GNU dialect of ISO C11.  The name gnu1x is deprecated.
1424
1425           gnu17
1426           gnu18
1427               GNU dialect of ISO C17.  This is the default for C code.
1428
1429           gnu2x
1430               The next version of the ISO C standard, still under
1431               development, plus GNU extensions.  The support for this version
1432               is experimental and incomplete.
1433
1434           c++98
1435           c++03
1436               The 1998 ISO C++ standard plus the 2003 technical corrigendum
1437               and some additional defect reports. Same as -ansi for C++ code.
1438
1439           gnu++98
1440           gnu++03
1441               GNU dialect of -std=c++98.
1442
1443           c++11
1444           c++0x
1445               The 2011 ISO C++ standard plus amendments.  The name c++0x is
1446               deprecated.
1447
1448           gnu++11
1449           gnu++0x
1450               GNU dialect of -std=c++11.  The name gnu++0x is deprecated.
1451
1452           c++14
1453           c++1y
1454               The 2014 ISO C++ standard plus amendments.  The name c++1y is
1455               deprecated.
1456
1457           gnu++14
1458           gnu++1y
1459               GNU dialect of -std=c++14.  This is the default for C++ code.
1460               The name gnu++1y is deprecated.
1461
1462           c++17
1463           c++1z
1464               The 2017 ISO C++ standard plus amendments.  The name c++1z is
1465               deprecated.
1466
1467           gnu++17
1468           gnu++1z
1469               GNU dialect of -std=c++17.  The name gnu++1z is deprecated.
1470
1471           c++2a
1472               The next revision of the ISO C++ standard, tentatively planned
1473               for 2020.  Support is highly experimental, and will almost
1474               certainly change in incompatible ways in future releases.
1475
1476           gnu++2a
1477               GNU dialect of -std=c++2a.  Support is highly experimental, and
1478               will almost certainly change in incompatible ways in future
1479               releases.
1480
1481       -fgnu89-inline
1482           The option -fgnu89-inline tells GCC to use the traditional GNU
1483           semantics for "inline" functions when in C99 mode.
1484
1485           Using this option is roughly equivalent to adding the "gnu_inline"
1486           function attribute to all inline functions.
1487
1488           The option -fno-gnu89-inline explicitly tells GCC to use the C99
1489           semantics for "inline" when in C99 or gnu99 mode (i.e., it
1490           specifies the default behavior).  This option is not supported in
1491           -std=c90 or -std=gnu90 mode.
1492
1493           The preprocessor macros "__GNUC_GNU_INLINE__" and
1494           "__GNUC_STDC_INLINE__" may be used to check which semantics are in
1495           effect for "inline" functions.
1496
1497       -fpermitted-flt-eval-methods=style
1498           ISO/IEC TS 18661-3 defines new permissible values for
1499           "FLT_EVAL_METHOD" that indicate that operations and constants with
1500           a semantic type that is an interchange or extended format should be
1501           evaluated to the precision and range of that type.  These new
1502           values are a superset of those permitted under C99/C11, which does
1503           not specify the meaning of other positive values of
1504           "FLT_EVAL_METHOD".  As such, code conforming to C11 may not have
1505           been written expecting the possibility of the new values.
1506
1507           -fpermitted-flt-eval-methods specifies whether the compiler should
1508           allow only the values of "FLT_EVAL_METHOD" specified in C99/C11, or
1509           the extended set of values specified in ISO/IEC TS 18661-3.
1510
1511           style is either "c11" or "ts-18661-3" as appropriate.
1512
1513           The default when in a standards compliant mode (-std=c11 or
1514           similar) is -fpermitted-flt-eval-methods=c11.  The default when in
1515           a GNU dialect (-std=gnu11 or similar) is
1516           -fpermitted-flt-eval-methods=ts-18661-3.
1517
1518       -aux-info filename
1519           Output to the given filename prototyped declarations for all
1520           functions declared and/or defined in a translation unit, including
1521           those in header files.  This option is silently ignored in any
1522           language other than C.
1523
1524           Besides declarations, the file indicates, in comments, the origin
1525           of each declaration (source file and line), whether the declaration
1526           was implicit, prototyped or unprototyped (I, N for new or O for
1527           old, respectively, in the first character after the line number and
1528           the colon), and whether it came from a declaration or a definition
1529           (C or F, respectively, in the following character).  In the case of
1530           function definitions, a K&R-style list of arguments followed by
1531           their declarations is also provided, inside comments, after the
1532           declaration.
1533
1534       -fallow-parameterless-variadic-functions
1535           Accept variadic functions without named parameters.
1536
1537           Although it is possible to define such a function, this is not very
1538           useful as it is not possible to read the arguments.  This is only
1539           supported for C as this construct is allowed by C++.
1540
1541       -fno-asm
1542           Do not recognize "asm", "inline" or "typeof" as a keyword, so that
1543           code can use these words as identifiers.  You can use the keywords
1544           "__asm__", "__inline__" and "__typeof__" instead.  -ansi implies
1545           -fno-asm.
1546
1547           In C++, this switch only affects the "typeof" keyword, since "asm"
1548           and "inline" are standard keywords.  You may want to use the
1549           -fno-gnu-keywords flag instead, which has the same effect.  In C99
1550           mode (-std=c99 or -std=gnu99), this switch only affects the "asm"
1551           and "typeof" keywords, since "inline" is a standard keyword in ISO
1552           C99.
1553
1554       -fno-builtin
1555       -fno-builtin-function
1556           Don't recognize built-in functions that do not begin with
1557           __builtin_ as prefix.
1558
1559           GCC normally generates special code to handle certain built-in
1560           functions more efficiently; for instance, calls to "alloca" may
1561           become single instructions which adjust the stack directly, and
1562           calls to "memcpy" may become inline copy loops.  The resulting code
1563           is often both smaller and faster, but since the function calls no
1564           longer appear as such, you cannot set a breakpoint on those calls,
1565           nor can you change the behavior of the functions by linking with a
1566           different library.  In addition, when a function is recognized as a
1567           built-in function, GCC may use information about that function to
1568           warn about problems with calls to that function, or to generate
1569           more efficient code, even if the resulting code still contains
1570           calls to that function.  For example, warnings are given with
1571           -Wformat for bad calls to "printf" when "printf" is built in and
1572           "strlen" is known not to modify global memory.
1573
1574           With the -fno-builtin-function option only the built-in function
1575           function is disabled.  function must not begin with __builtin_.  If
1576           a function is named that is not built-in in this version of GCC,
1577           this option is ignored.  There is no corresponding
1578           -fbuiltin-function option; if you wish to enable built-in functions
1579           selectively when using -fno-builtin or -ffreestanding, you may
1580           define macros such as:
1581
1582                   #define abs(n)          __builtin_abs ((n))
1583                   #define strcpy(d, s)    __builtin_strcpy ((d), (s))
1584
1585       -fgimple
1586           Enable parsing of function definitions marked with "__GIMPLE".
1587           This is an experimental feature that allows unit testing of GIMPLE
1588           passes.
1589
1590       -fhosted
1591           Assert that compilation targets a hosted environment.  This implies
1592           -fbuiltin.  A hosted environment is one in which the entire
1593           standard library is available, and in which "main" has a return
1594           type of "int".  Examples are nearly everything except a kernel.
1595           This is equivalent to -fno-freestanding.
1596
1597       -ffreestanding
1598           Assert that compilation targets a freestanding environment.  This
1599           implies -fno-builtin.  A freestanding environment is one in which
1600           the standard library may not exist, and program startup may not
1601           necessarily be at "main".  The most obvious example is an OS
1602           kernel.  This is equivalent to -fno-hosted.
1603
1604       -fopenacc
1605           Enable handling of OpenACC directives "#pragma acc" in C/C++ and
1606           "!$acc" in Fortran.  When -fopenacc is specified, the compiler
1607           generates accelerated code according to the OpenACC Application
1608           Programming Interface v2.0 <https://www.openacc.org>.  This option
1609           implies -pthread, and thus is only supported on targets that have
1610           support for -pthread.
1611
1612       -fopenacc-dim=geom
1613           Specify default compute dimensions for parallel offload regions
1614           that do not explicitly specify.  The geom value is a triple of
1615           ':'-separated sizes, in order 'gang', 'worker' and, 'vector'.  A
1616           size can be omitted, to use a target-specific default value.
1617
1618       -fopenmp
1619           Enable handling of OpenMP directives "#pragma omp" in C/C++ and
1620           "!$omp" in Fortran.  When -fopenmp is specified, the compiler
1621           generates parallel code according to the OpenMP Application Program
1622           Interface v4.5 <https://www.openmp.org>.  This option implies
1623           -pthread, and thus is only supported on targets that have support
1624           for -pthread. -fopenmp implies -fopenmp-simd.
1625
1626       -fopenmp-simd
1627           Enable handling of OpenMP's SIMD directives with "#pragma omp" in
1628           C/C++ and "!$omp" in Fortran. Other OpenMP directives are ignored.
1629
1630       -fgnu-tm
1631           When the option -fgnu-tm is specified, the compiler generates code
1632           for the Linux variant of Intel's current Transactional Memory ABI
1633           specification document (Revision 1.1, May 6 2009).  This is an
1634           experimental feature whose interface may change in future versions
1635           of GCC, as the official specification changes.  Please note that
1636           not all architectures are supported for this feature.
1637
1638           For more information on GCC's support for transactional memory,
1639
1640           Note that the transactional memory feature is not supported with
1641           non-call exceptions (-fnon-call-exceptions).
1642
1643       -fms-extensions
1644           Accept some non-standard constructs used in Microsoft header files.
1645
1646           In C++ code, this allows member names in structures to be similar
1647           to previous types declarations.
1648
1649                   typedef int UOW;
1650                   struct ABC {
1651                     UOW UOW;
1652                   };
1653
1654           Some cases of unnamed fields in structures and unions are only
1655           accepted with this option.
1656
1657           Note that this option is off for all targets but x86 targets using
1658           ms-abi.
1659
1660       -fplan9-extensions
1661           Accept some non-standard constructs used in Plan 9 code.
1662
1663           This enables -fms-extensions, permits passing pointers to
1664           structures with anonymous fields to functions that expect pointers
1665           to elements of the type of the field, and permits referring to
1666           anonymous fields declared using a typedef.    This is only
1667           supported for C, not C++.
1668
1669       -fcond-mismatch
1670           Allow conditional expressions with mismatched types in the second
1671           and third arguments.  The value of such an expression is void.
1672           This option is not supported for C++.
1673
1674       -flax-vector-conversions
1675           Allow implicit conversions between vectors with differing numbers
1676           of elements and/or incompatible element types.  This option should
1677           not be used for new code.
1678
1679       -funsigned-char
1680           Let the type "char" be unsigned, like "unsigned char".
1681
1682           Each kind of machine has a default for what "char" should be.  It
1683           is either like "unsigned char" by default or like "signed char" by
1684           default.
1685
1686           Ideally, a portable program should always use "signed char" or
1687           "unsigned char" when it depends on the signedness of an object.
1688           But many programs have been written to use plain "char" and expect
1689           it to be signed, or expect it to be unsigned, depending on the
1690           machines they were written for.  This option, and its inverse, let
1691           you make such a program work with the opposite default.
1692
1693           The type "char" is always a distinct type from each of "signed
1694           char" or "unsigned char", even though its behavior is always just
1695           like one of those two.
1696
1697       -fsigned-char
1698           Let the type "char" be signed, like "signed char".
1699
1700           Note that this is equivalent to -fno-unsigned-char, which is the
1701           negative form of -funsigned-char.  Likewise, the option
1702           -fno-signed-char is equivalent to -funsigned-char.
1703
1704       -fsigned-bitfields
1705       -funsigned-bitfields
1706       -fno-signed-bitfields
1707       -fno-unsigned-bitfields
1708           These options control whether a bit-field is signed or unsigned,
1709           when the declaration does not use either "signed" or "unsigned".
1710           By default, such a bit-field is signed, because this is consistent:
1711           the basic integer types such as "int" are signed types.
1712
1713       -fsso-struct=endianness
1714           Set the default scalar storage order of structures and unions to
1715           the specified endianness.  The accepted values are big-endian,
1716           little-endian and native for the native endianness of the target
1717           (the default).  This option is not supported for C++.
1718
1719           Warning: the -fsso-struct switch causes GCC to generate code that
1720           is not binary compatible with code generated without it if the
1721           specified endianness is not the native endianness of the target.
1722
1723   Options Controlling C++ Dialect
1724       This section describes the command-line options that are only
1725       meaningful for C++ programs.  You can also use most of the GNU compiler
1726       options regardless of what language your program is in.  For example,
1727       you might compile a file firstClass.C like this:
1728
1729               g++ -g -fstrict-enums -O -c firstClass.C
1730
1731       In this example, only -fstrict-enums is an option meant only for C++
1732       programs; you can use the other options with any language supported by
1733       GCC.
1734
1735       Some options for compiling C programs, such as -std, are also relevant
1736       for C++ programs.
1737
1738       Here is a list of options that are only for compiling C++ programs:
1739
1740       -fabi-version=n
1741           Use version n of the C++ ABI.  The default is version 0.
1742
1743           Version 0 refers to the version conforming most closely to the C++
1744           ABI specification.  Therefore, the ABI obtained using version 0
1745           will change in different versions of G++ as ABI bugs are fixed.
1746
1747           Version 1 is the version of the C++ ABI that first appeared in G++
1748           3.2.
1749
1750           Version 2 is the version of the C++ ABI that first appeared in G++
1751           3.4, and was the default through G++ 4.9.
1752
1753           Version 3 corrects an error in mangling a constant address as a
1754           template argument.
1755
1756           Version 4, which first appeared in G++ 4.5, implements a standard
1757           mangling for vector types.
1758
1759           Version 5, which first appeared in G++ 4.6, corrects the mangling
1760           of attribute const/volatile on function pointer types, decltype of
1761           a plain decl, and use of a function parameter in the declaration of
1762           another parameter.
1763
1764           Version 6, which first appeared in G++ 4.7, corrects the promotion
1765           behavior of C++11 scoped enums and the mangling of template
1766           argument packs, const/static_cast, prefix ++ and --, and a class
1767           scope function used as a template argument.
1768
1769           Version 7, which first appeared in G++ 4.8, that treats nullptr_t
1770           as a builtin type and corrects the mangling of lambdas in default
1771           argument scope.
1772
1773           Version 8, which first appeared in G++ 4.9, corrects the
1774           substitution behavior of function types with function-cv-
1775           qualifiers.
1776
1777           Version 9, which first appeared in G++ 5.2, corrects the alignment
1778           of "nullptr_t".
1779
1780           Version 10, which first appeared in G++ 6.1, adds mangling of
1781           attributes that affect type identity, such as ia32 calling
1782           convention attributes (e.g. stdcall).
1783
1784           Version 11, which first appeared in G++ 7, corrects the mangling of
1785           sizeof... expressions and operator names.  For multiple entities
1786           with the same name within a function, that are declared in
1787           different scopes, the mangling now changes starting with the
1788           twelfth occurrence.  It also implies -fnew-inheriting-ctors.
1789
1790           Version 12, which first appeared in G++ 8, corrects the calling
1791           conventions for empty classes on the x86_64 target and for classes
1792           with only deleted copy/move constructors.  It accidentally changes
1793           the calling convention for classes with a deleted copy constructor
1794           and a trivial move constructor.
1795
1796           Version 13, which first appeared in G++ 8.2, fixes the accidental
1797           change in version 12.
1798
1799           See also -Wabi.
1800
1801       -fabi-compat-version=n
1802           On targets that support strong aliases, G++ works around mangling
1803           changes by creating an alias with the correct mangled name when
1804           defining a symbol with an incorrect mangled name.  This switch
1805           specifies which ABI version to use for the alias.
1806
1807           With -fabi-version=0 (the default), this defaults to 11 (GCC 7
1808           compatibility).  If another ABI version is explicitly selected,
1809           this defaults to 0.  For compatibility with GCC versions 3.2
1810           through 4.9, use -fabi-compat-version=2.
1811
1812           If this option is not provided but -Wabi=n is, that version is used
1813           for compatibility aliases.  If this option is provided along with
1814           -Wabi (without the version), the version from this option is used
1815           for the warning.
1816
1817       -fno-access-control
1818           Turn off all access checking.  This switch is mainly useful for
1819           working around bugs in the access control code.
1820
1821       -faligned-new
1822           Enable support for C++17 "new" of types that require more alignment
1823           than "void* ::operator new(std::size_t)" provides.  A numeric
1824           argument such as "-faligned-new=32" can be used to specify how much
1825           alignment (in bytes) is provided by that function, but few users
1826           will need to override the default of "alignof(std::max_align_t)".
1827
1828           This flag is enabled by default for -std=c++17.
1829
1830       -fchar8_t
1831       -fno-char8_t
1832           Enable support for "char8_t" as adopted for C++2a.  This includes
1833           the addition of a new "char8_t" fundamental type, changes to the
1834           types of UTF-8 string and character literals, new signatures for
1835           user-defined literals, associated standard library updates, and new
1836           "__cpp_char8_t" and "__cpp_lib_char8_t" feature test macros.
1837
1838           This option enables functions to be overloaded for ordinary and
1839           UTF-8 strings:
1840
1841                   int f(const char *);    // #1
1842                   int f(const char8_t *); // #2
1843                   int v1 = f("text");     // Calls #1
1844                   int v2 = f(u8"text");   // Calls #2
1845
1846           and introduces new signatures for user-defined literals:
1847
1848                   int operator""_udl1(char8_t);
1849                   int v3 = u8'x'_udl1;
1850                   int operator""_udl2(const char8_t*, std::size_t);
1851                   int v4 = u8"text"_udl2;
1852                   template<typename T, T...> int operator""_udl3();
1853                   int v5 = u8"text"_udl3;
1854
1855           The change to the types of UTF-8 string and character literals
1856           introduces incompatibilities with ISO C++11 and later standards.
1857           For example, the following code is well-formed under ISO C++11, but
1858           is ill-formed when -fchar8_t is specified.
1859
1860                   char ca[] = u8"xx";     // error: char-array initialized from wide
1861                                           //        string
1862                   const char *cp = u8"xx";// error: invalid conversion from
1863                                           //        `const char8_t*' to `const char*'
1864                   int f(const char*);
1865                   auto v = f(u8"xx");     // error: invalid conversion from
1866                                           //        `const char8_t*' to `const char*'
1867                   std::string s{u8"xx"};  // error: no matching function for call to
1868                                           //        `std::basic_string<char>::basic_string()'
1869                   using namespace std::literals;
1870                   s = u8"xx"s;            // error: conversion from
1871                                           //        `basic_string<char8_t>' to non-scalar
1872                                           //        type `basic_string<char>' requested
1873
1874       -fcheck-new
1875           Check that the pointer returned by "operator new" is non-null
1876           before attempting to modify the storage allocated.  This check is
1877           normally unnecessary because the C++ standard specifies that
1878           "operator new" only returns 0 if it is declared "throw()", in which
1879           case the compiler always checks the return value even without this
1880           option.  In all other cases, when "operator new" has a non-empty
1881           exception specification, memory exhaustion is signalled by throwing
1882           "std::bad_alloc".  See also new (nothrow).
1883
1884       -fconcepts
1885           Enable support for the C++ Extensions for Concepts Technical
1886           Specification, ISO 19217 (2015), which allows code like
1887
1888                   template <class T> concept bool Addable = requires (T t) { t + t; };
1889                   template <Addable T> T add (T a, T b) { return a + b; }
1890
1891       -fconstexpr-depth=n
1892           Set the maximum nested evaluation depth for C++11 constexpr
1893           functions to n.  A limit is needed to detect endless recursion
1894           during constant expression evaluation.  The minimum specified by
1895           the standard is 512.
1896
1897       -fconstexpr-loop-limit=n
1898           Set the maximum number of iterations for a loop in C++14 constexpr
1899           functions to n.  A limit is needed to detect infinite loops during
1900           constant expression evaluation.  The default is 262144 (1<<18).
1901
1902       -fconstexpr-ops-limit=n
1903           Set the maximum number of operations during a single constexpr
1904           evaluation.  Even when number of iterations of a single loop is
1905           limited with the above limit, if there are several nested loops and
1906           each of them has many iterations but still smaller than the above
1907           limit, or if in a body of some loop or even outside of a loop too
1908           many expressions need to be evaluated, the resulting constexpr
1909           evaluation might take too long.  The default is 33554432 (1<<25).
1910
1911       -fdeduce-init-list
1912           Enable deduction of a template type parameter as
1913           "std::initializer_list" from a brace-enclosed initializer list,
1914           i.e.
1915
1916                   template <class T> auto forward(T t) -> decltype (realfn (t))
1917                   {
1918                     return realfn (t);
1919                   }
1920
1921                   void f()
1922                   {
1923                     forward({1,2}); // call forward<std::initializer_list<int>>
1924                   }
1925
1926           This deduction was implemented as a possible extension to the
1927           originally proposed semantics for the C++11 standard, but was not
1928           part of the final standard, so it is disabled by default.  This
1929           option is deprecated, and may be removed in a future version of
1930           G++.
1931
1932       -fno-elide-constructors
1933           The C++ standard allows an implementation to omit creating a
1934           temporary that is only used to initialize another object of the
1935           same type.  Specifying this option disables that optimization, and
1936           forces G++ to call the copy constructor in all cases.  This option
1937           also causes G++ to call trivial member functions which otherwise
1938           would be expanded inline.
1939
1940           In C++17, the compiler is required to omit these temporaries, but
1941           this option still affects trivial member functions.
1942
1943       -fno-enforce-eh-specs
1944           Don't generate code to check for violation of exception
1945           specifications at run time.  This option violates the C++ standard,
1946           but may be useful for reducing code size in production builds, much
1947           like defining "NDEBUG".  This does not give user code permission to
1948           throw exceptions in violation of the exception specifications; the
1949           compiler still optimizes based on the specifications, so throwing
1950           an unexpected exception results in undefined behavior at run time.
1951
1952       -fextern-tls-init
1953       -fno-extern-tls-init
1954           The C++11 and OpenMP standards allow "thread_local" and
1955           "threadprivate" variables to have dynamic (runtime) initialization.
1956           To support this, any use of such a variable goes through a wrapper
1957           function that performs any necessary initialization.  When the use
1958           and definition of the variable are in the same translation unit,
1959           this overhead can be optimized away, but when the use is in a
1960           different translation unit there is significant overhead even if
1961           the variable doesn't actually need dynamic initialization.  If the
1962           programmer can be sure that no use of the variable in a non-
1963           defining TU needs to trigger dynamic initialization (either because
1964           the variable is statically initialized, or a use of the variable in
1965           the defining TU will be executed before any uses in another TU),
1966           they can avoid this overhead with the -fno-extern-tls-init option.
1967
1968           On targets that support symbol aliases, the default is
1969           -fextern-tls-init.  On targets that do not support symbol aliases,
1970           the default is -fno-extern-tls-init.
1971
1972       -fno-gnu-keywords
1973           Do not recognize "typeof" as a keyword, so that code can use this
1974           word as an identifier.  You can use the keyword "__typeof__"
1975           instead.  This option is implied by the strict ISO C++ dialects:
1976           -ansi, -std=c++98, -std=c++11, etc.
1977
1978       -fno-implicit-templates
1979           Never emit code for non-inline templates that are instantiated
1980           implicitly (i.e. by use); only emit code for explicit
1981           instantiations.  If you use this option, you must take care to
1982           structure your code to include all the necessary explicit
1983           instantiations to avoid getting undefined symbols at link time.
1984
1985       -fno-implicit-inline-templates
1986           Don't emit code for implicit instantiations of inline templates,
1987           either.  The default is to handle inlines differently so that
1988           compiles with and without optimization need the same set of
1989           explicit instantiations.
1990
1991       -fno-implement-inlines
1992           To save space, do not emit out-of-line copies of inline functions
1993           controlled by "#pragma implementation".  This causes linker errors
1994           if these functions are not inlined everywhere they are called.
1995
1996       -fms-extensions
1997           Disable Wpedantic warnings about constructs used in MFC, such as
1998           implicit int and getting a pointer to member function via non-
1999           standard syntax.
2000
2001       -fnew-inheriting-ctors
2002           Enable the P0136 adjustment to the semantics of C++11 constructor
2003           inheritance.  This is part of C++17 but also considered to be a
2004           Defect Report against C++11 and C++14.  This flag is enabled by
2005           default unless -fabi-version=10 or lower is specified.
2006
2007       -fnew-ttp-matching
2008           Enable the P0522 resolution to Core issue 150, template template
2009           parameters and default arguments: this allows a template with
2010           default template arguments as an argument for a template template
2011           parameter with fewer template parameters.  This flag is enabled by
2012           default for -std=c++17.
2013
2014       -fno-nonansi-builtins
2015           Disable built-in declarations of functions that are not mandated by
2016           ANSI/ISO C.  These include "ffs", "alloca", "_exit", "index",
2017           "bzero", "conjf", and other related functions.
2018
2019       -fnothrow-opt
2020           Treat a "throw()" exception specification as if it were a
2021           "noexcept" specification to reduce or eliminate the text size
2022           overhead relative to a function with no exception specification.
2023           If the function has local variables of types with non-trivial
2024           destructors, the exception specification actually makes the
2025           function smaller because the EH cleanups for those variables can be
2026           optimized away.  The semantic effect is that an exception thrown
2027           out of a function with such an exception specification results in a
2028           call to "terminate" rather than "unexpected".
2029
2030       -fno-operator-names
2031           Do not treat the operator name keywords "and", "bitand", "bitor",
2032           "compl", "not", "or" and "xor" as synonyms as keywords.
2033
2034       -fno-optional-diags
2035           Disable diagnostics that the standard says a compiler does not need
2036           to issue.  Currently, the only such diagnostic issued by G++ is the
2037           one for a name having multiple meanings within a class.
2038
2039       -fpermissive
2040           Downgrade some diagnostics about nonconformant code from errors to
2041           warnings.  Thus, using -fpermissive allows some nonconforming code
2042           to compile.
2043
2044       -fno-pretty-templates
2045           When an error message refers to a specialization of a function
2046           template, the compiler normally prints the signature of the
2047           template followed by the template arguments and any typedefs or
2048           typenames in the signature (e.g. "void f(T) [with T = int]" rather
2049           than "void f(int)") so that it's clear which template is involved.
2050           When an error message refers to a specialization of a class
2051           template, the compiler omits any template arguments that match the
2052           default template arguments for that template.  If either of these
2053           behaviors make it harder to understand the error message rather
2054           than easier, you can use -fno-pretty-templates to disable them.
2055
2056       -frepo
2057           Enable automatic template instantiation at link time.  This option
2058           also implies -fno-implicit-templates.
2059
2060       -fno-rtti
2061           Disable generation of information about every class with virtual
2062           functions for use by the C++ run-time type identification features
2063           ("dynamic_cast" and "typeid").  If you don't use those parts of the
2064           language, you can save some space by using this flag.  Note that
2065           exception handling uses the same information, but G++ generates it
2066           as needed. The "dynamic_cast" operator can still be used for casts
2067           that do not require run-time type information, i.e. casts to "void
2068           *" or to unambiguous base classes.
2069
2070           Mixing code compiled with -frtti with that compiled with -fno-rtti
2071           may not work.  For example, programs may fail to link if a class
2072           compiled with -fno-rtti is used as a base for a class compiled with
2073           -frtti.
2074
2075       -fsized-deallocation
2076           Enable the built-in global declarations
2077
2078                   void operator delete (void *, std::size_t) noexcept;
2079                   void operator delete[] (void *, std::size_t) noexcept;
2080
2081           as introduced in C++14.  This is useful for user-defined
2082           replacement deallocation functions that, for example, use the size
2083           of the object to make deallocation faster.  Enabled by default
2084           under -std=c++14 and above.  The flag -Wsized-deallocation warns
2085           about places that might want to add a definition.
2086
2087       -fstrict-enums
2088           Allow the compiler to optimize using the assumption that a value of
2089           enumerated type can only be one of the values of the enumeration
2090           (as defined in the C++ standard; basically, a value that can be
2091           represented in the minimum number of bits needed to represent all
2092           the enumerators).  This assumption may not be valid if the program
2093           uses a cast to convert an arbitrary integer value to the enumerated
2094           type.
2095
2096       -fstrong-eval-order
2097           Evaluate member access, array subscripting, and shift expressions
2098           in left-to-right order, and evaluate assignment in right-to-left
2099           order, as adopted for C++17.  Enabled by default with -std=c++17.
2100           -fstrong-eval-order=some enables just the ordering of member access
2101           and shift expressions, and is the default without -std=c++17.
2102
2103       -ftemplate-backtrace-limit=n
2104           Set the maximum number of template instantiation notes for a single
2105           warning or error to n.  The default value is 10.
2106
2107       -ftemplate-depth=n
2108           Set the maximum instantiation depth for template classes to n.  A
2109           limit on the template instantiation depth is needed to detect
2110           endless recursions during template class instantiation.  ANSI/ISO
2111           C++ conforming programs must not rely on a maximum depth greater
2112           than 17 (changed to 1024 in C++11).  The default value is 900, as
2113           the compiler can run out of stack space before hitting 1024 in some
2114           situations.
2115
2116       -fno-threadsafe-statics
2117           Do not emit the extra code to use the routines specified in the C++
2118           ABI for thread-safe initialization of local statics.  You can use
2119           this option to reduce code size slightly in code that doesn't need
2120           to be thread-safe.
2121
2122       -fuse-cxa-atexit
2123           Register destructors for objects with static storage duration with
2124           the "__cxa_atexit" function rather than the "atexit" function.
2125           This option is required for fully standards-compliant handling of
2126           static destructors, but only works if your C library supports
2127           "__cxa_atexit".
2128
2129       -fno-use-cxa-get-exception-ptr
2130           Don't use the "__cxa_get_exception_ptr" runtime routine.  This
2131           causes "std::uncaught_exception" to be incorrect, but is necessary
2132           if the runtime routine is not available.
2133
2134       -fvisibility-inlines-hidden
2135           This switch declares that the user does not attempt to compare
2136           pointers to inline functions or methods where the addresses of the
2137           two functions are taken in different shared objects.
2138
2139           The effect of this is that GCC may, effectively, mark inline
2140           methods with "__attribute__ ((visibility ("hidden")))" so that they
2141           do not appear in the export table of a DSO and do not require a PLT
2142           indirection when used within the DSO.  Enabling this option can
2143           have a dramatic effect on load and link times of a DSO as it
2144           massively reduces the size of the dynamic export table when the
2145           library makes heavy use of templates.
2146
2147           The behavior of this switch is not quite the same as marking the
2148           methods as hidden directly, because it does not affect static
2149           variables local to the function or cause the compiler to deduce
2150           that the function is defined in only one shared object.
2151
2152           You may mark a method as having a visibility explicitly to negate
2153           the effect of the switch for that method.  For example, if you do
2154           want to compare pointers to a particular inline method, you might
2155           mark it as having default visibility.  Marking the enclosing class
2156           with explicit visibility has no effect.
2157
2158           Explicitly instantiated inline methods are unaffected by this
2159           option as their linkage might otherwise cross a shared library
2160           boundary.
2161
2162       -fvisibility-ms-compat
2163           This flag attempts to use visibility settings to make GCC's C++
2164           linkage model compatible with that of Microsoft Visual Studio.
2165
2166           The flag makes these changes to GCC's linkage model:
2167
2168           1.  It sets the default visibility to "hidden", like
2169               -fvisibility=hidden.
2170
2171           2.  Types, but not their members, are not hidden by default.
2172
2173           3.  The One Definition Rule is relaxed for types without explicit
2174               visibility specifications that are defined in more than one
2175               shared object: those declarations are permitted if they are
2176               permitted when this option is not used.
2177
2178           In new code it is better to use -fvisibility=hidden and export
2179           those classes that are intended to be externally visible.
2180           Unfortunately it is possible for code to rely, perhaps
2181           accidentally, on the Visual Studio behavior.
2182
2183           Among the consequences of these changes are that static data
2184           members of the same type with the same name but defined in
2185           different shared objects are different, so changing one does not
2186           change the other; and that pointers to function members defined in
2187           different shared objects may not compare equal.  When this flag is
2188           given, it is a violation of the ODR to define types with the same
2189           name differently.
2190
2191       -fno-weak
2192           Do not use weak symbol support, even if it is provided by the
2193           linker.  By default, G++ uses weak symbols if they are available.
2194           This option exists only for testing, and should not be used by end-
2195           users; it results in inferior code and has no benefits.  This
2196           option may be removed in a future release of G++.
2197
2198       -nostdinc++
2199           Do not search for header files in the standard directories specific
2200           to C++, but do still search the other standard directories.  (This
2201           option is used when building the C++ library.)
2202
2203       In addition, these optimization, warning, and code generation options
2204       have meanings only for C++ programs:
2205
2206       -Wabi (C, Objective-C, C++ and Objective-C++ only)
2207           Warn when G++ it generates code that is probably not compatible
2208           with the vendor-neutral C++ ABI.  Since G++ now defaults to
2209           updating the ABI with each major release, normally -Wabi will warn
2210           only if there is a check added later in a release series for an ABI
2211           issue discovered since the initial release.  -Wabi will warn about
2212           more things if an older ABI version is selected (with
2213           -fabi-version=n).
2214
2215           -Wabi can also be used with an explicit version number to warn
2216           about compatibility with a particular -fabi-version level, e.g.
2217           -Wabi=2 to warn about changes relative to -fabi-version=2.
2218
2219           If an explicit version number is provided and -fabi-compat-version
2220           is not specified, the version number from this option is used for
2221           compatibility aliases.  If no explicit version number is provided
2222           with this option, but -fabi-compat-version is specified, that
2223           version number is used for ABI warnings.
2224
2225           Although an effort has been made to warn about all such cases,
2226           there are probably some cases that are not warned about, even
2227           though G++ is generating incompatible code.  There may also be
2228           cases where warnings are emitted even though the code that is
2229           generated is compatible.
2230
2231           You should rewrite your code to avoid these warnings if you are
2232           concerned about the fact that code generated by G++ may not be
2233           binary compatible with code generated by other compilers.
2234
2235           Known incompatibilities in -fabi-version=2 (which was the default
2236           from GCC 3.4 to 4.9) include:
2237
2238           *   A template with a non-type template parameter of reference type
2239               was mangled incorrectly:
2240
2241                       extern int N;
2242                       template <int &> struct S {};
2243                       void n (S<N>) {2}
2244
2245               This was fixed in -fabi-version=3.
2246
2247           *   SIMD vector types declared using "__attribute ((vector_size))"
2248               were mangled in a non-standard way that does not allow for
2249               overloading of functions taking vectors of different sizes.
2250
2251               The mangling was changed in -fabi-version=4.
2252
2253           *   "__attribute ((const))" and "noreturn" were mangled as type
2254               qualifiers, and "decltype" of a plain declaration was folded
2255               away.
2256
2257               These mangling issues were fixed in -fabi-version=5.
2258
2259           *   Scoped enumerators passed as arguments to a variadic function
2260               are promoted like unscoped enumerators, causing "va_arg" to
2261               complain.  On most targets this does not actually affect the
2262               parameter passing ABI, as there is no way to pass an argument
2263               smaller than "int".
2264
2265               Also, the ABI changed the mangling of template argument packs,
2266               "const_cast", "static_cast", prefix increment/decrement, and a
2267               class scope function used as a template argument.
2268
2269               These issues were corrected in -fabi-version=6.
2270
2271           *   Lambdas in default argument scope were mangled incorrectly, and
2272               the ABI changed the mangling of "nullptr_t".
2273
2274               These issues were corrected in -fabi-version=7.
2275
2276           *   When mangling a function type with function-cv-qualifiers, the
2277               un-qualified function type was incorrectly treated as a
2278               substitution candidate.
2279
2280               This was fixed in -fabi-version=8, the default for GCC 5.1.
2281
2282           *   "decltype(nullptr)" incorrectly had an alignment of 1, leading
2283               to unaligned accesses.  Note that this did not affect the ABI
2284               of a function with a "nullptr_t" parameter, as parameters have
2285               a minimum alignment.
2286
2287               This was fixed in -fabi-version=9, the default for GCC 5.2.
2288
2289           *   Target-specific attributes that affect the identity of a type,
2290               such as ia32 calling conventions on a function type (stdcall,
2291               regparm, etc.), did not affect the mangled name, leading to
2292               name collisions when function pointers were used as template
2293               arguments.
2294
2295               This was fixed in -fabi-version=10, the default for GCC 6.1.
2296
2297           It also warns about psABI-related changes.  The known psABI changes
2298           at this point include:
2299
2300           *   For SysV/x86-64, unions with "long double" members are passed
2301               in memory as specified in psABI.  For example:
2302
2303                       union U {
2304                         long double ld;
2305                         int i;
2306                       };
2307
2308               "union U" is always passed in memory.
2309
2310       -Wabi-tag (C++ and Objective-C++ only)
2311           Warn when a type with an ABI tag is used in a context that does not
2312           have that ABI tag.  See C++ Attributes for more information about
2313           ABI tags.
2314
2315       -Wctor-dtor-privacy (C++ and Objective-C++ only)
2316           Warn when a class seems unusable because all the constructors or
2317           destructors in that class are private, and it has neither friends
2318           nor public static member functions.  Also warn if there are no non-
2319           private methods, and there's at least one private member function
2320           that isn't a constructor or destructor.
2321
2322       -Wdelete-non-virtual-dtor (C++ and Objective-C++ only)
2323           Warn when "delete" is used to destroy an instance of a class that
2324           has virtual functions and non-virtual destructor. It is unsafe to
2325           delete an instance of a derived class through a pointer to a base
2326           class if the base class does not have a virtual destructor.  This
2327           warning is enabled by -Wall.
2328
2329       -Wdeprecated-copy (C++ and Objective-C++ only)
2330           Warn that the implicit declaration of a copy constructor or copy
2331           assignment operator is deprecated if the class has a user-provided
2332           copy constructor or copy assignment operator, in C++11 and up.
2333           This warning is enabled by -Wextra.  With -Wdeprecated-copy-dtor,
2334           also deprecate if the class has a user-provided destructor.
2335
2336       -Wno-init-list-lifetime (C++ and Objective-C++ only)
2337           Do not warn about uses of "std::initializer_list" that are likely
2338           to result in dangling pointers.  Since the underlying array for an
2339           "initializer_list" is handled like a normal C++ temporary object,
2340           it is easy to inadvertently keep a pointer to the array past the
2341           end of the array's lifetime.  For example:
2342
2343           *   If a function returns a temporary "initializer_list", or a
2344               local "initializer_list" variable, the array's lifetime ends at
2345               the end of the return statement, so the value returned has a
2346               dangling pointer.
2347
2348           *   If a new-expression creates an "initializer_list", the array
2349               only lives until the end of the enclosing full-expression, so
2350               the "initializer_list" in the heap has a dangling pointer.
2351
2352           *   When an "initializer_list" variable is assigned from a brace-
2353               enclosed initializer list, the temporary array created for the
2354               right side of the assignment only lives until the end of the
2355               full-expression, so at the next statement the
2356               "initializer_list" variable has a dangling pointer.
2357
2358                       // li's initial underlying array lives as long as li
2359                       std::initializer_list<int> li = { 1,2,3 };
2360                       // assignment changes li to point to a temporary array
2361                       li = { 4, 5 };
2362                       // now the temporary is gone and li has a dangling pointer
2363                       int i = li.begin()[0] // undefined behavior
2364
2365           *   When a list constructor stores the "begin" pointer from the
2366               "initializer_list" argument, this doesn't extend the lifetime
2367               of the array, so if a class variable is constructed from a
2368               temporary "initializer_list", the pointer is left dangling by
2369               the end of the variable declaration statement.
2370
2371       -Wliteral-suffix (C++ and Objective-C++ only)
2372           Warn when a string or character literal is followed by a ud-suffix
2373           which does not begin with an underscore.  As a conforming
2374           extension, GCC treats such suffixes as separate preprocessing
2375           tokens in order to maintain backwards compatibility with code that
2376           uses formatting macros from "<inttypes.h>".  For example:
2377
2378                   #define __STDC_FORMAT_MACROS
2379                   #include <inttypes.h>
2380                   #include <stdio.h>
2381
2382                   int main() {
2383                     int64_t i64 = 123;
2384                     printf("My int64: %" PRId64"\n", i64);
2385                   }
2386
2387           In this case, "PRId64" is treated as a separate preprocessing
2388           token.
2389
2390           Additionally, warn when a user-defined literal operator is declared
2391           with a literal suffix identifier that doesn't begin with an
2392           underscore. Literal suffix identifiers that don't begin with an
2393           underscore are reserved for future standardization.
2394
2395           This warning is enabled by default.
2396
2397       -Wlto-type-mismatch
2398           During the link-time optimization warn about type mismatches in
2399           global declarations from different compilation units.  Requires
2400           -flto to be enabled.  Enabled by default.
2401
2402       -Wno-narrowing (C++ and Objective-C++ only)
2403           For C++11 and later standards, narrowing conversions are diagnosed
2404           by default, as required by the standard.  A narrowing conversion
2405           from a constant produces an error, and a narrowing conversion from
2406           a non-constant produces a warning, but -Wno-narrowing suppresses
2407           the diagnostic.  Note that this does not affect the meaning of
2408           well-formed code; narrowing conversions are still considered ill-
2409           formed in SFINAE contexts.
2410
2411           With -Wnarrowing in C++98, warn when a narrowing conversion
2412           prohibited by C++11 occurs within { }, e.g.
2413
2414                   int i = { 2.2 }; // error: narrowing from double to int
2415
2416           This flag is included in -Wall and -Wc++11-compat.
2417
2418       -Wnoexcept (C++ and Objective-C++ only)
2419           Warn when a noexcept-expression evaluates to false because of a
2420           call to a function that does not have a non-throwing exception
2421           specification (i.e. "throw()" or "noexcept") but is known by the
2422           compiler to never throw an exception.
2423
2424       -Wnoexcept-type (C++ and Objective-C++ only)
2425           Warn if the C++17 feature making "noexcept" part of a function type
2426           changes the mangled name of a symbol relative to C++14.  Enabled by
2427           -Wabi and -Wc++17-compat.
2428
2429           As an example:
2430
2431                   template <class T> void f(T t) { t(); };
2432                   void g() noexcept;
2433                   void h() { f(g); }
2434
2435           In C++14, "f" calls "f<void(*)()>", but in C++17 it calls
2436           "f<void(*)()noexcept>".
2437
2438       -Wclass-memaccess (C++ and Objective-C++ only)
2439           Warn when the destination of a call to a raw memory function such
2440           as "memset" or "memcpy" is an object of class type, and when
2441           writing into such an object might bypass the class non-trivial or
2442           deleted constructor or copy assignment, violate const-correctness
2443           or encapsulation, or corrupt virtual table pointers.  Modifying the
2444           representation of such objects may violate invariants maintained by
2445           member functions of the class.  For example, the call to "memset"
2446           below is undefined because it modifies a non-trivial class object
2447           and is, therefore, diagnosed.  The safe way to either initialize or
2448           clear the storage of objects of such types is by using the
2449           appropriate constructor or assignment operator, if one is
2450           available.
2451
2452                   std::string str = "abc";
2453                   memset (&str, 0, sizeof str);
2454
2455           The -Wclass-memaccess option is enabled by -Wall.  Explicitly
2456           casting the pointer to the class object to "void *" or to a type
2457           that can be safely accessed by the raw memory function suppresses
2458           the warning.
2459
2460       -Wnon-virtual-dtor (C++ and Objective-C++ only)
2461           Warn when a class has virtual functions and an accessible non-
2462           virtual destructor itself or in an accessible polymorphic base
2463           class, in which case it is possible but unsafe to delete an
2464           instance of a derived class through a pointer to the class itself
2465           or base class.  This warning is automatically enabled if -Weffc++
2466           is specified.
2467
2468       -Wregister (C++ and Objective-C++ only)
2469           Warn on uses of the "register" storage class specifier, except when
2470           it is part of the GNU Explicit Register Variables extension.  The
2471           use of the "register" keyword as storage class specifier has been
2472           deprecated in C++11 and removed in C++17.  Enabled by default with
2473           -std=c++17.
2474
2475       -Wreorder (C++ and Objective-C++ only)
2476           Warn when the order of member initializers given in the code does
2477           not match the order in which they must be executed.  For instance:
2478
2479                   struct A {
2480                     int i;
2481                     int j;
2482                     A(): j (0), i (1) { }
2483                   };
2484
2485           The compiler rearranges the member initializers for "i" and "j" to
2486           match the declaration order of the members, emitting a warning to
2487           that effect.  This warning is enabled by -Wall.
2488
2489       -Wno-pessimizing-move (C++ and Objective-C++ only)
2490           This warning warns when a call to "std::move" prevents copy
2491           elision.  A typical scenario when copy elision can occur is when
2492           returning in a function with a class return type, when the
2493           expression being returned is the name of a non-volatile automatic
2494           object, and is not a function parameter, and has the same type as
2495           the function return type.
2496
2497                   struct T {
2498                   ...
2499                   };
2500                   T fn()
2501                   {
2502                     T t;
2503                     ...
2504                     return std::move (t);
2505                   }
2506
2507           But in this example, the "std::move" call prevents copy elision.
2508
2509           This warning is enabled by -Wall.
2510
2511       -Wno-redundant-move (C++ and Objective-C++ only)
2512           This warning warns about redundant calls to "std::move"; that is,
2513           when a move operation would have been performed even without the
2514           "std::move" call.  This happens because the compiler is forced to
2515           treat the object as if it were an rvalue in certain situations such
2516           as returning a local variable, where copy elision isn't applicable.
2517           Consider:
2518
2519                   struct T {
2520                   ...
2521                   };
2522                   T fn(T t)
2523                   {
2524                     ...
2525                     return std::move (t);
2526                   }
2527
2528           Here, the "std::move" call is redundant.  Because G++ implements
2529           Core Issue 1579, another example is:
2530
2531                   struct T { // convertible to U
2532                   ...
2533                   };
2534                   struct U {
2535                   ...
2536                   };
2537                   U fn()
2538                   {
2539                     T t;
2540                     ...
2541                     return std::move (t);
2542                   }
2543
2544           In this example, copy elision isn't applicable because the type of
2545           the expression being returned and the function return type differ,
2546           yet G++ treats the return value as if it were designated by an
2547           rvalue.
2548
2549           This warning is enabled by -Wextra.
2550
2551       -fext-numeric-literals (C++ and Objective-C++ only)
2552           Accept imaginary, fixed-point, or machine-defined literal number
2553           suffixes as GNU extensions.  When this option is turned off these
2554           suffixes are treated as C++11 user-defined literal numeric
2555           suffixes.  This is on by default for all pre-C++11 dialects and all
2556           GNU dialects: -std=c++98, -std=gnu++98, -std=gnu++11, -std=gnu++14.
2557           This option is off by default for ISO C++11 onwards (-std=c++11,
2558           ...).
2559
2560       The following -W... options are not affected by -Wall.
2561
2562       -Weffc++ (C++ and Objective-C++ only)
2563           Warn about violations of the following style guidelines from Scott
2564           Meyers' Effective C++ series of books:
2565
2566           *   Define a copy constructor and an assignment operator for
2567               classes with dynamically-allocated memory.
2568
2569           *   Prefer initialization to assignment in constructors.
2570
2571           *   Have "operator=" return a reference to *this.
2572
2573           *   Don't try to return a reference when you must return an object.
2574
2575           *   Distinguish between prefix and postfix forms of increment and
2576               decrement operators.
2577
2578           *   Never overload "&&", "||", or ",".
2579
2580           This option also enables -Wnon-virtual-dtor, which is also one of
2581           the effective C++ recommendations.  However, the check is extended
2582           to warn about the lack of virtual destructor in accessible non-
2583           polymorphic bases classes too.
2584
2585           When selecting this option, be aware that the standard library
2586           headers do not obey all of these guidelines; use grep -v to filter
2587           out those warnings.
2588
2589       -Wstrict-null-sentinel (C++ and Objective-C++ only)
2590           Warn about the use of an uncasted "NULL" as sentinel.  When
2591           compiling only with GCC this is a valid sentinel, as "NULL" is
2592           defined to "__null".  Although it is a null pointer constant rather
2593           than a null pointer, it is guaranteed to be of the same size as a
2594           pointer.  But this use is not portable across different compilers.
2595
2596       -Wno-non-template-friend (C++ and Objective-C++ only)
2597           Disable warnings when non-template friend functions are declared
2598           within a template.  In very old versions of GCC that predate
2599           implementation of the ISO standard, declarations such as friend int
2600           foo(int), where the name of the friend is an unqualified-id, could
2601           be interpreted as a particular specialization of a template
2602           function; the warning exists to diagnose compatibility problems,
2603           and is enabled by default.
2604
2605       -Wold-style-cast (C++ and Objective-C++ only)
2606           Warn if an old-style (C-style) cast to a non-void type is used
2607           within a C++ program.  The new-style casts ("dynamic_cast",
2608           "static_cast", "reinterpret_cast", and "const_cast") are less
2609           vulnerable to unintended effects and much easier to search for.
2610
2611       -Woverloaded-virtual (C++ and Objective-C++ only)
2612           Warn when a function declaration hides virtual functions from a
2613           base class.  For example, in:
2614
2615                   struct A {
2616                     virtual void f();
2617                   };
2618
2619                   struct B: public A {
2620                     void f(int);
2621                   };
2622
2623           the "A" class version of "f" is hidden in "B", and code like:
2624
2625                   B* b;
2626                   b->f();
2627
2628           fails to compile.
2629
2630       -Wno-pmf-conversions (C++ and Objective-C++ only)
2631           Disable the diagnostic for converting a bound pointer to member
2632           function to a plain pointer.
2633
2634       -Wsign-promo (C++ and Objective-C++ only)
2635           Warn when overload resolution chooses a promotion from unsigned or
2636           enumerated type to a signed type, over a conversion to an unsigned
2637           type of the same size.  Previous versions of G++ tried to preserve
2638           unsignedness, but the standard mandates the current behavior.
2639
2640       -Wtemplates (C++ and Objective-C++ only)
2641           Warn when a primary template declaration is encountered.  Some
2642           coding rules disallow templates, and this may be used to enforce
2643           that rule.  The warning is inactive inside a system header file,
2644           such as the STL, so one can still use the STL.  One may also
2645           instantiate or specialize templates.
2646
2647       -Wmultiple-inheritance (C++ and Objective-C++ only)
2648           Warn when a class is defined with multiple direct base classes.
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 multiple inheritance.
2653
2654       -Wvirtual-inheritance
2655           Warn when a class is defined with a virtual direct base class.
2656           Some coding rules disallow multiple inheritance, and this may be
2657           used to enforce that rule.  The warning is inactive inside a system
2658           header file, such as the STL, so one can still use the STL.  One
2659           may also define classes that indirectly use virtual inheritance.
2660
2661       -Wnamespaces
2662           Warn when a namespace definition is opened.  Some coding rules
2663           disallow namespaces, and this may be used to enforce that rule.
2664           The warning is inactive inside a system header file, such as the
2665           STL, so one can still use the STL.  One may also use using
2666           directives and qualified names.
2667
2668       -Wno-terminate (C++ and Objective-C++ only)
2669           Disable the warning about a throw-expression that will immediately
2670           result in a call to "terminate".
2671
2672       -Wno-class-conversion (C++ and Objective-C++ only)
2673           Disable the warning about the case when a conversion function
2674           converts an object to the same type, to a base class of that type,
2675           or to void; such a conversion function will never be called.
2676
2677   Options Controlling Objective-C and Objective-C++ Dialects
2678       (NOTE: This manual does not describe the Objective-C and Objective-C++
2679       languages themselves.
2680
2681       This section describes the command-line options that are only
2682       meaningful for Objective-C and Objective-C++ programs.  You can also
2683       use most of the language-independent GNU compiler options.  For
2684       example, you might compile a file some_class.m like this:
2685
2686               gcc -g -fgnu-runtime -O -c some_class.m
2687
2688       In this example, -fgnu-runtime is an option meant only for Objective-C
2689       and Objective-C++ programs; you can use the other options with any
2690       language supported by GCC.
2691
2692       Note that since Objective-C is an extension of the C language,
2693       Objective-C compilations may also use options specific to the C front-
2694       end (e.g., -Wtraditional).  Similarly, Objective-C++ compilations may
2695       use C++-specific options (e.g., -Wabi).
2696
2697       Here is a list of options that are only for compiling Objective-C and
2698       Objective-C++ programs:
2699
2700       -fconstant-string-class=class-name
2701           Use class-name as the name of the class to instantiate for each
2702           literal string specified with the syntax "@"..."".  The default
2703           class name is "NXConstantString" if the GNU runtime is being used,
2704           and "NSConstantString" if the NeXT runtime is being used (see
2705           below).  The -fconstant-cfstrings option, if also present,
2706           overrides the -fconstant-string-class setting and cause "@"...""
2707           literals to be laid out as constant CoreFoundation strings.
2708
2709       -fgnu-runtime
2710           Generate object code compatible with the standard GNU Objective-C
2711           runtime.  This is the default for most types of systems.
2712
2713       -fnext-runtime
2714           Generate output compatible with the NeXT runtime.  This is the
2715           default for NeXT-based systems, including Darwin and Mac OS X.  The
2716           macro "__NEXT_RUNTIME__" is predefined if (and only if) this option
2717           is used.
2718
2719       -fno-nil-receivers
2720           Assume that all Objective-C message dispatches ("[receiver
2721           message:arg]") in this translation unit ensure that the receiver is
2722           not "nil".  This allows for more efficient entry points in the
2723           runtime to be used.  This option is only available in conjunction
2724           with the NeXT runtime and ABI version 0 or 1.
2725
2726       -fobjc-abi-version=n
2727           Use version n of the Objective-C ABI for the selected runtime.
2728           This option is currently supported only for the NeXT runtime.  In
2729           that case, Version 0 is the traditional (32-bit) ABI without
2730           support for properties and other Objective-C 2.0 additions.
2731           Version 1 is the traditional (32-bit) ABI with support for
2732           properties and other Objective-C 2.0 additions.  Version 2 is the
2733           modern (64-bit) ABI.  If nothing is specified, the default is
2734           Version 0 on 32-bit target machines, and Version 2 on 64-bit target
2735           machines.
2736
2737       -fobjc-call-cxx-cdtors
2738           For each Objective-C class, check if any of its instance variables
2739           is a C++ object with a non-trivial default constructor.  If so,
2740           synthesize a special "- (id) .cxx_construct" instance method which
2741           runs non-trivial default constructors on any such instance
2742           variables, in order, and then return "self".  Similarly, check if
2743           any instance variable is a C++ object with a non-trivial
2744           destructor, and if so, synthesize a special "- (void)
2745           .cxx_destruct" method which runs all such default destructors, in
2746           reverse order.
2747
2748           The "- (id) .cxx_construct" and "- (void) .cxx_destruct" methods
2749           thusly generated only operate on instance variables declared in the
2750           current Objective-C class, and not those inherited from
2751           superclasses.  It is the responsibility of the Objective-C runtime
2752           to invoke all such methods in an object's inheritance hierarchy.
2753           The "- (id) .cxx_construct" methods are invoked by the runtime
2754           immediately after a new object instance is allocated; the "- (void)
2755           .cxx_destruct" methods are invoked immediately before the runtime
2756           deallocates an object instance.
2757
2758           As of this writing, only the NeXT runtime on Mac OS X 10.4 and
2759           later has support for invoking the "- (id) .cxx_construct" and "-
2760           (void) .cxx_destruct" methods.
2761
2762       -fobjc-direct-dispatch
2763           Allow fast jumps to the message dispatcher.  On Darwin this is
2764           accomplished via the comm page.
2765
2766       -fobjc-exceptions
2767           Enable syntactic support for structured exception handling in
2768           Objective-C, similar to what is offered by C++.  This option is
2769           required to use the Objective-C keywords @try, @throw, @catch,
2770           @finally and @synchronized.  This option is available with both the
2771           GNU runtime and the NeXT runtime (but not available in conjunction
2772           with the NeXT runtime on Mac OS X 10.2 and earlier).
2773
2774       -fobjc-gc
2775           Enable garbage collection (GC) in Objective-C and Objective-C++
2776           programs.  This option is only available with the NeXT runtime; the
2777           GNU runtime has a different garbage collection implementation that
2778           does not require special compiler flags.
2779
2780       -fobjc-nilcheck
2781           For the NeXT runtime with version 2 of the ABI, check for a nil
2782           receiver in method invocations before doing the actual method call.
2783           This is the default and can be disabled using -fno-objc-nilcheck.
2784           Class methods and super calls are never checked for nil in this way
2785           no matter what this flag is set to.  Currently this flag does
2786           nothing when the GNU runtime, or an older version of the NeXT
2787           runtime ABI, is used.
2788
2789       -fobjc-std=objc1
2790           Conform to the language syntax of Objective-C 1.0, the language
2791           recognized by GCC 4.0.  This only affects the Objective-C additions
2792           to the C/C++ language; it does not affect conformance to C/C++
2793           standards, which is controlled by the separate C/C++ dialect option
2794           flags.  When this option is used with the Objective-C or
2795           Objective-C++ compiler, any Objective-C syntax that is not
2796           recognized by GCC 4.0 is rejected.  This is useful if you need to
2797           make sure that your Objective-C code can be compiled with older
2798           versions of GCC.
2799
2800       -freplace-objc-classes
2801           Emit a special marker instructing ld(1) not to statically link in
2802           the resulting object file, and allow dyld(1) to load it in at run
2803           time instead.  This is used in conjunction with the Fix-and-
2804           Continue debugging mode, where the object file in question may be
2805           recompiled and dynamically reloaded in the course of program
2806           execution, without the need to restart the program itself.
2807           Currently, Fix-and-Continue functionality is only available in
2808           conjunction with the NeXT runtime on Mac OS X 10.3 and later.
2809
2810       -fzero-link
2811           When compiling for the NeXT runtime, the compiler ordinarily
2812           replaces calls to "objc_getClass("...")" (when the name of the
2813           class is known at compile time) with static class references that
2814           get initialized at load time, which improves run-time performance.
2815           Specifying the -fzero-link flag suppresses this behavior and causes
2816           calls to "objc_getClass("...")"  to be retained.  This is useful in
2817           Zero-Link debugging mode, since it allows for individual class
2818           implementations to be modified during program execution.  The GNU
2819           runtime currently always retains calls to "objc_get_class("...")"
2820           regardless of command-line options.
2821
2822       -fno-local-ivars
2823           By default instance variables in Objective-C can be accessed as if
2824           they were local variables from within the methods of the class
2825           they're declared in.  This can lead to shadowing between instance
2826           variables and other variables declared either locally inside a
2827           class method or globally with the same name.  Specifying the
2828           -fno-local-ivars flag disables this behavior thus avoiding variable
2829           shadowing issues.
2830
2831       -fivar-visibility=[public|protected|private|package]
2832           Set the default instance variable visibility to the specified
2833           option so that instance variables declared outside the scope of any
2834           access modifier directives default to the specified visibility.
2835
2836       -gen-decls
2837           Dump interface declarations for all classes seen in the source file
2838           to a file named sourcename.decl.
2839
2840       -Wassign-intercept (Objective-C and Objective-C++ only)
2841           Warn whenever an Objective-C assignment is being intercepted by the
2842           garbage collector.
2843
2844       -Wno-protocol (Objective-C and Objective-C++ only)
2845           If a class is declared to implement a protocol, a warning is issued
2846           for every method in the protocol that is not implemented by the
2847           class.  The default behavior is to issue a warning for every method
2848           not explicitly implemented in the class, even if a method
2849           implementation is inherited from the superclass.  If you use the
2850           -Wno-protocol option, then methods inherited from the superclass
2851           are considered to be implemented, and no warning is issued for
2852           them.
2853
2854       -Wselector (Objective-C and Objective-C++ only)
2855           Warn if multiple methods of different types for the same selector
2856           are found during compilation.  The check is performed on the list
2857           of methods in the final stage of compilation.  Additionally, a
2858           check is performed for each selector appearing in a
2859           "@selector(...)"  expression, and a corresponding method for that
2860           selector has been found during compilation.  Because these checks
2861           scan the method table only at the end of compilation, these
2862           warnings are not produced if the final stage of compilation is not
2863           reached, for example because an error is found during compilation,
2864           or because the -fsyntax-only option is being used.
2865
2866       -Wstrict-selector-match (Objective-C and Objective-C++ only)
2867           Warn if multiple methods with differing argument and/or return
2868           types are found for a given selector when attempting to send a
2869           message using this selector to a receiver of type "id" or "Class".
2870           When this flag is off (which is the default behavior), the compiler
2871           omits such warnings if any differences found are confined to types
2872           that share the same size and alignment.
2873
2874       -Wundeclared-selector (Objective-C and Objective-C++ only)
2875           Warn if a "@selector(...)" expression referring to an undeclared
2876           selector is found.  A selector is considered undeclared if no
2877           method with that name has been declared before the "@selector(...)"
2878           expression, either explicitly in an @interface or @protocol
2879           declaration, or implicitly in an @implementation section.  This
2880           option always performs its checks as soon as a "@selector(...)"
2881           expression is found, while -Wselector only performs its checks in
2882           the final stage of compilation.  This also enforces the coding
2883           style convention that methods and selectors must be declared before
2884           being used.
2885
2886       -print-objc-runtime-info
2887           Generate C header describing the largest structure that is passed
2888           by value, if any.
2889
2890   Options to Control Diagnostic Messages Formatting
2891       Traditionally, diagnostic messages have been formatted irrespective of
2892       the output device's aspect (e.g. its width, ...).  You can use the
2893       options described below to control the formatting algorithm for
2894       diagnostic messages, e.g. how many characters per line, how often
2895       source location information should be reported.  Note that some
2896       language front ends may not honor these options.
2897
2898       -fmessage-length=n
2899           Try to format error messages so that they fit on lines of about n
2900           characters.  If n is zero, then no line-wrapping is done; each
2901           error message appears on a single line.  This is the default for
2902           all front ends.
2903
2904           Note - this option also affects the display of the #error and
2905           #warning pre-processor directives, and the deprecated
2906           function/type/variable attribute.  It does not however affect the
2907           pragma GCC warning and pragma GCC error pragmas.
2908
2909       -fdiagnostics-show-location=once
2910           Only meaningful in line-wrapping mode.  Instructs the diagnostic
2911           messages reporter to emit source location information once; that
2912           is, in case the message is too long to fit on a single physical
2913           line and has to be wrapped, the source location won't be emitted
2914           (as prefix) again, over and over, in subsequent continuation lines.
2915           This is the default behavior.
2916
2917       -fdiagnostics-show-location=every-line
2918           Only meaningful in line-wrapping mode.  Instructs the diagnostic
2919           messages reporter to emit the same source location information (as
2920           prefix) for physical lines that result from the process of breaking
2921           a message which is too long to fit on a single line.
2922
2923       -fdiagnostics-color[=WHEN]
2924       -fno-diagnostics-color
2925           Use color in diagnostics.  WHEN is never, always, or auto.  The
2926           default depends on how the compiler has been configured, it can be
2927           any of the above WHEN options or also never if GCC_COLORS
2928           environment variable isn't present in the environment, and auto
2929           otherwise.  auto means to use color only when the standard error is
2930           a terminal.  The forms -fdiagnostics-color and
2931           -fno-diagnostics-color are aliases for -fdiagnostics-color=always
2932           and -fdiagnostics-color=never, respectively.
2933
2934           The colors are defined by the environment variable GCC_COLORS.  Its
2935           value is a colon-separated list of capabilities and Select Graphic
2936           Rendition (SGR) substrings. SGR commands are interpreted by the
2937           terminal or terminal emulator.  (See the section in the
2938           documentation of your text terminal for permitted values and their
2939           meanings as character attributes.)  These substring values are
2940           integers in decimal representation and can be concatenated with
2941           semicolons.  Common values to concatenate include 1 for bold, 4 for
2942           underline, 5 for blink, 7 for inverse, 39 for default foreground
2943           color, 30 to 37 for foreground colors, 90 to 97 for 16-color mode
2944           foreground colors, 38;5;0 to 38;5;255 for 88-color and 256-color
2945           modes foreground colors, 49 for default background color, 40 to 47
2946           for background colors, 100 to 107 for 16-color mode background
2947           colors, and 48;5;0 to 48;5;255 for 88-color and 256-color modes
2948           background colors.
2949
2950           The default GCC_COLORS is
2951
2952                   error=01;31:warning=01;35:note=01;36:range1=32:range2=34:locus=01:\
2953                   quote=01:fixit-insert=32:fixit-delete=31:\
2954                   diff-filename=01:diff-hunk=32:diff-delete=31:diff-insert=32:\
2955                   type-diff=01;32
2956
2957           where 01;31 is bold red, 01;35 is bold magenta, 01;36 is bold cyan,
2958           32 is green, 34 is blue, 01 is bold, and 31 is red.  Setting
2959           GCC_COLORS to the empty string disables colors.  Supported
2960           capabilities are as follows.
2961
2962           "error="
2963               SGR substring for error: markers.
2964
2965           "warning="
2966               SGR substring for warning: markers.
2967
2968           "note="
2969               SGR substring for note: markers.
2970
2971           "range1="
2972               SGR substring for first additional range.
2973
2974           "range2="
2975               SGR substring for second additional range.
2976
2977           "locus="
2978               SGR substring for location information, file:line or
2979               file:line:column etc.
2980
2981           "quote="
2982               SGR substring for information printed within quotes.
2983
2984           "fixit-insert="
2985               SGR substring for fix-it hints suggesting text to be inserted
2986               or replaced.
2987
2988           "fixit-delete="
2989               SGR substring for fix-it hints suggesting text to be deleted.
2990
2991           "diff-filename="
2992               SGR substring for filename headers within generated patches.
2993
2994           "diff-hunk="
2995               SGR substring for the starts of hunks within generated patches.
2996
2997           "diff-delete="
2998               SGR substring for deleted lines within generated patches.
2999
3000           "diff-insert="
3001               SGR substring for inserted lines within generated patches.
3002
3003           "type-diff="
3004               SGR substring for highlighting mismatching types within
3005               template arguments in the C++ frontend.
3006
3007       -fno-diagnostics-show-option
3008           By default, each diagnostic emitted includes text indicating the
3009           command-line option that directly controls the diagnostic (if such
3010           an option is known to the diagnostic machinery).  Specifying the
3011           -fno-diagnostics-show-option flag suppresses that behavior.
3012
3013       -fno-diagnostics-show-caret
3014           By default, each diagnostic emitted includes the original source
3015           line and a caret ^ indicating the column.  This option suppresses
3016           this information.  The source line is truncated to n characters, if
3017           the -fmessage-length=n option is given.  When the output is done to
3018           the terminal, the width is limited to the width given by the
3019           COLUMNS environment variable or, if not set, to the terminal width.
3020
3021       -fno-diagnostics-show-labels
3022           By default, when printing source code (via
3023           -fdiagnostics-show-caret), diagnostics can label ranges of source
3024           code with pertinent information, such as the types of expressions:
3025
3026                       printf ("foo %s bar", long_i + long_j);
3027                                    ~^       ~~~~~~~~~~~~~~~
3028                                     |              |
3029                                     char *         long int
3030
3031           This option suppresses the printing of these labels (in the example
3032           above, the vertical bars and the "char *" and "long int" text).
3033
3034       -fno-diagnostics-show-line-numbers
3035           By default, when printing source code (via
3036           -fdiagnostics-show-caret), a left margin is printed, showing line
3037           numbers.  This option suppresses this left margin.
3038
3039       -fdiagnostics-minimum-margin-width=width
3040           This option controls the minimum width of the left margin printed
3041           by -fdiagnostics-show-line-numbers.  It defaults to 6.
3042
3043       -fdiagnostics-parseable-fixits
3044           Emit fix-it hints in a machine-parseable format, suitable for
3045           consumption by IDEs.  For each fix-it, a line will be printed after
3046           the relevant diagnostic, starting with the string "fix-it:".  For
3047           example:
3048
3049                   fix-it:"test.c":{45:3-45:21}:"gtk_widget_show_all"
3050
3051           The location is expressed as a half-open range, expressed as a
3052           count of bytes, starting at byte 1 for the initial column.  In the
3053           above example, bytes 3 through 20 of line 45 of "test.c" are to be
3054           replaced with the given string:
3055
3056                   00000000011111111112222222222
3057                   12345678901234567890123456789
3058                     gtk_widget_showall (dlg);
3059                     ^^^^^^^^^^^^^^^^^^
3060                     gtk_widget_show_all
3061
3062           The filename and replacement string escape backslash as "\\", tab
3063           as "\t", newline as "\n", double quotes as "\"", non-printable
3064           characters as octal (e.g. vertical tab as "\013").
3065
3066           An empty replacement string indicates that the given range is to be
3067           removed.  An empty range (e.g. "45:3-45:3") indicates that the
3068           string is to be inserted at the given position.
3069
3070       -fdiagnostics-generate-patch
3071           Print fix-it hints to stderr in unified diff format, after any
3072           diagnostics are printed.  For example:
3073
3074                   --- test.c
3075                   +++ test.c
3076                   @ -42,5 +42,5 @
3077
3078                    void show_cb(GtkDialog *dlg)
3079                    {
3080                   -  gtk_widget_showall(dlg);
3081                   +  gtk_widget_show_all(dlg);
3082                    }
3083
3084           The diff may or may not be colorized, following the same rules as
3085           for diagnostics (see -fdiagnostics-color).
3086
3087       -fdiagnostics-show-template-tree
3088           In the C++ frontend, when printing diagnostics showing mismatching
3089           template types, such as:
3090
3091                     could not convert 'std::map<int, std::vector<double> >()'
3092                       from 'map<[...],vector<double>>' to 'map<[...],vector<float>>
3093
3094           the -fdiagnostics-show-template-tree flag enables printing a tree-
3095           like structure showing the common and differing parts of the types,
3096           such as:
3097
3098                     map<
3099                       [...],
3100                       vector<
3101                         [double != float]>>
3102
3103           The parts that differ are highlighted with color ("double" and
3104           "float" in this case).
3105
3106       -fno-elide-type
3107           By default when the C++ frontend prints diagnostics showing
3108           mismatching template types, common parts of the types are printed
3109           as "[...]" to simplify the error message.  For example:
3110
3111                     could not convert 'std::map<int, std::vector<double> >()'
3112                       from 'map<[...],vector<double>>' to 'map<[...],vector<float>>
3113
3114           Specifying the -fno-elide-type flag suppresses that behavior.  This
3115           flag also affects the output of the
3116           -fdiagnostics-show-template-tree flag.
3117
3118       -fno-show-column
3119           Do not print column numbers in diagnostics.  This may be necessary
3120           if diagnostics are being scanned by a program that does not
3121           understand the column numbers, such as dejagnu.
3122
3123       -fdiagnostics-format=FORMAT
3124           Select a different format for printing diagnostics.  FORMAT is text
3125           or json.  The default is text.
3126
3127           The json format consists of a top-level JSON array containing JSON
3128           objects representing the diagnostics.
3129
3130           The JSON is emitted as one line, without formatting; the examples
3131           below have been formatted for clarity.
3132
3133           Diagnostics can have child diagnostics.  For example, this error
3134           and note:
3135
3136                   misleading-indentation.c:15:3: warning: this 'if' clause does not
3137                     guard... [-Wmisleading-indentation]
3138                      15 |   if (flag)
3139                         |   ^~
3140                   misleading-indentation.c:17:5: note: ...this statement, but the latter
3141                     is misleadingly indented as if it were guarded by the 'if'
3142                      17 |     y = 2;
3143                         |     ^
3144
3145           might be printed in JSON form (after formatting) like this:
3146
3147                   [
3148                       {
3149                           "kind": "warning",
3150                           "locations": [
3151                               {
3152                                   "caret": {
3153                                       "column": 3,
3154                                       "file": "misleading-indentation.c",
3155                                       "line": 15
3156                                   },
3157                                   "finish": {
3158                                       "column": 4,
3159                                       "file": "misleading-indentation.c",
3160                                       "line": 15
3161                                   }
3162                               }
3163                           ],
3164                           "message": "this \u2018if\u2019 clause does not guard...",
3165                           "option": "-Wmisleading-indentation",
3166                           "children": [
3167                               {
3168                                   "kind": "note",
3169                                   "locations": [
3170                                       {
3171                                           "caret": {
3172                                               "column": 5,
3173                                               "file": "misleading-indentation.c",
3174                                               "line": 17
3175                                           }
3176                                       }
3177                                   ],
3178                                   "message": "...this statement, but the latter is ..."
3179                               }
3180                           ]
3181                       },
3182                       ...
3183                   ]
3184
3185           where the "note" is a child of the "warning".
3186
3187           A diagnostic has a "kind".  If this is "warning", then there is an
3188           "option" key describing the command-line option controlling the
3189           warning.
3190
3191           A diagnostic can contain zero or more locations.  Each location has
3192           up to three positions within it: a "caret" position and optional
3193           "start" and "finish" positions.  A location can also have an
3194           optional "label" string.  For example, this error:
3195
3196                   bad-binary-ops.c:64:23: error: invalid operands to binary + (have 'S' {aka
3197                      'struct s'} and 'T' {aka 'struct t'})
3198                      64 |   return callee_4a () + callee_4b ();
3199                         |          ~~~~~~~~~~~~ ^ ~~~~~~~~~~~~
3200                         |          |              |
3201                         |          |              T {aka struct t}
3202                         |          S {aka struct s}
3203
3204           has three locations.  Its primary location is at the "+" token at
3205           column 23.  It has two secondary locations, describing the left and
3206           right-hand sides of the expression, which have labels.  It might be
3207           printed in JSON form as:
3208
3209                       {
3210                           "children": [],
3211                           "kind": "error",
3212                           "locations": [
3213                               {
3214                                   "caret": {
3215                                       "column": 23, "file": "bad-binary-ops.c", "line": 64
3216                                   }
3217                               },
3218                               {
3219                                   "caret": {
3220                                       "column": 10, "file": "bad-binary-ops.c", "line": 64
3221                                   },
3222                                   "finish": {
3223                                       "column": 21, "file": "bad-binary-ops.c", "line": 64
3224                                   },
3225                                   "label": "S {aka struct s}"
3226                               },
3227                               {
3228                                   "caret": {
3229                                       "column": 25, "file": "bad-binary-ops.c", "line": 64
3230                                   },
3231                                   "finish": {
3232                                       "column": 36, "file": "bad-binary-ops.c", "line": 64
3233                                   },
3234                                   "label": "T {aka struct t}"
3235                               }
3236                           ],
3237                           "message": "invalid operands to binary + ..."
3238                       }
3239
3240           If a diagnostic contains fix-it hints, it has a "fixits" array,
3241           consisting of half-open intervals, similar to the output of
3242           -fdiagnostics-parseable-fixits.  For example, this diagnostic with
3243           a replacement fix-it hint:
3244
3245                   demo.c:8:15: error: 'struct s' has no member named 'colour'; did you
3246                     mean 'color'?
3247                       8 |   return ptr->colour;
3248                         |               ^~~~~~
3249                         |               color
3250
3251           might be printed in JSON form as:
3252
3253                       {
3254                           "children": [],
3255                           "fixits": [
3256                               {
3257                                   "next": {
3258                                       "column": 21,
3259                                       "file": "demo.c",
3260                                       "line": 8
3261                                   },
3262                                   "start": {
3263                                       "column": 15,
3264                                       "file": "demo.c",
3265                                       "line": 8
3266                                   },
3267                                   "string": "color"
3268                               }
3269                           ],
3270                           "kind": "error",
3271                           "locations": [
3272                               {
3273                                   "caret": {
3274                                       "column": 15,
3275                                       "file": "demo.c",
3276                                       "line": 8
3277                                   },
3278                                   "finish": {
3279                                       "column": 20,
3280                                       "file": "demo.c",
3281                                       "line": 8
3282                                   }
3283                               }
3284                           ],
3285                           "message": "\u2018struct s\u2019 has no member named ..."
3286                       }
3287
3288           where the fix-it hint suggests replacing the text from "start" up
3289           to but not including "next" with "string"'s value.  Deletions are
3290           expressed via an empty value for "string", insertions by having
3291           "start" equal "next".
3292
3293   Options to Request or Suppress Warnings
3294       Warnings are diagnostic messages that report constructions that are not
3295       inherently erroneous but that are risky or suggest there may have been
3296       an error.
3297
3298       The following language-independent options do not enable specific
3299       warnings but control the kinds of diagnostics produced by GCC.
3300
3301       -fsyntax-only
3302           Check the code for syntax errors, but don't do anything beyond
3303           that.
3304
3305       -fmax-errors=n
3306           Limits the maximum number of error messages to n, at which point
3307           GCC bails out rather than attempting to continue processing the
3308           source code.  If n is 0 (the default), there is no limit on the
3309           number of error messages produced.  If -Wfatal-errors is also
3310           specified, then -Wfatal-errors takes precedence over this option.
3311
3312       -w  Inhibit all warning messages.
3313
3314       -Werror
3315           Make all warnings into errors.
3316
3317       -Werror=
3318           Make the specified warning into an error.  The specifier for a
3319           warning is appended; for example -Werror=switch turns the warnings
3320           controlled by -Wswitch into errors.  This switch takes a negative
3321           form, to be used to negate -Werror for specific warnings; for
3322           example -Wno-error=switch makes -Wswitch warnings not be errors,
3323           even when -Werror is in effect.
3324
3325           The warning message for each controllable warning includes the
3326           option that controls the warning.  That option can then be used
3327           with -Werror= and -Wno-error= as described above.  (Printing of the
3328           option in the warning message can be disabled using the
3329           -fno-diagnostics-show-option flag.)
3330
3331           Note that specifying -Werror=foo automatically implies -Wfoo.
3332           However, -Wno-error=foo does not imply anything.
3333
3334       -Wfatal-errors
3335           This option causes the compiler to abort compilation on the first
3336           error occurred rather than trying to keep going and printing
3337           further error messages.
3338
3339       You can request many specific warnings with options beginning with -W,
3340       for example -Wimplicit to request warnings on implicit declarations.
3341       Each of these specific warning options also has a negative form
3342       beginning -Wno- to turn off warnings; for example, -Wno-implicit.  This
3343       manual lists only one of the two forms, whichever is not the default.
3344       For further language-specific options also refer to C++ Dialect Options
3345       and Objective-C and Objective-C++ Dialect Options.
3346
3347       Some options, such as -Wall and -Wextra, turn on other options, such as
3348       -Wunused, which may turn on further options, such as -Wunused-value.
3349       The combined effect of positive and negative forms is that more
3350       specific options have priority over less specific ones, independently
3351       of their position in the command-line. For options of the same
3352       specificity, the last one takes effect. Options enabled or disabled via
3353       pragmas take effect as if they appeared at the end of the command-line.
3354
3355       When an unrecognized warning option is requested (e.g.,
3356       -Wunknown-warning), GCC emits a diagnostic stating that the option is
3357       not recognized.  However, if the -Wno- form is used, the behavior is
3358       slightly different: no diagnostic is produced for -Wno-unknown-warning
3359       unless other diagnostics are being produced.  This allows the use of
3360       new -Wno- options with old compilers, but if something goes wrong, the
3361       compiler warns that an unrecognized option is present.
3362
3363       -Wpedantic
3364       -pedantic
3365           Issue all the warnings demanded by strict ISO C and ISO C++; reject
3366           all programs that use forbidden extensions, and some other programs
3367           that do not follow ISO C and ISO C++.  For ISO C, follows the
3368           version of the ISO C standard specified by any -std option used.
3369
3370           Valid ISO C and ISO C++ programs should compile properly with or
3371           without this option (though a rare few require -ansi or a -std
3372           option specifying the required version of ISO C).  However, without
3373           this option, certain GNU extensions and traditional C and C++
3374           features are supported as well.  With this option, they are
3375           rejected.
3376
3377           -Wpedantic does not cause warning messages for use of the alternate
3378           keywords whose names begin and end with __.  Pedantic warnings are
3379           also disabled in the expression that follows "__extension__".
3380           However, only system header files should use these escape routes;
3381           application programs should avoid them.
3382
3383           Some users try to use -Wpedantic to check programs for strict ISO C
3384           conformance.  They soon find that it does not do quite what they
3385           want: it finds some non-ISO practices, but not all---only those for
3386           which ISO C requires a diagnostic, and some others for which
3387           diagnostics have been added.
3388
3389           A feature to report any failure to conform to ISO C might be useful
3390           in some instances, but would require considerable additional work
3391           and would be quite different from -Wpedantic.  We don't have plans
3392           to support such a feature in the near future.
3393
3394           Where the standard specified with -std represents a GNU extended
3395           dialect of C, such as gnu90 or gnu99, there is a corresponding base
3396           standard, the version of ISO C on which the GNU extended dialect is
3397           based.  Warnings from -Wpedantic are given where they are required
3398           by the base standard.  (It does not make sense for such warnings to
3399           be given only for features not in the specified GNU C dialect,
3400           since by definition the GNU dialects of C include all features the
3401           compiler supports with the given option, and there would be nothing
3402           to warn about.)
3403
3404       -pedantic-errors
3405           Give an error whenever the base standard (see -Wpedantic) requires
3406           a diagnostic, in some cases where there is undefined behavior at
3407           compile-time and in some other cases that do not prevent
3408           compilation of programs that are valid according to the standard.
3409           This is not equivalent to -Werror=pedantic, since there are errors
3410           enabled by this option and not enabled by the latter and vice
3411           versa.
3412
3413       -Wall
3414           This enables all the warnings about constructions that some users
3415           consider questionable, and that are easy to avoid (or modify to
3416           prevent the warning), even in conjunction with macros.  This also
3417           enables some language-specific warnings described in C++ Dialect
3418           Options and Objective-C and Objective-C++ Dialect Options.
3419
3420           -Wall turns on the following warning flags:
3421
3422           -Waddress -Warray-bounds=1 (only with -O2) -Wbool-compare
3423           -Wbool-operation -Wc++11-compat  -Wc++14-compat -Wcatch-value (C++
3424           and Objective-C++ only) -Wchar-subscripts -Wcomment
3425           -Wduplicate-decl-specifier (C and Objective-C only) -Wenum-compare
3426           (in C/ObjC; this is on by default in C++) -Wformat
3427           -Wint-in-bool-context -Wimplicit (C and Objective-C only)
3428           -Wimplicit-int (C and Objective-C only)
3429           -Wimplicit-function-declaration (C and Objective-C only)
3430           -Winit-self (only for C++) -Wlogical-not-parentheses -Wmain (only
3431           for C/ObjC and unless -ffreestanding) -Wmaybe-uninitialized
3432           -Wmemset-elt-size -Wmemset-transposed-args -Wmisleading-indentation
3433           (only for C/C++) -Wmissing-attributes -Wmissing-braces (only for
3434           C/ObjC) -Wmultistatement-macros -Wnarrowing (only for C++)
3435           -Wnonnull -Wnonnull-compare -Wopenmp-simd -Wparentheses
3436           -Wpessimizing-move (only for C++) -Wpointer-sign -Wreorder
3437           -Wrestrict -Wreturn-type -Wsequence-point -Wsign-compare (only in
3438           C++) -Wsizeof-pointer-div -Wsizeof-pointer-memaccess
3439           -Wstrict-aliasing -Wstrict-overflow=1 -Wswitch
3440           -Wtautological-compare -Wtrigraphs -Wuninitialized
3441           -Wunknown-pragmas -Wunused-function -Wunused-label -Wunused-value
3442           -Wunused-variable -Wvolatile-register-var
3443
3444           Note that some warning flags are not implied by -Wall.  Some of
3445           them warn about constructions that users generally do not consider
3446           questionable, but which occasionally you might wish to check for;
3447           others warn about constructions that are necessary or hard to avoid
3448           in some cases, and there is no simple way to modify the code to
3449           suppress the warning. Some of them are enabled by -Wextra but many
3450           of them must be enabled individually.
3451
3452       -Wextra
3453           This enables some extra warning flags that are not enabled by
3454           -Wall. (This option used to be called -W.  The older name is still
3455           supported, but the newer name is more descriptive.)
3456
3457           -Wclobbered -Wcast-function-type -Wdeprecated-copy (C++ only)
3458           -Wempty-body -Wignored-qualifiers -Wimplicit-fallthrough=3
3459           -Wmissing-field-initializers -Wmissing-parameter-type (C only)
3460           -Wold-style-declaration (C only) -Woverride-init -Wsign-compare (C
3461           only) -Wredundant-move (only for C++) -Wtype-limits -Wuninitialized
3462           -Wshift-negative-value (in C++03 and in C99 and newer)
3463           -Wunused-parameter (only with -Wunused or -Wall)
3464           -Wunused-but-set-parameter (only with -Wunused or -Wall)
3465
3466           The option -Wextra also prints warning messages for the following
3467           cases:
3468
3469           *   A pointer is compared against integer zero with "<", "<=", ">",
3470               or ">=".
3471
3472           *   (C++ only) An enumerator and a non-enumerator both appear in a
3473               conditional expression.
3474
3475           *   (C++ only) Ambiguous virtual bases.
3476
3477           *   (C++ only) Subscripting an array that has been declared
3478               "register".
3479
3480           *   (C++ only) Taking the address of a variable that has been
3481               declared "register".
3482
3483           *   (C++ only) A base class is not initialized in the copy
3484               constructor of a derived class.
3485
3486       -Wchar-subscripts
3487           Warn if an array subscript has type "char".  This is a common cause
3488           of error, as programmers often forget that this type is signed on
3489           some machines.  This warning is enabled by -Wall.
3490
3491       -Wno-coverage-mismatch
3492           Warn if feedback profiles do not match when using the -fprofile-use
3493           option.  If a source file is changed between compiling with
3494           -fprofile-generate and with -fprofile-use, the files with the
3495           profile feedback can fail to match the source file and GCC cannot
3496           use the profile feedback information.  By default, this warning is
3497           enabled and is treated as an error.  -Wno-coverage-mismatch can be
3498           used to disable the warning or -Wno-error=coverage-mismatch can be
3499           used to disable the error.  Disabling the error for this warning
3500           can result in poorly optimized code and is useful only in the case
3501           of very minor changes such as bug fixes to an existing code-base.
3502           Completely disabling the warning is not recommended.
3503
3504       -Wno-cpp
3505           (C, Objective-C, C++, Objective-C++ and Fortran only)
3506
3507           Suppress warning messages emitted by "#warning" directives.
3508
3509       -Wdouble-promotion (C, C++, Objective-C and Objective-C++ only)
3510           Give a warning when a value of type "float" is implicitly promoted
3511           to "double".  CPUs with a 32-bit "single-precision" floating-point
3512           unit implement "float" in hardware, but emulate "double" in
3513           software.  On such a machine, doing computations using "double"
3514           values is much more expensive because of the overhead required for
3515           software emulation.
3516
3517           It is easy to accidentally do computations with "double" because
3518           floating-point literals are implicitly of type "double".  For
3519           example, in:
3520
3521                   float area(float radius)
3522                   {
3523                      return 3.14159 * radius * radius;
3524                   }
3525
3526           the compiler performs the entire computation with "double" because
3527           the floating-point literal is a "double".
3528
3529       -Wduplicate-decl-specifier (C and Objective-C only)
3530           Warn if a declaration has duplicate "const", "volatile", "restrict"
3531           or "_Atomic" specifier.  This warning is enabled by -Wall.
3532
3533       -Wformat
3534       -Wformat=n
3535           Check calls to "printf" and "scanf", etc., to make sure that the
3536           arguments supplied have types appropriate to the format string
3537           specified, and that the conversions specified in the format string
3538           make sense.  This includes standard functions, and others specified
3539           by format attributes, in the "printf", "scanf", "strftime" and
3540           "strfmon" (an X/Open extension, not in the C standard) families (or
3541           other target-specific families).  Which functions are checked
3542           without format attributes having been specified depends on the
3543           standard version selected, and such checks of functions without the
3544           attribute specified are disabled by -ffreestanding or -fno-builtin.
3545
3546           The formats are checked against the format features supported by
3547           GNU libc version 2.2.  These include all ISO C90 and C99 features,
3548           as well as features from the Single Unix Specification and some BSD
3549           and GNU extensions.  Other library implementations may not support
3550           all these features; GCC does not support warning about features
3551           that go beyond a particular library's limitations.  However, if
3552           -Wpedantic is used with -Wformat, warnings are given about format
3553           features not in the selected standard version (but not for
3554           "strfmon" formats, since those are not in any version of the C
3555           standard).
3556
3557           -Wformat=1
3558           -Wformat
3559               Option -Wformat is equivalent to -Wformat=1, and -Wno-format is
3560               equivalent to -Wformat=0.  Since -Wformat also checks for null
3561               format arguments for several functions, -Wformat also implies
3562               -Wnonnull.  Some aspects of this level of format checking can
3563               be disabled by the options: -Wno-format-contains-nul,
3564               -Wno-format-extra-args, and -Wno-format-zero-length.  -Wformat
3565               is enabled by -Wall.
3566
3567           -Wno-format-contains-nul
3568               If -Wformat is specified, do not warn about format strings that
3569               contain NUL bytes.
3570
3571           -Wno-format-extra-args
3572               If -Wformat is specified, do not warn about excess arguments to
3573               a "printf" or "scanf" format function.  The C standard
3574               specifies that such arguments are ignored.
3575
3576               Where the unused arguments lie between used arguments that are
3577               specified with $ operand number specifications, normally
3578               warnings are still given, since the implementation could not
3579               know what type to pass to "va_arg" to skip the unused
3580               arguments.  However, in the case of "scanf" formats, this
3581               option suppresses the warning if the unused arguments are all
3582               pointers, since the Single Unix Specification says that such
3583               unused arguments are allowed.
3584
3585           -Wformat-overflow
3586           -Wformat-overflow=level
3587               Warn about calls to formatted input/output functions such as
3588               "sprintf" and "vsprintf" that might overflow the destination
3589               buffer.  When the exact number of bytes written by a format
3590               directive cannot be determined at compile-time it is estimated
3591               based on heuristics that depend on the level argument and on
3592               optimization.  While enabling optimization will in most cases
3593               improve the accuracy of the warning, it may also result in
3594               false positives.
3595
3596               -Wformat-overflow
3597               -Wformat-overflow=1
3598                   Level 1 of -Wformat-overflow enabled by -Wformat employs a
3599                   conservative approach that warns only about calls that most
3600                   likely overflow the buffer.  At this level, numeric
3601                   arguments to format directives with unknown values are
3602                   assumed to have the value of one, and strings of unknown
3603                   length to be empty.  Numeric arguments that are known to be
3604                   bounded to a subrange of their type, or string arguments
3605                   whose output is bounded either by their directive's
3606                   precision or by a finite set of string literals, are
3607                   assumed to take on the value within the range that results
3608                   in the most bytes on output.  For example, the call to
3609                   "sprintf" below is diagnosed because even with both a and b
3610                   equal to zero, the terminating NUL character ('\0')
3611                   appended by the function to the destination buffer will be
3612                   written past its end.  Increasing the size of the buffer by
3613                   a single byte is sufficient to avoid the warning, though it
3614                   may not be sufficient to avoid the overflow.
3615
3616                           void f (int a, int b)
3617                           {
3618                             char buf [13];
3619                             sprintf (buf, "a = %i, b = %i\n", a, b);
3620                           }
3621
3622               -Wformat-overflow=2
3623                   Level 2 warns also about calls that might overflow the
3624                   destination buffer given an argument of sufficient length
3625                   or magnitude.  At level 2, unknown numeric arguments are
3626                   assumed to have the minimum representable value for signed
3627                   types with a precision greater than 1, and the maximum
3628                   representable value otherwise.  Unknown string arguments
3629                   whose length cannot be assumed to be bounded either by the
3630                   directive's precision, or by a finite set of string
3631                   literals they may evaluate to, or the character array they
3632                   may point to, are assumed to be 1 character long.
3633
3634                   At level 2, the call in the example above is again
3635                   diagnosed, but this time because with a equal to a 32-bit
3636                   "INT_MIN" the first %i directive will write some of its
3637                   digits beyond the end of the destination buffer.  To make
3638                   the call safe regardless of the values of the two
3639                   variables, the size of the destination buffer must be
3640                   increased to at least 34 bytes.  GCC includes the minimum
3641                   size of the buffer in an informational note following the
3642                   warning.
3643
3644                   An alternative to increasing the size of the destination
3645                   buffer is to constrain the range of formatted values.  The
3646                   maximum length of string arguments can be bounded by
3647                   specifying the precision in the format directive.  When
3648                   numeric arguments of format directives can be assumed to be
3649                   bounded by less than the precision of their type, choosing
3650                   an appropriate length modifier to the format specifier will
3651                   reduce the required buffer size.  For example, if a and b
3652                   in the example above can be assumed to be within the
3653                   precision of the "short int" type then using either the %hi
3654                   format directive or casting the argument to "short" reduces
3655                   the maximum required size of the buffer to 24 bytes.
3656
3657                           void f (int a, int b)
3658                           {
3659                             char buf [23];
3660                             sprintf (buf, "a = %hi, b = %i\n", a, (short)b);
3661                           }
3662
3663           -Wno-format-zero-length
3664               If -Wformat is specified, do not warn about zero-length
3665               formats.  The C standard specifies that zero-length formats are
3666               allowed.
3667
3668           -Wformat=2
3669               Enable -Wformat plus additional format checks.  Currently
3670               equivalent to -Wformat -Wformat-nonliteral -Wformat-security
3671               -Wformat-y2k.
3672
3673           -Wformat-nonliteral
3674               If -Wformat is specified, also warn if the format string is not
3675               a string literal and so cannot be checked, unless the format
3676               function takes its format arguments as a "va_list".
3677
3678           -Wformat-security
3679               If -Wformat is specified, also warn about uses of format
3680               functions that represent possible security problems.  At
3681               present, this warns about calls to "printf" and "scanf"
3682               functions where the format string is not a string literal and
3683               there are no format arguments, as in "printf (foo);".  This may
3684               be a security hole if the format string came from untrusted
3685               input and contains %n.  (This is currently a subset of what
3686               -Wformat-nonliteral warns about, but in future warnings may be
3687               added to -Wformat-security that are not included in
3688               -Wformat-nonliteral.)
3689
3690           -Wformat-signedness
3691               If -Wformat is specified, also warn if the format string
3692               requires an unsigned argument and the argument is signed and
3693               vice versa.
3694
3695           -Wformat-truncation
3696           -Wformat-truncation=level
3697               Warn about calls to formatted input/output functions such as
3698               "snprintf" and "vsnprintf" that might result in output
3699               truncation.  When the exact number of bytes written by a format
3700               directive cannot be determined at compile-time it is estimated
3701               based on heuristics that depend on the level argument and on
3702               optimization.  While enabling optimization will in most cases
3703               improve the accuracy of the warning, it may also result in
3704               false positives.  Except as noted otherwise, the option uses
3705               the same logic -Wformat-overflow.
3706
3707               -Wformat-truncation
3708               -Wformat-truncation=1
3709                   Level 1 of -Wformat-truncation enabled by -Wformat employs
3710                   a conservative approach that warns only about calls to
3711                   bounded functions whose return value is unused and that
3712                   will most likely result in output truncation.
3713
3714               -Wformat-truncation=2
3715                   Level 2 warns also about calls to bounded functions whose
3716                   return value is used and that might result in truncation
3717                   given an argument of sufficient length or magnitude.
3718
3719           -Wformat-y2k
3720               If -Wformat is specified, also warn about "strftime" formats
3721               that may yield only a two-digit year.
3722
3723       -Wnonnull
3724           Warn about passing a null pointer for arguments marked as requiring
3725           a non-null value by the "nonnull" function attribute.
3726
3727           -Wnonnull is included in -Wall and -Wformat.  It can be disabled
3728           with the -Wno-nonnull option.
3729
3730       -Wnonnull-compare
3731           Warn when comparing an argument marked with the "nonnull" function
3732           attribute against null inside the function.
3733
3734           -Wnonnull-compare is included in -Wall.  It can be disabled with
3735           the -Wno-nonnull-compare option.
3736
3737       -Wnull-dereference
3738           Warn if the compiler detects paths that trigger erroneous or
3739           undefined behavior due to dereferencing a null pointer.  This
3740           option is only active when -fdelete-null-pointer-checks is active,
3741           which is enabled by optimizations in most targets.  The precision
3742           of the warnings depends on the optimization options used.
3743
3744       -Winit-self (C, C++, Objective-C and Objective-C++ only)
3745           Warn about uninitialized variables that are initialized with
3746           themselves.  Note this option can only be used with the
3747           -Wuninitialized option.
3748
3749           For example, GCC warns about "i" being uninitialized in the
3750           following snippet only when -Winit-self has been specified:
3751
3752                   int f()
3753                   {
3754                     int i = i;
3755                     return i;
3756                   }
3757
3758           This warning is enabled by -Wall in C++.
3759
3760       -Wimplicit-int (C and Objective-C only)
3761           Warn when a declaration does not specify a type.  This warning is
3762           enabled by -Wall.
3763
3764       -Wimplicit-function-declaration (C and Objective-C only)
3765           Give a warning whenever a function is used before being declared.
3766           In C99 mode (-std=c99 or -std=gnu99), this warning is enabled by
3767           default and it is made into an error by -pedantic-errors. This
3768           warning is also enabled by -Wall.
3769
3770       -Wimplicit (C and Objective-C only)
3771           Same as -Wimplicit-int and -Wimplicit-function-declaration.  This
3772           warning is enabled by -Wall.
3773
3774       -Wimplicit-fallthrough
3775           -Wimplicit-fallthrough is the same as -Wimplicit-fallthrough=3 and
3776           -Wno-implicit-fallthrough is the same as -Wimplicit-fallthrough=0.
3777
3778       -Wimplicit-fallthrough=n
3779           Warn when a switch case falls through.  For example:
3780
3781                   switch (cond)
3782                     {
3783                     case 1:
3784                       a = 1;
3785                       break;
3786                     case 2:
3787                       a = 2;
3788                     case 3:
3789                       a = 3;
3790                       break;
3791                     }
3792
3793           This warning does not warn when the last statement of a case cannot
3794           fall through, e.g. when there is a return statement or a call to
3795           function declared with the noreturn attribute.
3796           -Wimplicit-fallthrough= also takes into account control flow
3797           statements, such as ifs, and only warns when appropriate.  E.g.
3798
3799                   switch (cond)
3800                     {
3801                     case 1:
3802                       if (i > 3) {
3803                         bar (5);
3804                         break;
3805                       } else if (i < 1) {
3806                         bar (0);
3807                       } else
3808                         return;
3809                     default:
3810                       ...
3811                     }
3812
3813           Since there are occasions where a switch case fall through is
3814           desirable, GCC provides an attribute, "__attribute__
3815           ((fallthrough))", that is to be used along with a null statement to
3816           suppress this warning that would normally occur:
3817
3818                   switch (cond)
3819                     {
3820                     case 1:
3821                       bar (0);
3822                       __attribute__ ((fallthrough));
3823                     default:
3824                       ...
3825                     }
3826
3827           C++17 provides a standard way to suppress the
3828           -Wimplicit-fallthrough warning using "[[fallthrough]];" instead of
3829           the GNU attribute.  In C++11 or C++14 users can use
3830           "[[gnu::fallthrough]];", which is a GNU extension.  Instead of
3831           these attributes, it is also possible to add a fallthrough comment
3832           to silence the warning.  The whole body of the C or C++ style
3833           comment should match the given regular expressions listed below.
3834           The option argument n specifies what kind of comments are accepted:
3835
3836           *<-Wimplicit-fallthrough=0 disables the warning altogether.>
3837           *<-Wimplicit-fallthrough=1 matches ".*" regular>
3838               expression, any comment is used as fallthrough comment.
3839
3840           *<-Wimplicit-fallthrough=2 case insensitively matches>
3841               ".*falls?[ \t-]*thr(ough|u).*" regular expression.
3842
3843           *<-Wimplicit-fallthrough=3 case sensitively matches one of the>
3844               following regular expressions:
3845
3846               *<"-fallthrough">
3847               *<"@fallthrough@">
3848               *<"lint -fallthrough[ \t]*">
3849               *<"[ \t.!]*(ELSE,? |INTENTIONAL(LY)? )?FALL(S |
3850               |-)?THR(OUGH|U)[ \t.!]*(-[^\n\r]*)?">
3851               *<"[ \t.!]*(Else,? |Intentional(ly)? )?Fall((s |
3852               |-)[Tt]|t)hr(ough|u)[ \t.!]*(-[^\n\r]*)?">
3853               *<"[ \t.!]*([Ee]lse,? |[Ii]ntentional(ly)? )?fall(s |
3854               |-)?thr(ough|u)[ \t.!]*(-[^\n\r]*)?">
3855           *<-Wimplicit-fallthrough=4 case sensitively matches one of the>
3856               following regular expressions:
3857
3858               *<"-fallthrough">
3859               *<"@fallthrough@">
3860               *<"lint -fallthrough[ \t]*">
3861               *<"[ \t]*FALLTHR(OUGH|U)[ \t]*">
3862           *<-Wimplicit-fallthrough=5 doesn't recognize any comments as>
3863               fallthrough comments, only attributes disable the warning.
3864
3865           The comment needs to be followed after optional whitespace and
3866           other comments by "case" or "default" keywords or by a user label
3867           that precedes some "case" or "default" label.
3868
3869                   switch (cond)
3870                     {
3871                     case 1:
3872                       bar (0);
3873                       /* FALLTHRU */
3874                     default:
3875                       ...
3876                     }
3877
3878           The -Wimplicit-fallthrough=3 warning is enabled by -Wextra.
3879
3880       -Wif-not-aligned (C, C++, Objective-C and Objective-C++ only)
3881           Control if warning triggered by the "warn_if_not_aligned" attribute
3882           should be issued.  This is enabled by default.  Use
3883           -Wno-if-not-aligned to disable it.
3884
3885       -Wignored-qualifiers (C and C++ only)
3886           Warn if the return type of a function has a type qualifier such as
3887           "const".  For ISO C such a type qualifier has no effect, since the
3888           value returned by a function is not an lvalue.  For C++, the
3889           warning is only emitted for scalar types or "void".  ISO C
3890           prohibits qualified "void" return types on function definitions, so
3891           such return types always receive a warning even without this
3892           option.
3893
3894           This warning is also enabled by -Wextra.
3895
3896       -Wignored-attributes (C and C++ only)
3897           Warn when an attribute is ignored.  This is different from the
3898           -Wattributes option in that it warns whenever the compiler decides
3899           to drop an attribute, not that the attribute is either unknown,
3900           used in a wrong place, etc.  This warning is enabled by default.
3901
3902       -Wmain
3903           Warn if the type of "main" is suspicious.  "main" should be a
3904           function with external linkage, returning int, taking either zero
3905           arguments, two, or three arguments of appropriate types.  This
3906           warning is enabled by default in C++ and is enabled by either -Wall
3907           or -Wpedantic.
3908
3909       -Wmisleading-indentation (C and C++ only)
3910           Warn when the indentation of the code does not reflect the block
3911           structure.  Specifically, a warning is issued for "if", "else",
3912           "while", and "for" clauses with a guarded statement that does not
3913           use braces, followed by an unguarded statement with the same
3914           indentation.
3915
3916           In the following example, the call to "bar" is misleadingly
3917           indented as if it were guarded by the "if" conditional.
3918
3919                     if (some_condition ())
3920                       foo ();
3921                       bar ();  /* Gotcha: this is not guarded by the "if".  */
3922
3923           In the case of mixed tabs and spaces, the warning uses the
3924           -ftabstop= option to determine if the statements line up
3925           (defaulting to 8).
3926
3927           The warning is not issued for code involving multiline preprocessor
3928           logic such as the following example.
3929
3930                     if (flagA)
3931                       foo (0);
3932                   #if SOME_CONDITION_THAT_DOES_NOT_HOLD
3933                     if (flagB)
3934                   #endif
3935                       foo (1);
3936
3937           The warning is not issued after a "#line" directive, since this
3938           typically indicates autogenerated code, and no assumptions can be
3939           made about the layout of the file that the directive references.
3940
3941           This warning is enabled by -Wall in C and C++.
3942
3943       -Wmissing-attributes
3944           Warn when a declaration of a function is missing one or more
3945           attributes that a related function is declared with and whose
3946           absence may adversely affect the correctness or efficiency of
3947           generated code.  For example, the warning is issued for
3948           declarations of aliases that use attributes to specify less
3949           restrictive requirements than those of their targets.  This
3950           typically represents a potential optimization opportunity.  By
3951           contrast, the -Wattribute-alias=2 option controls warnings issued
3952           when the alias is more restrictive than the target, which could
3953           lead to incorrect code generation.  Attributes considered include
3954           "alloc_align", "alloc_size", "cold", "const", "hot", "leaf",
3955           "malloc", "nonnull", "noreturn", "nothrow", "pure",
3956           "returns_nonnull", and "returns_twice".
3957
3958           In C++, the warning is issued when an explicit specialization of a
3959           primary template declared with attribute "alloc_align",
3960           "alloc_size", "assume_aligned", "format", "format_arg", "malloc",
3961           or "nonnull" is declared without it.  Attributes "deprecated",
3962           "error", and "warning" suppress the warning..
3963
3964           You can use the "copy" attribute to apply the same set of
3965           attributes to a declaration as that on another declaration without
3966           explicitly enumerating the attributes. This attribute can be
3967           applied to declarations of functions, variables, or types.
3968
3969           -Wmissing-attributes is enabled by -Wall.
3970
3971           For example, since the declaration of the primary function template
3972           below makes use of both attribute "malloc" and "alloc_size" the
3973           declaration of the explicit specialization of the template is
3974           diagnosed because it is missing one of the attributes.
3975
3976                   template <class T>
3977                   T* __attribute__ ((malloc, alloc_size (1)))
3978                   allocate (size_t);
3979
3980                   template <>
3981                   void* __attribute__ ((malloc))   // missing alloc_size
3982                   allocate<void> (size_t);
3983
3984       -Wmissing-braces
3985           Warn if an aggregate or union initializer is not fully bracketed.
3986           In the following example, the initializer for "a" is not fully
3987           bracketed, but that for "b" is fully bracketed.  This warning is
3988           enabled by -Wall in C.
3989
3990                   int a[2][2] = { 0, 1, 2, 3 };
3991                   int b[2][2] = { { 0, 1 }, { 2, 3 } };
3992
3993           This warning is enabled by -Wall.
3994
3995       -Wmissing-include-dirs (C, C++, Objective-C and Objective-C++ only)
3996           Warn if a user-supplied include directory does not exist.
3997
3998       -Wmissing-profile
3999           Warn if feedback profiles are missing when using the -fprofile-use
4000           option.  This option diagnoses those cases where a new function or
4001           a new file is added to the user code between compiling with
4002           -fprofile-generate and with -fprofile-use, without regenerating the
4003           profiles.  In these cases, the profile feedback data files do not
4004           contain any profile feedback information for the newly added
4005           function or file respectively.  Also, in the case when profile
4006           count data (.gcda) files are removed, GCC cannot use any profile
4007           feedback information.  In all these cases, warnings are issued to
4008           inform the user that a profile generation step is due.
4009           -Wno-missing-profile can be used to disable the warning.  Ignoring
4010           the warning can result in poorly optimized code.  Completely
4011           disabling the warning is not recommended and should be done only
4012           when non-existent profile data is justified.
4013
4014       -Wmultistatement-macros
4015           Warn about unsafe multiple statement macros that appear to be
4016           guarded by a clause such as "if", "else", "for", "switch", or
4017           "while", in which only the first statement is actually guarded
4018           after the macro is expanded.
4019
4020           For example:
4021
4022                   #define DOIT x++; y++
4023                   if (c)
4024                     DOIT;
4025
4026           will increment "y" unconditionally, not just when "c" holds.  The
4027           can usually be fixed by wrapping the macro in a do-while loop:
4028
4029                   #define DOIT do { x++; y++; } while (0)
4030                   if (c)
4031                     DOIT;
4032
4033           This warning is enabled by -Wall in C and C++.
4034
4035       -Wparentheses
4036           Warn if parentheses are omitted in certain contexts, such as when
4037           there is an assignment in a context where a truth value is
4038           expected, or when operators are nested whose precedence people
4039           often get confused about.
4040
4041           Also warn if a comparison like "x<=y<=z" appears; this is
4042           equivalent to "(x<=y ? 1 : 0) <= z", which is a different
4043           interpretation from that of ordinary mathematical notation.
4044
4045           Also warn for dangerous uses of the GNU extension to "?:" with
4046           omitted middle operand. When the condition in the "?": operator is
4047           a boolean expression, the omitted value is always 1.  Often
4048           programmers expect it to be a value computed inside the conditional
4049           expression instead.
4050
4051           For C++ this also warns for some cases of unnecessary parentheses
4052           in declarations, which can indicate an attempt at a function call
4053           instead of a declaration:
4054
4055                   {
4056                     // Declares a local variable called mymutex.
4057                     std::unique_lock<std::mutex> (mymutex);
4058                     // User meant std::unique_lock<std::mutex> lock (mymutex);
4059                   }
4060
4061           This warning is enabled by -Wall.
4062
4063       -Wsequence-point
4064           Warn about code that may have undefined semantics because of
4065           violations of sequence point rules in the C and C++ standards.
4066
4067           The C and C++ standards define the order in which expressions in a
4068           C/C++ program are evaluated in terms of sequence points, which
4069           represent a partial ordering between the execution of parts of the
4070           program: those executed before the sequence point, and those
4071           executed after it.  These occur after the evaluation of a full
4072           expression (one which is not part of a larger expression), after
4073           the evaluation of the first operand of a "&&", "||", "? :" or ","
4074           (comma) operator, before a function is called (but after the
4075           evaluation of its arguments and the expression denoting the called
4076           function), and in certain other places.  Other than as expressed by
4077           the sequence point rules, the order of evaluation of subexpressions
4078           of an expression is not specified.  All these rules describe only a
4079           partial order rather than a total order, since, for example, if two
4080           functions are called within one expression with no sequence point
4081           between them, the order in which the functions are called is not
4082           specified.  However, the standards committee have ruled that
4083           function calls do not overlap.
4084
4085           It is not specified when between sequence points modifications to
4086           the values of objects take effect.  Programs whose behavior depends
4087           on this have undefined behavior; the C and C++ standards specify
4088           that "Between the previous and next sequence point an object shall
4089           have its stored value modified at most once by the evaluation of an
4090           expression.  Furthermore, the prior value shall be read only to
4091           determine the value to be stored.".  If a program breaks these
4092           rules, the results on any particular implementation are entirely
4093           unpredictable.
4094
4095           Examples of code with undefined behavior are "a = a++;", "a[n] =
4096           b[n++]" and "a[i++] = i;".  Some more complicated cases are not
4097           diagnosed by this option, and it may give an occasional false
4098           positive result, but in general it has been found fairly effective
4099           at detecting this sort of problem in programs.
4100
4101           The C++17 standard will define the order of evaluation of operands
4102           in more cases: in particular it requires that the right-hand side
4103           of an assignment be evaluated before the left-hand side, so the
4104           above examples are no longer undefined.  But this warning will
4105           still warn about them, to help people avoid writing code that is
4106           undefined in C and earlier revisions of C++.
4107
4108           The standard is worded confusingly, therefore there is some debate
4109           over the precise meaning of the sequence point rules in subtle
4110           cases.  Links to discussions of the problem, including proposed
4111           formal definitions, may be found on the GCC readings page, at
4112           <http://gcc.gnu.org/readings.html>.
4113
4114           This warning is enabled by -Wall for C and C++.
4115
4116       -Wno-return-local-addr
4117           Do not warn about returning a pointer (or in C++, a reference) to a
4118           variable that goes out of scope after the function returns.
4119
4120       -Wreturn-type
4121           Warn whenever a function is defined with a return type that
4122           defaults to "int".  Also warn about any "return" statement with no
4123           return value in a function whose return type is not "void" (falling
4124           off the end of the function body is considered returning without a
4125           value).
4126
4127           For C only, warn about a "return" statement with an expression in a
4128           function whose return type is "void", unless the expression type is
4129           also "void".  As a GNU extension, the latter case is accepted
4130           without a warning unless -Wpedantic is used.  Attempting to use the
4131           return value of a non-"void" function other than "main" that flows
4132           off the end by reaching the closing curly brace that terminates the
4133           function is undefined.
4134
4135           Unlike in C, in C++, flowing off the end of a non-"void" function
4136           other than "main" results in undefined behavior even when the value
4137           of the function is not used.
4138
4139           This warning is enabled by default in C++ and by -Wall otherwise.
4140
4141       -Wshift-count-negative
4142           Warn if shift count is negative. This warning is enabled by
4143           default.
4144
4145       -Wshift-count-overflow
4146           Warn if shift count >= width of type. This warning is enabled by
4147           default.
4148
4149       -Wshift-negative-value
4150           Warn if left shifting a negative value.  This warning is enabled by
4151           -Wextra in C99 and C++11 modes (and newer).
4152
4153       -Wshift-overflow
4154       -Wshift-overflow=n
4155           Warn about left shift overflows.  This warning is enabled by
4156           default in C99 and C++11 modes (and newer).
4157
4158           -Wshift-overflow=1
4159               This is the warning level of -Wshift-overflow and is enabled by
4160               default in C99 and C++11 modes (and newer).  This warning level
4161               does not warn about left-shifting 1 into the sign bit.
4162               (However, in C, such an overflow is still rejected in contexts
4163               where an integer constant expression is required.)  No warning
4164               is emitted in C++2A mode (and newer), as signed left shifts
4165               always wrap.
4166
4167           -Wshift-overflow=2
4168               This warning level also warns about left-shifting 1 into the
4169               sign bit, unless C++14 mode (or newer) is active.
4170
4171       -Wswitch
4172           Warn whenever a "switch" statement has an index of enumerated type
4173           and lacks a "case" for one or more of the named codes of that
4174           enumeration.  (The presence of a "default" label prevents this
4175           warning.)  "case" labels outside the enumeration range also provoke
4176           warnings when this option is used (even if there is a "default"
4177           label).  This warning is enabled by -Wall.
4178
4179       -Wswitch-default
4180           Warn whenever a "switch" statement does not have a "default" case.
4181
4182       -Wswitch-enum
4183           Warn whenever a "switch" statement has an index of enumerated type
4184           and lacks a "case" for one or more of the named codes of that
4185           enumeration.  "case" labels outside the enumeration range also
4186           provoke warnings when this option is used.  The only difference
4187           between -Wswitch and this option is that this option gives a
4188           warning about an omitted enumeration code even if there is a
4189           "default" label.
4190
4191       -Wswitch-bool
4192           Warn whenever a "switch" statement has an index of boolean type and
4193           the case values are outside the range of a boolean type.  It is
4194           possible to suppress this warning by casting the controlling
4195           expression to a type other than "bool".  For example:
4196
4197                   switch ((int) (a == 4))
4198                     {
4199                     ...
4200                     }
4201
4202           This warning is enabled by default for C and C++ programs.
4203
4204       -Wswitch-unreachable
4205           Warn whenever a "switch" statement contains statements between the
4206           controlling expression and the first case label, which will never
4207           be executed.  For example:
4208
4209                   switch (cond)
4210                     {
4211                      i = 15;
4212                     ...
4213                      case 5:
4214                     ...
4215                     }
4216
4217           -Wswitch-unreachable does not warn if the statement between the
4218           controlling expression and the first case label is just a
4219           declaration:
4220
4221                   switch (cond)
4222                     {
4223                      int i;
4224                     ...
4225                      case 5:
4226                      i = 5;
4227                     ...
4228                     }
4229
4230           This warning is enabled by default for C and C++ programs.
4231
4232       -Wsync-nand (C and C++ only)
4233           Warn when "__sync_fetch_and_nand" and "__sync_nand_and_fetch"
4234           built-in functions are used.  These functions changed semantics in
4235           GCC 4.4.
4236
4237       -Wunused-but-set-parameter
4238           Warn whenever a function parameter is assigned to, but otherwise
4239           unused (aside from its declaration).
4240
4241           To suppress this warning use the "unused" attribute.
4242
4243           This warning is also enabled by -Wunused together with -Wextra.
4244
4245       -Wunused-but-set-variable
4246           Warn whenever a local variable is assigned to, but otherwise unused
4247           (aside from its declaration).  This warning is enabled by -Wall.
4248
4249           To suppress this warning use the "unused" attribute.
4250
4251           This warning is also enabled by -Wunused, which is enabled by
4252           -Wall.
4253
4254       -Wunused-function
4255           Warn whenever a static function is declared but not defined or a
4256           non-inline static function is unused.  This warning is enabled by
4257           -Wall.
4258
4259       -Wunused-label
4260           Warn whenever a label is declared but not used.  This warning is
4261           enabled by -Wall.
4262
4263           To suppress this warning use the "unused" attribute.
4264
4265       -Wunused-local-typedefs (C, Objective-C, C++ and Objective-C++ only)
4266           Warn when a typedef locally defined in a function is not used.
4267           This warning is enabled by -Wall.
4268
4269       -Wunused-parameter
4270           Warn whenever a function parameter is unused aside from its
4271           declaration.
4272
4273           To suppress this warning use the "unused" attribute.
4274
4275       -Wno-unused-result
4276           Do not warn if a caller of a function marked with attribute
4277           "warn_unused_result" does not use its return value. The default is
4278           -Wunused-result.
4279
4280       -Wunused-variable
4281           Warn whenever a local or static variable is unused aside from its
4282           declaration. This option implies -Wunused-const-variable=1 for C,
4283           but not for C++. This warning is enabled by -Wall.
4284
4285           To suppress this warning use the "unused" attribute.
4286
4287       -Wunused-const-variable
4288       -Wunused-const-variable=n
4289           Warn whenever a constant static variable is unused aside from its
4290           declaration.  -Wunused-const-variable=1 is enabled by
4291           -Wunused-variable for C, but not for C++. In C this declares
4292           variable storage, but in C++ this is not an error since const
4293           variables take the place of "#define"s.
4294
4295           To suppress this warning use the "unused" attribute.
4296
4297           -Wunused-const-variable=1
4298               This is the warning level that is enabled by -Wunused-variable
4299               for C.  It warns only about unused static const variables
4300               defined in the main compilation unit, but not about static
4301               const variables declared in any header included.
4302
4303           -Wunused-const-variable=2
4304               This warning level also warns for unused constant static
4305               variables in headers (excluding system headers).  This is the
4306               warning level of -Wunused-const-variable and must be explicitly
4307               requested since in C++ this isn't an error and in C it might be
4308               harder to clean up all headers included.
4309
4310       -Wunused-value
4311           Warn whenever a statement computes a result that is explicitly not
4312           used. To suppress this warning cast the unused expression to
4313           "void". This includes an expression-statement or the left-hand side
4314           of a comma expression that contains no side effects. For example,
4315           an expression such as "x[i,j]" causes a warning, while
4316           "x[(void)i,j]" does not.
4317
4318           This warning is enabled by -Wall.
4319
4320       -Wunused
4321           All the above -Wunused options combined.
4322
4323           In order to get a warning about an unused function parameter, you
4324           must either specify -Wextra -Wunused (note that -Wall implies
4325           -Wunused), or separately specify -Wunused-parameter.
4326
4327       -Wuninitialized
4328           Warn if an automatic variable is used without first being
4329           initialized or if a variable may be clobbered by a "setjmp" call.
4330           In C++, warn if a non-static reference or non-static "const" member
4331           appears in a class without constructors.
4332
4333           If you want to warn about code that uses the uninitialized value of
4334           the variable in its own initializer, use the -Winit-self option.
4335
4336           These warnings occur for individual uninitialized or clobbered
4337           elements of structure, union or array variables as well as for
4338           variables that are uninitialized or clobbered as a whole.  They do
4339           not occur for variables or elements declared "volatile".  Because
4340           these warnings depend on optimization, the exact variables or
4341           elements for which there are warnings depends on the precise
4342           optimization options and version of GCC used.
4343
4344           Note that there may be no warning about a variable that is used
4345           only to compute a value that itself is never used, because such
4346           computations may be deleted by data flow analysis before the
4347           warnings are printed.
4348
4349       -Winvalid-memory-model
4350           Warn for invocations of __atomic Builtins, __sync Builtins, and the
4351           C11 atomic generic functions with a memory consistency argument
4352           that is either invalid for the operation or outside the range of
4353           values of the "memory_order" enumeration.  For example, since the
4354           "__atomic_store" and "__atomic_store_n" built-ins are only defined
4355           for the relaxed, release, and sequentially consistent memory orders
4356           the following code is diagnosed:
4357
4358                   void store (int *i)
4359                   {
4360                     __atomic_store_n (i, 0, memory_order_consume);
4361                   }
4362
4363           -Winvalid-memory-model is enabled by default.
4364
4365       -Wmaybe-uninitialized
4366           For an automatic (i.e. local) variable, if there exists a path from
4367           the function entry to a use of the variable that is initialized,
4368           but there exist some other paths for which the variable is not
4369           initialized, the compiler emits a warning if it cannot prove the
4370           uninitialized paths are not executed at run time.
4371
4372           These warnings are only possible in optimizing compilation, because
4373           otherwise GCC does not keep track of the state of variables.
4374
4375           These warnings are made optional because GCC may not be able to
4376           determine when the code is correct in spite of appearing to have an
4377           error.  Here is one example of how this can happen:
4378
4379                   {
4380                     int x;
4381                     switch (y)
4382                       {
4383                       case 1: x = 1;
4384                         break;
4385                       case 2: x = 4;
4386                         break;
4387                       case 3: x = 5;
4388                       }
4389                     foo (x);
4390                   }
4391
4392           If the value of "y" is always 1, 2 or 3, then "x" is always
4393           initialized, but GCC doesn't know this. To suppress the warning,
4394           you need to provide a default case with assert(0) or similar code.
4395
4396           This option also warns when a non-volatile automatic variable might
4397           be changed by a call to "longjmp".  The compiler sees only the
4398           calls to "setjmp".  It cannot know where "longjmp" will be called;
4399           in fact, a signal handler could call it at any point in the code.
4400           As a result, you may get a warning even when there is in fact no
4401           problem because "longjmp" cannot in fact be called at the place
4402           that would cause a problem.
4403
4404           Some spurious warnings can be avoided if you declare all the
4405           functions you use that never return as "noreturn".
4406
4407           This warning is enabled by -Wall or -Wextra.
4408
4409       -Wunknown-pragmas
4410           Warn when a "#pragma" directive is encountered that is not
4411           understood by GCC.  If this command-line option is used, warnings
4412           are even issued for unknown pragmas in system header files.  This
4413           is not the case if the warnings are only enabled by the -Wall
4414           command-line option.
4415
4416       -Wno-pragmas
4417           Do not warn about misuses of pragmas, such as incorrect parameters,
4418           invalid syntax, or conflicts between pragmas.  See also
4419           -Wunknown-pragmas.
4420
4421       -Wno-prio-ctor-dtor
4422           Do not warn if a priority from 0 to 100 is used for constructor or
4423           destructor.  The use of constructor and destructor attributes allow
4424           you to assign a priority to the constructor/destructor to control
4425           its order of execution before "main" is called or after it returns.
4426           The priority values must be greater than 100 as the compiler
4427           reserves priority values between 0--100 for the implementation.
4428
4429       -Wstrict-aliasing
4430           This option is only active when -fstrict-aliasing is active.  It
4431           warns about code that might break the strict aliasing rules that
4432           the compiler is using for optimization.  The warning does not catch
4433           all cases, but does attempt to catch the more common pitfalls.  It
4434           is included in -Wall.  It is equivalent to -Wstrict-aliasing=3
4435
4436       -Wstrict-aliasing=n
4437           This option is only active when -fstrict-aliasing is active.  It
4438           warns about code that might break the strict aliasing rules that
4439           the compiler is using for optimization.  Higher levels correspond
4440           to higher accuracy (fewer false positives).  Higher levels also
4441           correspond to more effort, similar to the way -O works.
4442           -Wstrict-aliasing is equivalent to -Wstrict-aliasing=3.
4443
4444           Level 1: Most aggressive, quick, least accurate.  Possibly useful
4445           when higher levels do not warn but -fstrict-aliasing still breaks
4446           the code, as it has very few false negatives.  However, it has many
4447           false positives.  Warns for all pointer conversions between
4448           possibly incompatible types, even if never dereferenced.  Runs in
4449           the front end only.
4450
4451           Level 2: Aggressive, quick, not too precise.  May still have many
4452           false positives (not as many as level 1 though), and few false
4453           negatives (but possibly more than level 1).  Unlike level 1, it
4454           only warns when an address is taken.  Warns about incomplete types.
4455           Runs in the front end only.
4456
4457           Level 3 (default for -Wstrict-aliasing): Should have very few false
4458           positives and few false negatives.  Slightly slower than levels 1
4459           or 2 when optimization is enabled.  Takes care of the common
4460           pun+dereference pattern in the front end: "*(int*)&some_float".  If
4461           optimization is enabled, it also runs in the back end, where it
4462           deals with multiple statement cases using flow-sensitive points-to
4463           information.  Only warns when the converted pointer is
4464           dereferenced.  Does not warn about incomplete types.
4465
4466       -Wstrict-overflow
4467       -Wstrict-overflow=n
4468           This option is only active when signed overflow is undefined.  It
4469           warns about cases where the compiler optimizes based on the
4470           assumption that signed overflow does not occur.  Note that it does
4471           not warn about all cases where the code might overflow: it only
4472           warns about cases where the compiler implements some optimization.
4473           Thus this warning depends on the optimization level.
4474
4475           An optimization that assumes that signed overflow does not occur is
4476           perfectly safe if the values of the variables involved are such
4477           that overflow never does, in fact, occur.  Therefore this warning
4478           can easily give a false positive: a warning about code that is not
4479           actually a problem.  To help focus on important issues, several
4480           warning levels are defined.  No warnings are issued for the use of
4481           undefined signed overflow when estimating how many iterations a
4482           loop requires, in particular when determining whether a loop will
4483           be executed at all.
4484
4485           -Wstrict-overflow=1
4486               Warn about cases that are both questionable and easy to avoid.
4487               For example the compiler simplifies "x + 1 > x" to 1.  This
4488               level of -Wstrict-overflow is enabled by -Wall; higher levels
4489               are not, and must be explicitly requested.
4490
4491           -Wstrict-overflow=2
4492               Also warn about other cases where a comparison is simplified to
4493               a constant.  For example: "abs (x) >= 0".  This can only be
4494               simplified when signed integer overflow is undefined, because
4495               "abs (INT_MIN)" overflows to "INT_MIN", which is less than
4496               zero.  -Wstrict-overflow (with no level) is the same as
4497               -Wstrict-overflow=2.
4498
4499           -Wstrict-overflow=3
4500               Also warn about other cases where a comparison is simplified.
4501               For example: "x + 1 > 1" is simplified to "x > 0".
4502
4503           -Wstrict-overflow=4
4504               Also warn about other simplifications not covered by the above
4505               cases.  For example: "(x * 10) / 5" is simplified to "x * 2".
4506
4507           -Wstrict-overflow=5
4508               Also warn about cases where the compiler reduces the magnitude
4509               of a constant involved in a comparison.  For example: "x + 2 >
4510               y" is simplified to "x + 1 >= y".  This is reported only at the
4511               highest warning level because this simplification applies to
4512               many comparisons, so this warning level gives a very large
4513               number of false positives.
4514
4515       -Wstringop-overflow
4516       -Wstringop-overflow=type
4517           Warn for calls to string manipulation functions such as "memcpy"
4518           and "strcpy" that are determined to overflow the destination
4519           buffer.  The optional argument is one greater than the type of
4520           Object Size Checking to perform to determine the size of the
4521           destination.  The argument is meaningful only for functions that
4522           operate on character arrays but not for raw memory functions like
4523           "memcpy" which always make use of Object Size type-0.  The option
4524           also warns for calls that specify a size in excess of the largest
4525           possible object or at most "SIZE_MAX / 2" bytes.  The option
4526           produces the best results with optimization enabled but can detect
4527           a small subset of simple buffer overflows even without optimization
4528           in calls to the GCC built-in functions like "__builtin_memcpy" that
4529           correspond to the standard functions.  In any case, the option
4530           warns about just a subset of buffer overflows detected by the
4531           corresponding overflow checking built-ins.  For example, the option
4532           will issue a warning for the "strcpy" call below because it copies
4533           at least 5 characters (the string "blue" including the terminating
4534           NUL) into the buffer of size 4.
4535
4536                   enum Color { blue, purple, yellow };
4537                   const char* f (enum Color clr)
4538                   {
4539                     static char buf [4];
4540                     const char *str;
4541                     switch (clr)
4542                       {
4543                         case blue: str = "blue"; break;
4544                         case purple: str = "purple"; break;
4545                         case yellow: str = "yellow"; break;
4546                       }
4547
4548                     return strcpy (buf, str);   // warning here
4549                   }
4550
4551           Option -Wstringop-overflow=2 is enabled by default.
4552
4553           -Wstringop-overflow
4554           -Wstringop-overflow=1
4555               The -Wstringop-overflow=1 option uses type-zero Object Size
4556               Checking to determine the sizes of destination objects.  This
4557               is the default setting of the option.  At this setting the
4558               option will not warn for writes past the end of subobjects of
4559               larger objects accessed by pointers unless the size of the
4560               largest surrounding object is known.  When the destination may
4561               be one of several objects it is assumed to be the largest one
4562               of them.  On Linux systems, when optimization is enabled at
4563               this setting the option warns for the same code as when the
4564               "_FORTIFY_SOURCE" macro is defined to a non-zero value.
4565
4566           -Wstringop-overflow=2
4567               The -Wstringop-overflow=2 option uses type-one Object Size
4568               Checking to determine the sizes of destination objects.  At
4569               this setting the option will warn about overflows when writing
4570               to members of the largest complete objects whose exact size is
4571               known.  It will, however, not warn for excessive writes to the
4572               same members of unknown objects referenced by pointers since
4573               they may point to arrays containing unknown numbers of
4574               elements.
4575
4576           -Wstringop-overflow=3
4577               The -Wstringop-overflow=3 option uses type-two Object Size
4578               Checking to determine the sizes of destination objects.  At
4579               this setting the option warns about overflowing the smallest
4580               object or data member.  This is the most restrictive setting of
4581               the option that may result in warnings for safe code.
4582
4583           -Wstringop-overflow=4
4584               The -Wstringop-overflow=4 option uses type-three Object Size
4585               Checking to determine the sizes of destination objects.  At
4586               this setting the option will warn about overflowing any data
4587               members, and when the destination is one of several objects it
4588               uses the size of the largest of them to decide whether to issue
4589               a warning.  Similarly to -Wstringop-overflow=3 this setting of
4590               the option may result in warnings for benign code.
4591
4592       -Wstringop-truncation
4593           Warn for calls to bounded string manipulation functions such as
4594           "strncat", "strncpy", and "stpncpy" that may either truncate the
4595           copied string or leave the destination unchanged.
4596
4597           In the following example, the call to "strncat" specifies a bound
4598           that is less than the length of the source string.  As a result,
4599           the copy of the source will be truncated and so the call is
4600           diagnosed.  To avoid the warning use "bufsize - strlen (buf) - 1)"
4601           as the bound.
4602
4603                   void append (char *buf, size_t bufsize)
4604                   {
4605                     strncat (buf, ".txt", 3);
4606                   }
4607
4608           As another example, the following call to "strncpy" results in
4609           copying to "d" just the characters preceding the terminating NUL,
4610           without appending the NUL to the end.  Assuming the result of
4611           "strncpy" is necessarily a NUL-terminated string is a common
4612           mistake, and so the call is diagnosed.  To avoid the warning when
4613           the result is not expected to be NUL-terminated, call "memcpy"
4614           instead.
4615
4616                   void copy (char *d, const char *s)
4617                   {
4618                     strncpy (d, s, strlen (s));
4619                   }
4620
4621           In the following example, the call to "strncpy" specifies the size
4622           of the destination buffer as the bound.  If the length of the
4623           source string is equal to or greater than this size the result of
4624           the copy will not be NUL-terminated.  Therefore, the call is also
4625           diagnosed.  To avoid the warning, specify "sizeof buf - 1" as the
4626           bound and set the last element of the buffer to "NUL".
4627
4628                   void copy (const char *s)
4629                   {
4630                     char buf[80];
4631                     strncpy (buf, s, sizeof buf);
4632                     ...
4633                   }
4634
4635           In situations where a character array is intended to store a
4636           sequence of bytes with no terminating "NUL" such an array may be
4637           annotated with attribute "nonstring" to avoid this warning.  Such
4638           arrays, however, are not suitable arguments to functions that
4639           expect "NUL"-terminated strings.  To help detect accidental misuses
4640           of such arrays GCC issues warnings unless it can prove that the use
4641           is safe.
4642
4643       -Wsuggest-attribute=[pure|const|noreturn|format|cold|malloc]
4644           Warn for cases where adding an attribute may be beneficial. The
4645           attributes currently supported are listed below.
4646
4647           -Wsuggest-attribute=pure
4648           -Wsuggest-attribute=const
4649           -Wsuggest-attribute=noreturn
4650           -Wmissing-noreturn
4651           -Wsuggest-attribute=malloc
4652               Warn about functions that might be candidates for attributes
4653               "pure", "const" or "noreturn" or "malloc". The compiler only
4654               warns for functions visible in other compilation units or (in
4655               the case of "pure" and "const") if it cannot prove that the
4656               function returns normally. A function returns normally if it
4657               doesn't contain an infinite loop or return abnormally by
4658               throwing, calling "abort" or trapping.  This analysis requires
4659               option -fipa-pure-const, which is enabled by default at -O and
4660               higher.  Higher optimization levels improve the accuracy of the
4661               analysis.
4662
4663           -Wsuggest-attribute=format
4664           -Wmissing-format-attribute
4665               Warn about function pointers that might be candidates for
4666               "format" attributes.  Note these are only possible candidates,
4667               not absolute ones.  GCC guesses that function pointers with
4668               "format" attributes that are used in assignment,
4669               initialization, parameter passing or return statements should
4670               have a corresponding "format" attribute in the resulting type.
4671               I.e. the left-hand side of the assignment or initialization,
4672               the type of the parameter variable, or the return type of the
4673               containing function respectively should also have a "format"
4674               attribute to avoid the warning.
4675
4676               GCC also warns about function definitions that might be
4677               candidates for "format" attributes.  Again, these are only
4678               possible candidates.  GCC guesses that "format" attributes
4679               might be appropriate for any function that calls a function
4680               like "vprintf" or "vscanf", but this might not always be the
4681               case, and some functions for which "format" attributes are
4682               appropriate may not be detected.
4683
4684           -Wsuggest-attribute=cold
4685               Warn about functions that might be candidates for "cold"
4686               attribute.  This is based on static detection and generally
4687               will only warn about functions which always leads to a call to
4688               another "cold" function such as wrappers of C++ "throw" or
4689               fatal error reporting functions leading to "abort".
4690
4691       -Wsuggest-final-types
4692           Warn about types with virtual methods where code quality would be
4693           improved if the type were declared with the C++11 "final"
4694           specifier, or, if possible, declared in an anonymous namespace.
4695           This allows GCC to more aggressively devirtualize the polymorphic
4696           calls. This warning is more effective with link time optimization,
4697           where the information about the class hierarchy graph is more
4698           complete.
4699
4700       -Wsuggest-final-methods
4701           Warn about virtual methods where code quality would be improved if
4702           the method were declared with the C++11 "final" specifier, or, if
4703           possible, its type were declared in an anonymous namespace or with
4704           the "final" specifier.  This warning is more effective with link-
4705           time optimization, where the information about the class hierarchy
4706           graph is more complete. It is recommended to first consider
4707           suggestions of -Wsuggest-final-types and then rebuild with new
4708           annotations.
4709
4710       -Wsuggest-override
4711           Warn about overriding virtual functions that are not marked with
4712           the override keyword.
4713
4714       -Walloc-zero
4715           Warn about calls to allocation functions decorated with attribute
4716           "alloc_size" that specify zero bytes, including those to the built-
4717           in forms of the functions "aligned_alloc", "alloca", "calloc",
4718           "malloc", and "realloc".  Because the behavior of these functions
4719           when called with a zero size differs among implementations (and in
4720           the case of "realloc" has been deprecated) relying on it may result
4721           in subtle portability bugs and should be avoided.
4722
4723       -Walloc-size-larger-than=byte-size
4724           Warn about calls to functions decorated with attribute "alloc_size"
4725           that attempt to allocate objects larger than the specified number
4726           of bytes, or where the result of the size computation in an integer
4727           type with infinite precision would exceed the value of PTRDIFF_MAX
4728           on the target.  -Walloc-size-larger-than=PTRDIFF_MAX is enabled by
4729           default.  Warnings controlled by the option can be disabled either
4730           by specifying byte-size of SIZE_MAX or more or by
4731           -Wno-alloc-size-larger-than.
4732
4733       -Wno-alloc-size-larger-than
4734           Disable -Walloc-size-larger-than= warnings.  The option is
4735           equivalent to -Walloc-size-larger-than=SIZE_MAX or larger.
4736
4737       -Walloca
4738           This option warns on all uses of "alloca" in the source.
4739
4740       -Walloca-larger-than=byte-size
4741           This option warns on calls to "alloca" with an integer argument
4742           whose value is either zero, or that is not bounded by a controlling
4743           predicate that limits its value to at most byte-size.  It also
4744           warns for calls to "alloca" where the bound value is unknown.
4745           Arguments of non-integer types are considered unbounded even if
4746           they appear to be constrained to the expected range.
4747
4748           For example, a bounded case of "alloca" could be:
4749
4750                   void func (size_t n)
4751                   {
4752                     void *p;
4753                     if (n <= 1000)
4754                       p = alloca (n);
4755                     else
4756                       p = malloc (n);
4757                     f (p);
4758                   }
4759
4760           In the above example, passing "-Walloca-larger-than=1000" would not
4761           issue a warning because the call to "alloca" is known to be at most
4762           1000 bytes.  However, if "-Walloca-larger-than=500" were passed,
4763           the compiler would emit a warning.
4764
4765           Unbounded uses, on the other hand, are uses of "alloca" with no
4766           controlling predicate constraining its integer argument.  For
4767           example:
4768
4769                   void func ()
4770                   {
4771                     void *p = alloca (n);
4772                     f (p);
4773                   }
4774
4775           If "-Walloca-larger-than=500" were passed, the above would trigger
4776           a warning, but this time because of the lack of bounds checking.
4777
4778           Note, that even seemingly correct code involving signed integers
4779           could cause a warning:
4780
4781                   void func (signed int n)
4782                   {
4783                     if (n < 500)
4784                       {
4785                         p = alloca (n);
4786                         f (p);
4787                       }
4788                   }
4789
4790           In the above example, n could be negative, causing a larger than
4791           expected argument to be implicitly cast into the "alloca" call.
4792
4793           This option also warns when "alloca" is used in a loop.
4794
4795           -Walloca-larger-than=PTRDIFF_MAX is enabled by default but is
4796           usually only effective  when -ftree-vrp is active (default for -O2
4797           and above).
4798
4799           See also -Wvla-larger-than=byte-size.
4800
4801       -Wno-alloca-larger-than
4802           Disable -Walloca-larger-than= warnings.  The option is equivalent
4803           to -Walloca-larger-than=SIZE_MAX or larger.
4804
4805       -Warray-bounds
4806       -Warray-bounds=n
4807           This option is only active when -ftree-vrp is active (default for
4808           -O2 and above). It warns about subscripts to arrays that are always
4809           out of bounds. This warning is enabled by -Wall.
4810
4811           -Warray-bounds=1
4812               This is the warning level of -Warray-bounds and is enabled by
4813               -Wall; higher levels are not, and must be explicitly requested.
4814
4815           -Warray-bounds=2
4816               This warning level also warns about out of bounds access for
4817               arrays at the end of a struct and for arrays accessed through
4818               pointers. This warning level may give a larger number of false
4819               positives and is deactivated by default.
4820
4821       -Wattribute-alias=n
4822       -Wno-attribute-alias
4823           Warn about declarations using the "alias" and similar attributes
4824           whose target is incompatible with the type of the alias.
4825
4826           -Wattribute-alias=1
4827               The default warning level of the -Wattribute-alias option
4828               diagnoses incompatibilities between the type of the alias
4829               declaration and that of its target.  Such incompatibilities are
4830               typically indicative of bugs.
4831
4832           -Wattribute-alias=2
4833               At this level -Wattribute-alias also diagnoses cases where the
4834               attributes of the alias declaration are more restrictive than
4835               the attributes applied to its target.  These mismatches can
4836               potentially result in incorrect code generation.  In other
4837               cases they may be benign and could be resolved simply by adding
4838               the missing attribute to the target.  For comparison, see the
4839               -Wmissing-attributes option, which controls diagnostics when
4840               the alias declaration is less restrictive than the target,
4841               rather than more restrictive.
4842
4843               Attributes considered include "alloc_align", "alloc_size",
4844               "cold", "const", "hot", "leaf", "malloc", "nonnull",
4845               "noreturn", "nothrow", "pure", "returns_nonnull", and
4846               "returns_twice".
4847
4848           -Wattribute-alias is equivalent to -Wattribute-alias=1.  This is
4849           the default.  You can disable these warnings with either
4850           -Wno-attribute-alias or -Wattribute-alias=0.
4851
4852       -Wbool-compare
4853           Warn about boolean expression compared with an integer value
4854           different from "true"/"false".  For instance, the following
4855           comparison is always false:
4856
4857                   int n = 5;
4858                   ...
4859                   if ((n > 1) == 2) { ... }
4860
4861           This warning is enabled by -Wall.
4862
4863       -Wbool-operation
4864           Warn about suspicious operations on expressions of a boolean type.
4865           For instance, bitwise negation of a boolean is very likely a bug in
4866           the program.  For C, this warning also warns about incrementing or
4867           decrementing a boolean, which rarely makes sense.  (In C++,
4868           decrementing a boolean is always invalid.  Incrementing a boolean
4869           is invalid in C++17, and deprecated otherwise.)
4870
4871           This warning is enabled by -Wall.
4872
4873       -Wduplicated-branches
4874           Warn when an if-else has identical branches.  This warning detects
4875           cases like
4876
4877                   if (p != NULL)
4878                     return 0;
4879                   else
4880                     return 0;
4881
4882           It doesn't warn when both branches contain just a null statement.
4883           This warning also warn for conditional operators:
4884
4885                     int i = x ? *p : *p;
4886
4887       -Wduplicated-cond
4888           Warn about duplicated conditions in an if-else-if chain.  For
4889           instance, warn for the following code:
4890
4891                   if (p->q != NULL) { ... }
4892                   else if (p->q != NULL) { ... }
4893
4894       -Wframe-address
4895           Warn when the __builtin_frame_address or __builtin_return_address
4896           is called with an argument greater than 0.  Such calls may return
4897           indeterminate values or crash the program.  The warning is included
4898           in -Wall.
4899
4900       -Wno-discarded-qualifiers (C and Objective-C only)
4901           Do not warn if type qualifiers on pointers are being discarded.
4902           Typically, the compiler warns if a "const char *" variable is
4903           passed to a function that takes a "char *" parameter.  This option
4904           can be used to suppress such a warning.
4905
4906       -Wno-discarded-array-qualifiers (C and Objective-C only)
4907           Do not warn if type qualifiers on arrays which are pointer targets
4908           are being discarded. Typically, the compiler warns if a "const int
4909           (*)[]" variable is passed to a function that takes a "int (*)[]"
4910           parameter.  This option can be used to suppress such a warning.
4911
4912       -Wno-incompatible-pointer-types (C and Objective-C only)
4913           Do not warn when there is a conversion between pointers that have
4914           incompatible types.  This warning is for cases not covered by
4915           -Wno-pointer-sign, which warns for pointer argument passing or
4916           assignment with different signedness.
4917
4918       -Wno-int-conversion (C and Objective-C only)
4919           Do not warn about incompatible integer to pointer and pointer to
4920           integer conversions.  This warning is about implicit conversions;
4921           for explicit conversions the warnings -Wno-int-to-pointer-cast and
4922           -Wno-pointer-to-int-cast may be used.
4923
4924       -Wno-div-by-zero
4925           Do not warn about compile-time integer division by zero.  Floating-
4926           point division by zero is not warned about, as it can be a
4927           legitimate way of obtaining infinities and NaNs.
4928
4929       -Wsystem-headers
4930           Print warning messages for constructs found in system header files.
4931           Warnings from system headers are normally suppressed, on the
4932           assumption that they usually do not indicate real problems and
4933           would only make the compiler output harder to read.  Using this
4934           command-line option tells GCC to emit warnings from system headers
4935           as if they occurred in user code.  However, note that using -Wall
4936           in conjunction with this option does not warn about unknown pragmas
4937           in system headers---for that, -Wunknown-pragmas must also be used.
4938
4939       -Wtautological-compare
4940           Warn if a self-comparison always evaluates to true or false.  This
4941           warning detects various mistakes such as:
4942
4943                   int i = 1;
4944                   ...
4945                   if (i > i) { ... }
4946
4947           This warning also warns about bitwise comparisons that always
4948           evaluate to true or false, for instance:
4949
4950                   if ((a & 16) == 10) { ... }
4951
4952           will always be false.
4953
4954           This warning is enabled by -Wall.
4955
4956       -Wtrampolines
4957           Warn about trampolines generated for pointers to nested functions.
4958           A trampoline is a small piece of data or code that is created at
4959           run time on the stack when the address of a nested function is
4960           taken, and is used to call the nested function indirectly.  For
4961           some targets, it is made up of data only and thus requires no
4962           special treatment.  But, for most targets, it is made up of code
4963           and thus requires the stack to be made executable in order for the
4964           program to work properly.
4965
4966       -Wfloat-equal
4967           Warn if floating-point values are used in equality comparisons.
4968
4969           The idea behind this is that sometimes it is convenient (for the
4970           programmer) to consider floating-point values as approximations to
4971           infinitely precise real numbers.  If you are doing this, then you
4972           need to compute (by analyzing the code, or in some other way) the
4973           maximum or likely maximum error that the computation introduces,
4974           and allow for it when performing comparisons (and when producing
4975           output, but that's a different problem).  In particular, instead of
4976           testing for equality, you should check to see whether the two
4977           values have ranges that overlap; and this is done with the
4978           relational operators, so equality comparisons are probably
4979           mistaken.
4980
4981       -Wtraditional (C and Objective-C only)
4982           Warn about certain constructs that behave differently in
4983           traditional and ISO C.  Also warn about ISO C constructs that have
4984           no traditional C equivalent, and/or problematic constructs that
4985           should be avoided.
4986
4987           *   Macro parameters that appear within string literals in the
4988               macro body.  In traditional C macro replacement takes place
4989               within string literals, but in ISO C it does not.
4990
4991           *   In traditional C, some preprocessor directives did not exist.
4992               Traditional preprocessors only considered a line to be a
4993               directive if the # appeared in column 1 on the line.  Therefore
4994               -Wtraditional warns about directives that traditional C
4995               understands but ignores because the # does not appear as the
4996               first character on the line.  It also suggests you hide
4997               directives like "#pragma" not understood by traditional C by
4998               indenting them.  Some traditional implementations do not
4999               recognize "#elif", so this option suggests avoiding it
5000               altogether.
5001
5002           *   A function-like macro that appears without arguments.
5003
5004           *   The unary plus operator.
5005
5006           *   The U integer constant suffix, or the F or L floating-point
5007               constant suffixes.  (Traditional C does support the L suffix on
5008               integer constants.)  Note, these suffixes appear in macros
5009               defined in the system headers of most modern systems, e.g. the
5010               _MIN/_MAX macros in "<limits.h>".  Use of these macros in user
5011               code might normally lead to spurious warnings, however GCC's
5012               integrated preprocessor has enough context to avoid warning in
5013               these cases.
5014
5015           *   A function declared external in one block and then used after
5016               the end of the block.
5017
5018           *   A "switch" statement has an operand of type "long".
5019
5020           *   A non-"static" function declaration follows a "static" one.
5021               This construct is not accepted by some traditional C compilers.
5022
5023           *   The ISO type of an integer constant has a different width or
5024               signedness from its traditional type.  This warning is only
5025               issued if the base of the constant is ten.  I.e. hexadecimal or
5026               octal values, which typically represent bit patterns, are not
5027               warned about.
5028
5029           *   Usage of ISO string concatenation is detected.
5030
5031           *   Initialization of automatic aggregates.
5032
5033           *   Identifier conflicts with labels.  Traditional C lacks a
5034               separate namespace for labels.
5035
5036           *   Initialization of unions.  If the initializer is zero, the
5037               warning is omitted.  This is done under the assumption that the
5038               zero initializer in user code appears conditioned on e.g.
5039               "__STDC__" to avoid missing initializer warnings and relies on
5040               default initialization to zero in the traditional C case.
5041
5042           *   Conversions by prototypes between fixed/floating-point values
5043               and vice versa.  The absence of these prototypes when compiling
5044               with traditional C causes serious problems.  This is a subset
5045               of the possible conversion warnings; for the full set use
5046               -Wtraditional-conversion.
5047
5048           *   Use of ISO C style function definitions.  This warning
5049               intentionally is not issued for prototype declarations or
5050               variadic functions because these ISO C features appear in your
5051               code when using libiberty's traditional C compatibility macros,
5052               "PARAMS" and "VPARAMS".  This warning is also bypassed for
5053               nested functions because that feature is already a GCC
5054               extension and thus not relevant to traditional C compatibility.
5055
5056       -Wtraditional-conversion (C and Objective-C only)
5057           Warn if a prototype causes a type conversion that is different from
5058           what would happen to the same argument in the absence of a
5059           prototype.  This includes conversions of fixed point to floating
5060           and vice versa, and conversions changing the width or signedness of
5061           a fixed-point argument except when the same as the default
5062           promotion.
5063
5064       -Wdeclaration-after-statement (C and Objective-C only)
5065           Warn when a declaration is found after a statement in a block.
5066           This construct, known from C++, was introduced with ISO C99 and is
5067           by default allowed in GCC.  It is not supported by ISO C90.
5068
5069       -Wshadow
5070           Warn whenever a local variable or type declaration shadows another
5071           variable, parameter, type, class member (in C++), or instance
5072           variable (in Objective-C) or whenever a built-in function is
5073           shadowed. Note that in C++, the compiler warns if a local variable
5074           shadows an explicit typedef, but not if it shadows a
5075           struct/class/enum.  Same as -Wshadow=global.
5076
5077       -Wno-shadow-ivar (Objective-C only)
5078           Do not warn whenever a local variable shadows an instance variable
5079           in an Objective-C method.
5080
5081       -Wshadow=global
5082           The default for -Wshadow. Warns for any (global) shadowing.
5083
5084       -Wshadow=local
5085           Warn when a local variable shadows another local variable or
5086           parameter.  This warning is enabled by -Wshadow=global.
5087
5088       -Wshadow=compatible-local
5089           Warn when a local variable shadows another local variable or
5090           parameter whose type is compatible with that of the shadowing
5091           variable. In C++, type compatibility here means the type of the
5092           shadowing variable can be converted to that of the shadowed
5093           variable. The creation of this flag (in addition to -Wshadow=local)
5094           is based on the idea that when a local variable shadows another one
5095           of incompatible type, it is most likely intentional, not a bug or
5096           typo, as shown in the following example:
5097
5098                   for (SomeIterator i = SomeObj.begin(); i != SomeObj.end(); ++i)
5099                   {
5100                     for (int i = 0; i < N; ++i)
5101                     {
5102                       ...
5103                     }
5104                     ...
5105                   }
5106
5107           Since the two variable "i" in the example above have incompatible
5108           types, enabling only -Wshadow=compatible-local will not emit a
5109           warning.  Because their types are incompatible, if a programmer
5110           accidentally uses one in place of the other, type checking will
5111           catch that and emit an error or warning. So not warning (about
5112           shadowing) in this case will not lead to undetected bugs. Use of
5113           this flag instead of -Wshadow=local can possibly reduce the number
5114           of warnings triggered by intentional shadowing.
5115
5116           This warning is enabled by -Wshadow=local.
5117
5118       -Wlarger-than=byte-size
5119           Warn whenever an object is defined whose size exceeds byte-size.
5120           -Wlarger-than=PTRDIFF_MAX is enabled by default.  Warnings
5121           controlled by the option can be disabled either by specifying byte-
5122           size of SIZE_MAX or more or by -Wno-larger-than.
5123
5124       -Wno-larger-than
5125           Disable -Wlarger-than= warnings.  The option is equivalent to
5126           -Wlarger-than=SIZE_MAX or larger.
5127
5128       -Wframe-larger-than=byte-size
5129           Warn if the size of a function frame exceeds byte-size.  The
5130           computation done to determine the stack frame size is approximate
5131           and not conservative.  The actual requirements may be somewhat
5132           greater than byte-size even if you do not get a warning.  In
5133           addition, any space allocated via "alloca", variable-length arrays,
5134           or related constructs is not included by the compiler when
5135           determining whether or not to issue a warning.
5136           -Wframe-larger-than=PTRDIFF_MAX is enabled by default.  Warnings
5137           controlled by the option can be disabled either by specifying byte-
5138           size of SIZE_MAX or more or by -Wno-frame-larger-than.
5139
5140       -Wno-frame-larger-than
5141           Disable -Wframe-larger-than= warnings.  The option is equivalent to
5142           -Wframe-larger-than=SIZE_MAX or larger.
5143
5144       -Wno-free-nonheap-object
5145           Do not warn when attempting to free an object that was not
5146           allocated on the heap.
5147
5148       -Wstack-usage=byte-size
5149           Warn if the stack usage of a function might exceed byte-size.  The
5150           computation done to determine the stack usage is conservative.  Any
5151           space allocated via "alloca", variable-length arrays, or related
5152           constructs is included by the compiler when determining whether or
5153           not to issue a warning.
5154
5155           The message is in keeping with the output of -fstack-usage.
5156
5157           *   If the stack usage is fully static but exceeds the specified
5158               amount, it's:
5159
5160                         warning: stack usage is 1120 bytes
5161
5162           *   If the stack usage is (partly) dynamic but bounded, it's:
5163
5164                         warning: stack usage might be 1648 bytes
5165
5166           *   If the stack usage is (partly) dynamic and not bounded, it's:
5167
5168                         warning: stack usage might be unbounded
5169
5170           -Wstack-usage=PTRDIFF_MAX is enabled by default.  Warnings
5171           controlled by the option can be disabled either by specifying byte-
5172           size of SIZE_MAX or more or by -Wno-stack-usage.
5173
5174       -Wno-stack-usage
5175           Disable -Wstack-usage= warnings.  The option is equivalent to
5176           -Wstack-usage=SIZE_MAX or larger.
5177
5178       -Wunsafe-loop-optimizations
5179           Warn if the loop cannot be optimized because the compiler cannot
5180           assume anything on the bounds of the loop indices.  With
5181           -funsafe-loop-optimizations warn if the compiler makes such
5182           assumptions.
5183
5184       -Wno-pedantic-ms-format (MinGW targets only)
5185           When used in combination with -Wformat and -pedantic without GNU
5186           extensions, this option disables the warnings about non-ISO
5187           "printf" / "scanf" format width specifiers "I32", "I64", and "I"
5188           used on Windows targets, which depend on the MS runtime.
5189
5190       -Waligned-new
5191           Warn about a new-expression of a type that requires greater
5192           alignment than the "alignof(std::max_align_t)" but uses an
5193           allocation function without an explicit alignment parameter. This
5194           option is enabled by -Wall.
5195
5196           Normally this only warns about global allocation functions, but
5197           -Waligned-new=all also warns about class member allocation
5198           functions.
5199
5200       -Wplacement-new
5201       -Wplacement-new=n
5202           Warn about placement new expressions with undefined behavior, such
5203           as constructing an object in a buffer that is smaller than the type
5204           of the object.  For example, the placement new expression below is
5205           diagnosed because it attempts to construct an array of 64 integers
5206           in a buffer only 64 bytes large.
5207
5208                   char buf [64];
5209                   new (buf) int[64];
5210
5211           This warning is enabled by default.
5212
5213           -Wplacement-new=1
5214               This is the default warning level of -Wplacement-new.  At this
5215               level the warning is not issued for some strictly undefined
5216               constructs that GCC allows as extensions for compatibility with
5217               legacy code.  For example, the following "new" expression is
5218               not diagnosed at this level even though it has undefined
5219               behavior according to the C++ standard because it writes past
5220               the end of the one-element array.
5221
5222                       struct S { int n, a[1]; };
5223                       S *s = (S *)malloc (sizeof *s + 31 * sizeof s->a[0]);
5224                       new (s->a)int [32]();
5225
5226           -Wplacement-new=2
5227               At this level, in addition to diagnosing all the same
5228               constructs as at level 1, a diagnostic is also issued for
5229               placement new expressions that construct an object in the last
5230               member of structure whose type is an array of a single element
5231               and whose size is less than the size of the object being
5232               constructed.  While the previous example would be diagnosed,
5233               the following construct makes use of the flexible member array
5234               extension to avoid the warning at level 2.
5235
5236                       struct S { int n, a[]; };
5237                       S *s = (S *)malloc (sizeof *s + 32 * sizeof s->a[0]);
5238                       new (s->a)int [32]();
5239
5240       -Wpointer-arith
5241           Warn about anything that depends on the "size of" a function type
5242           or of "void".  GNU C assigns these types a size of 1, for
5243           convenience in calculations with "void *" pointers and pointers to
5244           functions.  In C++, warn also when an arithmetic operation involves
5245           "NULL".  This warning is also enabled by -Wpedantic.
5246
5247       -Wpointer-compare
5248           Warn if a pointer is compared with a zero character constant.  This
5249           usually means that the pointer was meant to be dereferenced.  For
5250           example:
5251
5252                   const char *p = foo ();
5253                   if (p == '\0')
5254                     return 42;
5255
5256           Note that the code above is invalid in C++11.
5257
5258           This warning is enabled by default.
5259
5260       -Wtype-limits
5261           Warn if a comparison is always true or always false due to the
5262           limited range of the data type, but do not warn for constant
5263           expressions.  For example, warn if an unsigned variable is compared
5264           against zero with "<" or ">=".  This warning is also enabled by
5265           -Wextra.
5266
5267       -Wabsolute-value (C and Objective-C only)
5268           Warn for calls to standard functions that compute the absolute
5269           value of an argument when a more appropriate standard function is
5270           available.  For example, calling "abs(3.14)" triggers the warning
5271           because the appropriate function to call to compute the absolute
5272           value of a double argument is "fabs".  The option also triggers
5273           warnings when the argument in a call to such a function has an
5274           unsigned type.  This warning can be suppressed with an explicit
5275           type cast and it is also enabled by -Wextra.
5276
5277       -Wcomment
5278       -Wcomments
5279           Warn whenever a comment-start sequence /* appears in a /* comment,
5280           or whenever a backslash-newline appears in a // comment.  This
5281           warning is enabled by -Wall.
5282
5283       -Wtrigraphs
5284           Warn if any trigraphs are encountered that might change the meaning
5285           of the program.  Trigraphs within comments are not warned about,
5286           except those that would form escaped newlines.
5287
5288           This option is implied by -Wall.  If -Wall is not given, this
5289           option is still enabled unless trigraphs are enabled.  To get
5290           trigraph conversion without warnings, but get the other -Wall
5291           warnings, use -trigraphs -Wall -Wno-trigraphs.
5292
5293       -Wundef
5294           Warn if an undefined identifier is evaluated in an "#if" directive.
5295           Such identifiers are replaced with zero.
5296
5297       -Wexpansion-to-defined
5298           Warn whenever defined is encountered in the expansion of a macro
5299           (including the case where the macro is expanded by an #if
5300           directive).  Such usage is not portable.  This warning is also
5301           enabled by -Wpedantic and -Wextra.
5302
5303       -Wunused-macros
5304           Warn about macros defined in the main file that are unused.  A
5305           macro is used if it is expanded or tested for existence at least
5306           once.  The preprocessor also warns if the macro has not been used
5307           at the time it is redefined or undefined.
5308
5309           Built-in macros, macros defined on the command line, and macros
5310           defined in include files are not warned about.
5311
5312           Note: If a macro is actually used, but only used in skipped
5313           conditional blocks, then the preprocessor reports it as unused.  To
5314           avoid the warning in such a case, you might improve the scope of
5315           the macro's definition by, for example, moving it into the first
5316           skipped block.  Alternatively, you could provide a dummy use with
5317           something like:
5318
5319                   #if defined the_macro_causing_the_warning
5320                   #endif
5321
5322       -Wno-endif-labels
5323           Do not warn whenever an "#else" or an "#endif" are followed by
5324           text.  This sometimes happens in older programs with code of the
5325           form
5326
5327                   #if FOO
5328                   ...
5329                   #else FOO
5330                   ...
5331                   #endif FOO
5332
5333           The second and third "FOO" should be in comments.  This warning is
5334           on by default.
5335
5336       -Wbad-function-cast (C and Objective-C only)
5337           Warn when a function call is cast to a non-matching type.  For
5338           example, warn if a call to a function returning an integer type is
5339           cast to a pointer type.
5340
5341       -Wc90-c99-compat (C and Objective-C only)
5342           Warn about features not present in ISO C90, but present in ISO C99.
5343           For instance, warn about use of variable length arrays, "long long"
5344           type, "bool" type, compound literals, designated initializers, and
5345           so on.  This option is independent of the standards mode.  Warnings
5346           are disabled in the expression that follows "__extension__".
5347
5348       -Wc99-c11-compat (C and Objective-C only)
5349           Warn about features not present in ISO C99, but present in ISO C11.
5350           For instance, warn about use of anonymous structures and unions,
5351           "_Atomic" type qualifier, "_Thread_local" storage-class specifier,
5352           "_Alignas" specifier, "Alignof" operator, "_Generic" keyword, and
5353           so on.  This option is independent of the standards mode.  Warnings
5354           are disabled in the expression that follows "__extension__".
5355
5356       -Wc++-compat (C and Objective-C only)
5357           Warn about ISO C constructs that are outside of the common subset
5358           of ISO C and ISO C++, e.g. request for implicit conversion from
5359           "void *" to a pointer to non-"void" type.
5360
5361       -Wc++11-compat (C++ and Objective-C++ only)
5362           Warn about C++ constructs whose meaning differs between ISO C++
5363           1998 and ISO C++ 2011, e.g., identifiers in ISO C++ 1998 that are
5364           keywords in ISO C++ 2011.  This warning turns on -Wnarrowing and is
5365           enabled by -Wall.
5366
5367       -Wc++14-compat (C++ and Objective-C++ only)
5368           Warn about C++ constructs whose meaning differs between ISO C++
5369           2011 and ISO C++ 2014.  This warning is enabled by -Wall.
5370
5371       -Wc++17-compat (C++ and Objective-C++ only)
5372           Warn about C++ constructs whose meaning differs between ISO C++
5373           2014 and ISO C++ 2017.  This warning is enabled by -Wall.
5374
5375       -Wcast-qual
5376           Warn whenever a pointer is cast so as to remove a type qualifier
5377           from the target type.  For example, warn if a "const char *" is
5378           cast to an ordinary "char *".
5379
5380           Also warn when making a cast that introduces a type qualifier in an
5381           unsafe way.  For example, casting "char **" to "const char **" is
5382           unsafe, as in this example:
5383
5384                     /* p is char ** value.  */
5385                     const char **q = (const char **) p;
5386                     /* Assignment of readonly string to const char * is OK.  */
5387                     *q = "string";
5388                     /* Now char** pointer points to read-only memory.  */
5389                     **p = 'b';
5390
5391       -Wcast-align
5392           Warn whenever a pointer is cast such that the required alignment of
5393           the target is increased.  For example, warn if a "char *" is cast
5394           to an "int *" on machines where integers can only be accessed at
5395           two- or four-byte boundaries.
5396
5397       -Wcast-align=strict
5398           Warn whenever a pointer is cast such that the required alignment of
5399           the target is increased.  For example, warn if a "char *" is cast
5400           to an "int *" regardless of the target machine.
5401
5402       -Wcast-function-type
5403           Warn when a function pointer is cast to an incompatible function
5404           pointer.  In a cast involving function types with a variable
5405           argument list only the types of initial arguments that are provided
5406           are considered.  Any parameter of pointer-type matches any other
5407           pointer-type.  Any benign differences in integral types are
5408           ignored, like "int" vs. "long" on ILP32 targets.  Likewise type
5409           qualifiers are ignored.  The function type "void (*) (void)" is
5410           special and matches everything, which can be used to suppress this
5411           warning.  In a cast involving pointer to member types this warning
5412           warns whenever the type cast is changing the pointer to member
5413           type.  This warning is enabled by -Wextra.
5414
5415       -Wwrite-strings
5416           When compiling C, give string constants the type "const
5417           char[length]" so that copying the address of one into a non-"const"
5418           "char *" pointer produces a warning.  These warnings help you find
5419           at compile time code that can try to write into a string constant,
5420           but only if you have been very careful about using "const" in
5421           declarations and prototypes.  Otherwise, it is just a nuisance.
5422           This is why we did not make -Wall request these warnings.
5423
5424           When compiling C++, warn about the deprecated conversion from
5425           string literals to "char *".  This warning is enabled by default
5426           for C++ programs.
5427
5428       -Wcatch-value
5429       -Wcatch-value=n (C++ and Objective-C++ only)
5430           Warn about catch handlers that do not catch via reference.  With
5431           -Wcatch-value=1 (or -Wcatch-value for short) warn about polymorphic
5432           class types that are caught by value.  With -Wcatch-value=2 warn
5433           about all class types that are caught by value. With
5434           -Wcatch-value=3 warn about all types that are not caught by
5435           reference. -Wcatch-value is enabled by -Wall.
5436
5437       -Wclobbered
5438           Warn for variables that might be changed by "longjmp" or "vfork".
5439           This warning is also enabled by -Wextra.
5440
5441       -Wconditionally-supported (C++ and Objective-C++ only)
5442           Warn for conditionally-supported (C++11 [intro.defs]) constructs.
5443
5444       -Wconversion
5445           Warn for implicit conversions that may alter a value. This includes
5446           conversions between real and integer, like "abs (x)" when "x" is
5447           "double"; conversions between signed and unsigned, like "unsigned
5448           ui = -1"; and conversions to smaller types, like "sqrtf (M_PI)". Do
5449           not warn for explicit casts like "abs ((int) x)" and "ui =
5450           (unsigned) -1", or if the value is not changed by the conversion
5451           like in "abs (2.0)".  Warnings about conversions between signed and
5452           unsigned integers can be disabled by using -Wno-sign-conversion.
5453
5454           For C++, also warn for confusing overload resolution for user-
5455           defined conversions; and conversions that never use a type
5456           conversion operator: conversions to "void", the same type, a base
5457           class or a reference to them. Warnings about conversions between
5458           signed and unsigned integers are disabled by default in C++ unless
5459           -Wsign-conversion is explicitly enabled.
5460
5461       -Wno-conversion-null (C++ and Objective-C++ only)
5462           Do not warn for conversions between "NULL" and non-pointer types.
5463           -Wconversion-null is enabled by default.
5464
5465       -Wzero-as-null-pointer-constant (C++ and Objective-C++ only)
5466           Warn when a literal 0 is used as null pointer constant.  This can
5467           be useful to facilitate the conversion to "nullptr" in C++11.
5468
5469       -Wsubobject-linkage (C++ and Objective-C++ only)
5470           Warn if a class type has a base or a field whose type uses the
5471           anonymous namespace or depends on a type with no linkage.  If a
5472           type A depends on a type B with no or internal linkage, defining it
5473           in multiple translation units would be an ODR violation because the
5474           meaning of B is different in each translation unit.  If A only
5475           appears in a single translation unit, the best way to silence the
5476           warning is to give it internal linkage by putting it in an
5477           anonymous namespace as well.  The compiler doesn't give this
5478           warning for types defined in the main .C file, as those are
5479           unlikely to have multiple definitions.  -Wsubobject-linkage is
5480           enabled by default.
5481
5482       -Wdangling-else
5483           Warn about constructions where there may be confusion to which "if"
5484           statement an "else" branch belongs.  Here is an example of such a
5485           case:
5486
5487                   {
5488                     if (a)
5489                       if (b)
5490                         foo ();
5491                     else
5492                       bar ();
5493                   }
5494
5495           In C/C++, every "else" branch belongs to the innermost possible
5496           "if" statement, which in this example is "if (b)".  This is often
5497           not what the programmer expected, as illustrated in the above
5498           example by indentation the programmer chose.  When there is the
5499           potential for this confusion, GCC issues a warning when this flag
5500           is specified.  To eliminate the warning, add explicit braces around
5501           the innermost "if" statement so there is no way the "else" can
5502           belong to the enclosing "if".  The resulting code looks like this:
5503
5504                   {
5505                     if (a)
5506                       {
5507                         if (b)
5508                           foo ();
5509                         else
5510                           bar ();
5511                       }
5512                   }
5513
5514           This warning is enabled by -Wparentheses.
5515
5516       -Wdate-time
5517           Warn when macros "__TIME__", "__DATE__" or "__TIMESTAMP__" are
5518           encountered as they might prevent bit-wise-identical reproducible
5519           compilations.
5520
5521       -Wdelete-incomplete (C++ and Objective-C++ only)
5522           Warn when deleting a pointer to incomplete type, which may cause
5523           undefined behavior at runtime.  This warning is enabled by default.
5524
5525       -Wuseless-cast (C++ and Objective-C++ only)
5526           Warn when an expression is casted to its own type.
5527
5528       -Wempty-body
5529           Warn if an empty body occurs in an "if", "else" or "do while"
5530           statement.  This warning is also enabled by -Wextra.
5531
5532       -Wenum-compare
5533           Warn about a comparison between values of different enumerated
5534           types.  In C++ enumerated type mismatches in conditional
5535           expressions are also diagnosed and the warning is enabled by
5536           default.  In C this warning is enabled by -Wall.
5537
5538       -Wextra-semi (C++, Objective-C++ only)
5539           Warn about redundant semicolon after in-class function definition.
5540
5541       -Wjump-misses-init (C, Objective-C only)
5542           Warn if a "goto" statement or a "switch" statement jumps forward
5543           across the initialization of a variable, or jumps backward to a
5544           label after the variable has been initialized.  This only warns
5545           about variables that are initialized when they are declared.  This
5546           warning is only supported for C and Objective-C; in C++ this sort
5547           of branch is an error in any case.
5548
5549           -Wjump-misses-init is included in -Wc++-compat.  It can be disabled
5550           with the -Wno-jump-misses-init option.
5551
5552       -Wsign-compare
5553           Warn when a comparison between signed and unsigned values could
5554           produce an incorrect result when the signed value is converted to
5555           unsigned.  In C++, this warning is also enabled by -Wall.  In C, it
5556           is also enabled by -Wextra.
5557
5558       -Wsign-conversion
5559           Warn for implicit conversions that may change the sign of an
5560           integer value, like assigning a signed integer expression to an
5561           unsigned integer variable. An explicit cast silences the warning.
5562           In C, this option is enabled also by -Wconversion.
5563
5564       -Wfloat-conversion
5565           Warn for implicit conversions that reduce the precision of a real
5566           value.  This includes conversions from real to integer, and from
5567           higher precision real to lower precision real values.  This option
5568           is also enabled by -Wconversion.
5569
5570       -Wno-scalar-storage-order
5571           Do not warn on suspicious constructs involving reverse scalar
5572           storage order.
5573
5574       -Wsized-deallocation (C++ and Objective-C++ only)
5575           Warn about a definition of an unsized deallocation function
5576
5577                   void operator delete (void *) noexcept;
5578                   void operator delete[] (void *) noexcept;
5579
5580           without a definition of the corresponding sized deallocation
5581           function
5582
5583                   void operator delete (void *, std::size_t) noexcept;
5584                   void operator delete[] (void *, std::size_t) noexcept;
5585
5586           or vice versa.  Enabled by -Wextra along with -fsized-deallocation.
5587
5588       -Wsizeof-pointer-div
5589           Warn for suspicious divisions of two sizeof expressions that divide
5590           the pointer size by the element size, which is the usual way to
5591           compute the array size but won't work out correctly with pointers.
5592           This warning warns e.g. about "sizeof (ptr) / sizeof (ptr[0])" if
5593           "ptr" is not an array, but a pointer.  This warning is enabled by
5594           -Wall.
5595
5596       -Wsizeof-pointer-memaccess
5597           Warn for suspicious length parameters to certain string and memory
5598           built-in functions if the argument uses "sizeof".  This warning
5599           triggers for example for "memset (ptr, 0, sizeof (ptr));" if "ptr"
5600           is not an array, but a pointer, and suggests a possible fix, or
5601           about "memcpy (&foo, ptr, sizeof (&foo));".
5602           -Wsizeof-pointer-memaccess also warns about calls to bounded string
5603           copy functions like "strncat" or "strncpy" that specify as the
5604           bound a "sizeof" expression of the source array.  For example, in
5605           the following function the call to "strncat" specifies the size of
5606           the source string as the bound.  That is almost certainly a mistake
5607           and so the call is diagnosed.
5608
5609                   void make_file (const char *name)
5610                   {
5611                     char path[PATH_MAX];
5612                     strncpy (path, name, sizeof path - 1);
5613                     strncat (path, ".text", sizeof ".text");
5614                     ...
5615                   }
5616
5617           The -Wsizeof-pointer-memaccess option is enabled by -Wall.
5618
5619       -Wsizeof-array-argument
5620           Warn when the "sizeof" operator is applied to a parameter that is
5621           declared as an array in a function definition.  This warning is
5622           enabled by default for C and C++ programs.
5623
5624       -Wmemset-elt-size
5625           Warn for suspicious calls to the "memset" built-in function, if the
5626           first argument references an array, and the third argument is a
5627           number equal to the number of elements, but not equal to the size
5628           of the array in memory.  This indicates that the user has omitted a
5629           multiplication by the element size.  This warning is enabled by
5630           -Wall.
5631
5632       -Wmemset-transposed-args
5633           Warn for suspicious calls to the "memset" built-in function where
5634           the second argument is not zero and the third argument is zero.
5635           For example, the call "memset (buf, sizeof buf, 0)" is diagnosed
5636           because "memset (buf, 0, sizeof buf)" was meant instead.  The
5637           diagnostic is only emitted if the third argument is a literal zero.
5638           Otherwise, if it is an expression that is folded to zero, or a cast
5639           of zero to some type, it is far less likely that the arguments have
5640           been mistakenly transposed and no warning is emitted.  This warning
5641           is enabled by -Wall.
5642
5643       -Waddress
5644           Warn about suspicious uses of memory addresses. These include using
5645           the address of a function in a conditional expression, such as
5646           "void func(void); if (func)", and comparisons against the memory
5647           address of a string literal, such as "if (x == "abc")".  Such uses
5648           typically indicate a programmer error: the address of a function
5649           always evaluates to true, so their use in a conditional usually
5650           indicate that the programmer forgot the parentheses in a function
5651           call; and comparisons against string literals result in unspecified
5652           behavior and are not portable in C, so they usually indicate that
5653           the programmer intended to use "strcmp".  This warning is enabled
5654           by -Wall.
5655
5656       -Waddress-of-packed-member
5657           Warn when the address of packed member of struct or union is taken,
5658           which usually results in an unaligned pointer value.  This is
5659           enabled by default.
5660
5661       -Wlogical-op
5662           Warn about suspicious uses of logical operators in expressions.
5663           This includes using logical operators in contexts where a bit-wise
5664           operator is likely to be expected.  Also warns when the operands of
5665           a logical operator are the same:
5666
5667                   extern int a;
5668                   if (a < 0 && a < 0) { ... }
5669
5670       -Wlogical-not-parentheses
5671           Warn about logical not used on the left hand side operand of a
5672           comparison.  This option does not warn if the right operand is
5673           considered to be a boolean expression.  Its purpose is to detect
5674           suspicious code like the following:
5675
5676                   int a;
5677                   ...
5678                   if (!a > 1) { ... }
5679
5680           It is possible to suppress the warning by wrapping the LHS into
5681           parentheses:
5682
5683                   if ((!a) > 1) { ... }
5684
5685           This warning is enabled by -Wall.
5686
5687       -Waggregate-return
5688           Warn if any functions that return structures or unions are defined
5689           or called.  (In languages where you can return an array, this also
5690           elicits a warning.)
5691
5692       -Wno-aggressive-loop-optimizations
5693           Warn if in a loop with constant number of iterations the compiler
5694           detects undefined behavior in some statement during one or more of
5695           the iterations.
5696
5697       -Wno-attributes
5698           Do not warn if an unexpected "__attribute__" is used, such as
5699           unrecognized attributes, function attributes applied to variables,
5700           etc.  This does not stop errors for incorrect use of supported
5701           attributes.
5702
5703       -Wno-builtin-declaration-mismatch
5704           Warn if a built-in function is declared with an incompatible
5705           signature or as a non-function, or when a built-in function
5706           declared with a type that does not include a prototype is called
5707           with arguments whose promoted types do not match those expected by
5708           the function.  When -Wextra is specified, also warn when a built-in
5709           function that takes arguments is declared without a prototype.  The
5710           -Wno-builtin-declaration-mismatch warning is enabled by default.
5711           To avoid the warning include the appropriate header to bring the
5712           prototypes of built-in functions into scope.
5713
5714           For example, the call to "memset" below is diagnosed by the warning
5715           because the function expects a value of type "size_t" as its
5716           argument but the type of 32 is "int".  With -Wextra, the
5717           declaration of the function is diagnosed as well.
5718
5719                   extern void* memset ();
5720                   void f (void *d)
5721                   {
5722                     memset (d, '\0', 32);
5723                   }
5724
5725       -Wno-builtin-macro-redefined
5726           Do not warn if certain built-in macros are redefined.  This
5727           suppresses warnings for redefinition of "__TIMESTAMP__",
5728           "__TIME__", "__DATE__", "__FILE__", and "__BASE_FILE__".
5729
5730       -Wstrict-prototypes (C and Objective-C only)
5731           Warn if a function is declared or defined without specifying the
5732           argument types.  (An old-style function definition is permitted
5733           without a warning if preceded by a declaration that specifies the
5734           argument types.)
5735
5736       -Wold-style-declaration (C and Objective-C only)
5737           Warn for obsolescent usages, according to the C Standard, in a
5738           declaration. For example, warn if storage-class specifiers like
5739           "static" are not the first things in a declaration.  This warning
5740           is also enabled by -Wextra.
5741
5742       -Wold-style-definition (C and Objective-C only)
5743           Warn if an old-style function definition is used.  A warning is
5744           given even if there is a previous prototype.
5745
5746       -Wmissing-parameter-type (C and Objective-C only)
5747           A function parameter is declared without a type specifier in
5748           K&R-style functions:
5749
5750                   void foo(bar) { }
5751
5752           This warning is also enabled by -Wextra.
5753
5754       -Wmissing-prototypes (C and Objective-C only)
5755           Warn if a global function is defined without a previous prototype
5756           declaration.  This warning is issued even if the definition itself
5757           provides a prototype.  Use this option to detect global functions
5758           that do not have a matching prototype declaration in a header file.
5759           This option is not valid for C++ because all function declarations
5760           provide prototypes and a non-matching declaration declares an
5761           overload rather than conflict with an earlier declaration.  Use
5762           -Wmissing-declarations to detect missing declarations in C++.
5763
5764       -Wmissing-declarations
5765           Warn if a global function is defined without a previous
5766           declaration.  Do so even if the definition itself provides a
5767           prototype.  Use this option to detect global functions that are not
5768           declared in header files.  In C, no warnings are issued for
5769           functions with previous non-prototype declarations; use
5770           -Wmissing-prototypes to detect missing prototypes.  In C++, no
5771           warnings are issued for function templates, or for inline
5772           functions, or for functions in anonymous namespaces.
5773
5774       -Wmissing-field-initializers
5775           Warn if a structure's initializer has some fields missing.  For
5776           example, the following code causes such a warning, because "x.h" is
5777           implicitly zero:
5778
5779                   struct s { int f, g, h; };
5780                   struct s x = { 3, 4 };
5781
5782           This option does not warn about designated initializers, so the
5783           following modification does not trigger a warning:
5784
5785                   struct s { int f, g, h; };
5786                   struct s x = { .f = 3, .g = 4 };
5787
5788           In C this option does not warn about the universal zero initializer
5789           { 0 }:
5790
5791                   struct s { int f, g, h; };
5792                   struct s x = { 0 };
5793
5794           Likewise, in C++ this option does not warn about the empty { }
5795           initializer, for example:
5796
5797                   struct s { int f, g, h; };
5798                   s x = { };
5799
5800           This warning is included in -Wextra.  To get other -Wextra warnings
5801           without this one, use -Wextra -Wno-missing-field-initializers.
5802
5803       -Wno-multichar
5804           Do not warn if a multicharacter constant ('FOOF') is used.  Usually
5805           they indicate a typo in the user's code, as they have
5806           implementation-defined values, and should not be used in portable
5807           code.
5808
5809       -Wnormalized=[none|id|nfc|nfkc]
5810           In ISO C and ISO C++, two identifiers are different if they are
5811           different sequences of characters.  However, sometimes when
5812           characters outside the basic ASCII character set are used, you can
5813           have two different character sequences that look the same.  To
5814           avoid confusion, the ISO 10646 standard sets out some normalization
5815           rules which when applied ensure that two sequences that look the
5816           same are turned into the same sequence.  GCC can warn you if you
5817           are using identifiers that have not been normalized; this option
5818           controls that warning.
5819
5820           There are four levels of warning supported by GCC.  The default is
5821           -Wnormalized=nfc, which warns about any identifier that is not in
5822           the ISO 10646 "C" normalized form, NFC.  NFC is the recommended
5823           form for most uses.  It is equivalent to -Wnormalized.
5824
5825           Unfortunately, there are some characters allowed in identifiers by
5826           ISO C and ISO C++ that, when turned into NFC, are not allowed in
5827           identifiers.  That is, there's no way to use these symbols in
5828           portable ISO C or C++ and have all your identifiers in NFC.
5829           -Wnormalized=id suppresses the warning for these characters.  It is
5830           hoped that future versions of the standards involved will correct
5831           this, which is why this option is not the default.
5832
5833           You can switch the warning off for all characters by writing
5834           -Wnormalized=none or -Wno-normalized.  You should only do this if
5835           you are using some other normalization scheme (like "D"), because
5836           otherwise you can easily create bugs that are literally impossible
5837           to see.
5838
5839           Some characters in ISO 10646 have distinct meanings but look
5840           identical in some fonts or display methodologies, especially once
5841           formatting has been applied.  For instance "\u207F", "SUPERSCRIPT
5842           LATIN SMALL LETTER N", displays just like a regular "n" that has
5843           been placed in a superscript.  ISO 10646 defines the NFKC
5844           normalization scheme to convert all these into a standard form as
5845           well, and GCC warns if your code is not in NFKC if you use
5846           -Wnormalized=nfkc.  This warning is comparable to warning about
5847           every identifier that contains the letter O because it might be
5848           confused with the digit 0, and so is not the default, but may be
5849           useful as a local coding convention if the programming environment
5850           cannot be fixed to display these characters distinctly.
5851
5852       -Wno-attribute-warning
5853           Do not warn about usage of functions declared with "warning"
5854           attribute.  By default, this warning is enabled.
5855           -Wno-attribute-warning can be used to disable the warning or
5856           -Wno-error=attribute-warning can be used to disable the error when
5857           compiled with -Werror flag.
5858
5859       -Wno-deprecated
5860           Do not warn about usage of deprecated features.
5861
5862       -Wno-deprecated-declarations
5863           Do not warn about uses of functions, variables, and types marked as
5864           deprecated by using the "deprecated" attribute.
5865
5866       -Wno-overflow
5867           Do not warn about compile-time overflow in constant expressions.
5868
5869       -Wno-odr
5870           Warn about One Definition Rule violations during link-time
5871           optimization.  Requires -flto-odr-type-merging to be enabled.
5872           Enabled by default.
5873
5874       -Wopenmp-simd
5875           Warn if the vectorizer cost model overrides the OpenMP simd
5876           directive set by user.  The -fsimd-cost-model=unlimited option can
5877           be used to relax the cost model.
5878
5879       -Woverride-init (C and Objective-C only)
5880           Warn if an initialized field without side effects is overridden
5881           when using designated initializers.
5882
5883           This warning is included in -Wextra.  To get other -Wextra warnings
5884           without this one, use -Wextra -Wno-override-init.
5885
5886       -Woverride-init-side-effects (C and Objective-C only)
5887           Warn if an initialized field with side effects is overridden when
5888           using designated initializers.  This warning is enabled by default.
5889
5890       -Wpacked
5891           Warn if a structure is given the packed attribute, but the packed
5892           attribute has no effect on the layout or size of the structure.
5893           Such structures may be mis-aligned for little benefit.  For
5894           instance, in this code, the variable "f.x" in "struct bar" is
5895           misaligned even though "struct bar" does not itself have the packed
5896           attribute:
5897
5898                   struct foo {
5899                     int x;
5900                     char a, b, c, d;
5901                   } __attribute__((packed));
5902                   struct bar {
5903                     char z;
5904                     struct foo f;
5905                   };
5906
5907       -Wpacked-bitfield-compat
5908           The 4.1, 4.2 and 4.3 series of GCC ignore the "packed" attribute on
5909           bit-fields of type "char".  This has been fixed in GCC 4.4 but the
5910           change can lead to differences in the structure layout.  GCC
5911           informs you when the offset of such a field has changed in GCC 4.4.
5912           For example there is no longer a 4-bit padding between field "a"
5913           and "b" in this structure:
5914
5915                   struct foo
5916                   {
5917                     char a:4;
5918                     char b:8;
5919                   } __attribute__ ((packed));
5920
5921           This warning is enabled by default.  Use
5922           -Wno-packed-bitfield-compat to disable this warning.
5923
5924       -Wpacked-not-aligned (C, C++, Objective-C and Objective-C++ only)
5925           Warn if a structure field with explicitly specified alignment in a
5926           packed struct or union is misaligned.  For example, a warning will
5927           be issued on "struct S", like, "warning: alignment 1 of 'struct S'
5928           is less than 8", in this code:
5929
5930                   struct __attribute__ ((aligned (8))) S8 { char a[8]; };
5931                   struct __attribute__ ((packed)) S {
5932                     struct S8 s8;
5933                   };
5934
5935           This warning is enabled by -Wall.
5936
5937       -Wpadded
5938           Warn if padding is included in a structure, either to align an
5939           element of the structure or to align the whole structure.
5940           Sometimes when this happens it is possible to rearrange the fields
5941           of the structure to reduce the padding and so make the structure
5942           smaller.
5943
5944       -Wredundant-decls
5945           Warn if anything is declared more than once in the same scope, even
5946           in cases where multiple declaration is valid and changes nothing.
5947
5948       -Wno-restrict
5949           Warn when an object referenced by a "restrict"-qualified parameter
5950           (or, in C++, a "__restrict"-qualified parameter) is aliased by
5951           another argument, or when copies between such objects overlap.  For
5952           example, the call to the "strcpy" function below attempts to
5953           truncate the string by replacing its initial characters with the
5954           last four.  However, because the call writes the terminating NUL
5955           into "a[4]", the copies overlap and the call is diagnosed.
5956
5957                   void foo (void)
5958                   {
5959                     char a[] = "abcd1234";
5960                     strcpy (a, a + 4);
5961                     ...
5962                   }
5963
5964           The -Wrestrict option detects some instances of simple overlap even
5965           without optimization but works best at -O2 and above.  It is
5966           included in -Wall.
5967
5968       -Wnested-externs (C and Objective-C only)
5969           Warn if an "extern" declaration is encountered within a function.
5970
5971       -Wno-inherited-variadic-ctor
5972           Suppress warnings about use of C++11 inheriting constructors when
5973           the base class inherited from has a C variadic constructor; the
5974           warning is on by default because the ellipsis is not inherited.
5975
5976       -Winline
5977           Warn if a function that is declared as inline cannot be inlined.
5978           Even with this option, the compiler does not warn about failures to
5979           inline functions declared in system headers.
5980
5981           The compiler uses a variety of heuristics to determine whether or
5982           not to inline a function.  For example, the compiler takes into
5983           account the size of the function being inlined and the amount of
5984           inlining that has already been done in the current function.
5985           Therefore, seemingly insignificant changes in the source program
5986           can cause the warnings produced by -Winline to appear or disappear.
5987
5988       -Wno-invalid-offsetof (C++ and Objective-C++ only)
5989           Suppress warnings from applying the "offsetof" macro to a non-POD
5990           type.  According to the 2014 ISO C++ standard, applying "offsetof"
5991           to a non-standard-layout type is undefined.  In existing C++
5992           implementations, however, "offsetof" typically gives meaningful
5993           results.  This flag is for users who are aware that they are
5994           writing nonportable code and who have deliberately chosen to ignore
5995           the warning about it.
5996
5997           The restrictions on "offsetof" may be relaxed in a future version
5998           of the C++ standard.
5999
6000       -Wint-in-bool-context
6001           Warn for suspicious use of integer values where boolean values are
6002           expected, such as conditional expressions (?:) using non-boolean
6003           integer constants in boolean context, like "if (a <= b ? 2 : 3)".
6004           Or left shifting of signed integers in boolean context, like "for
6005           (a = 0; 1 << a; a++);".  Likewise for all kinds of multiplications
6006           regardless of the data type.  This warning is enabled by -Wall.
6007
6008       -Wno-int-to-pointer-cast
6009           Suppress warnings from casts to pointer type of an integer of a
6010           different size. In C++, casting to a pointer type of smaller size
6011           is an error. Wint-to-pointer-cast is enabled by default.
6012
6013       -Wno-pointer-to-int-cast (C and Objective-C only)
6014           Suppress warnings from casts from a pointer to an integer type of a
6015           different size.
6016
6017       -Winvalid-pch
6018           Warn if a precompiled header is found in the search path but cannot
6019           be used.
6020
6021       -Wlong-long
6022           Warn if "long long" type is used.  This is enabled by either
6023           -Wpedantic or -Wtraditional in ISO C90 and C++98 modes.  To inhibit
6024           the warning messages, use -Wno-long-long.
6025
6026       -Wvariadic-macros
6027           Warn if variadic macros are used in ISO C90 mode, or if the GNU
6028           alternate syntax is used in ISO C99 mode.  This is enabled by
6029           either -Wpedantic or -Wtraditional.  To inhibit the warning
6030           messages, use -Wno-variadic-macros.
6031
6032       -Wvarargs
6033           Warn upon questionable usage of the macros used to handle variable
6034           arguments like "va_start".  This is default.  To inhibit the
6035           warning messages, use -Wno-varargs.
6036
6037       -Wvector-operation-performance
6038           Warn if vector operation is not implemented via SIMD capabilities
6039           of the architecture.  Mainly useful for the performance tuning.
6040           Vector operation can be implemented "piecewise", which means that
6041           the scalar operation is performed on every vector element; "in
6042           parallel", which means that the vector operation is implemented
6043           using scalars of wider type, which normally is more performance
6044           efficient; and "as a single scalar", which means that vector fits
6045           into a scalar type.
6046
6047       -Wno-virtual-move-assign
6048           Suppress warnings about inheriting from a virtual base with a non-
6049           trivial C++11 move assignment operator.  This is dangerous because
6050           if the virtual base is reachable along more than one path, it is
6051           moved multiple times, which can mean both objects end up in the
6052           moved-from state.  If the move assignment operator is written to
6053           avoid moving from a moved-from object, this warning can be
6054           disabled.
6055
6056       -Wvla
6057           Warn if a variable-length array is used in the code.  -Wno-vla
6058           prevents the -Wpedantic warning of the variable-length array.
6059
6060       -Wvla-larger-than=byte-size
6061           If this option is used, the compiler will warn for declarations of
6062           variable-length arrays whose size is either unbounded, or bounded
6063           by an argument that allows the array size to exceed byte-size
6064           bytes.  This is similar to how -Walloca-larger-than=byte-size
6065           works, but with variable-length arrays.
6066
6067           Note that GCC may optimize small variable-length arrays of a known
6068           value into plain arrays, so this warning may not get triggered for
6069           such arrays.
6070
6071           -Wvla-larger-than=PTRDIFF_MAX is enabled by default but is
6072           typically only effective when -ftree-vrp is active (default for -O2
6073           and above).
6074
6075           See also -Walloca-larger-than=byte-size.
6076
6077       -Wno-vla-larger-than
6078           Disable -Wvla-larger-than= warnings.  The option is equivalent to
6079           -Wvla-larger-than=SIZE_MAX or larger.
6080
6081       -Wvolatile-register-var
6082           Warn if a register variable is declared volatile.  The volatile
6083           modifier does not inhibit all optimizations that may eliminate
6084           reads and/or writes to register variables.  This warning is enabled
6085           by -Wall.
6086
6087       -Wdisabled-optimization
6088           Warn if a requested optimization pass is disabled.  This warning
6089           does not generally indicate that there is anything wrong with your
6090           code; it merely indicates that GCC's optimizers are unable to
6091           handle the code effectively.  Often, the problem is that your code
6092           is too big or too complex; GCC refuses to optimize programs when
6093           the optimization itself is likely to take inordinate amounts of
6094           time.
6095
6096       -Wpointer-sign (C and Objective-C only)
6097           Warn for pointer argument passing or assignment with different
6098           signedness.  This option is only supported for C and Objective-C.
6099           It is implied by -Wall and by -Wpedantic, which can be disabled
6100           with -Wno-pointer-sign.
6101
6102       -Wstack-protector
6103           This option is only active when -fstack-protector is active.  It
6104           warns about functions that are not protected against stack
6105           smashing.
6106
6107       -Woverlength-strings
6108           Warn about string constants that are longer than the "minimum
6109           maximum" length specified in the C standard.  Modern compilers
6110           generally allow string constants that are much longer than the
6111           standard's minimum limit, but very portable programs should avoid
6112           using longer strings.
6113
6114           The limit applies after string constant concatenation, and does not
6115           count the trailing NUL.  In C90, the limit was 509 characters; in
6116           C99, it was raised to 4095.  C++98 does not specify a normative
6117           minimum maximum, so we do not diagnose overlength strings in C++.
6118
6119           This option is implied by -Wpedantic, and can be disabled with
6120           -Wno-overlength-strings.
6121
6122       -Wunsuffixed-float-constants (C and Objective-C only)
6123           Issue a warning for any floating constant that does not have a
6124           suffix.  When used together with -Wsystem-headers it warns about
6125           such constants in system header files.  This can be useful when
6126           preparing code to use with the "FLOAT_CONST_DECIMAL64" pragma from
6127           the decimal floating-point extension to C99.
6128
6129       -Wno-designated-init (C and Objective-C only)
6130           Suppress warnings when a positional initializer is used to
6131           initialize a structure that has been marked with the
6132           "designated_init" attribute.
6133
6134       -Whsa
6135           Issue a warning when HSAIL cannot be emitted for the compiled
6136           function or OpenMP construct.
6137
6138   Options for Debugging Your Program
6139       To tell GCC to emit extra information for use by a debugger, in almost
6140       all cases you need only to add -g to your other options.
6141
6142       GCC allows you to use -g with -O.  The shortcuts taken by optimized
6143       code may occasionally be surprising: some variables you declared may
6144       not exist at all; flow of control may briefly move where you did not
6145       expect it; some statements may not be executed because they compute
6146       constant results or their values are already at hand; some statements
6147       may execute in different places because they have been moved out of
6148       loops.  Nevertheless it is possible to debug optimized output.  This
6149       makes it reasonable to use the optimizer for programs that might have
6150       bugs.
6151
6152       If you are not using some other optimization option, consider using -Og
6153       with -g.  With no -O option at all, some compiler passes that collect
6154       information useful for debugging do not run at all, so that -Og may
6155       result in a better debugging experience.
6156
6157       -g  Produce debugging information in the operating system's native
6158           format (stabs, COFF, XCOFF, or DWARF).  GDB can work with this
6159           debugging information.
6160
6161           On most systems that use stabs format, -g enables use of extra
6162           debugging information that only GDB can use; this extra information
6163           makes debugging work better in GDB but probably makes other
6164           debuggers crash or refuse to read the program.  If you want to
6165           control for certain whether to generate the extra information, use
6166           -gstabs+, -gstabs, -gxcoff+, -gxcoff, or -gvms (see below).
6167
6168       -ggdb
6169           Produce debugging information for use by GDB.  This means to use
6170           the most expressive format available (DWARF, stabs, or the native
6171           format if neither of those are supported), including GDB extensions
6172           if at all possible.
6173
6174       -gdwarf
6175       -gdwarf-version
6176           Produce debugging information in DWARF format (if that is
6177           supported).  The value of version may be either 2, 3, 4 or 5; the
6178           default version for most targets is 4.  DWARF Version 5 is only
6179           experimental.
6180
6181           Note that with DWARF Version 2, some ports require and always use
6182           some non-conflicting DWARF 3 extensions in the unwind tables.
6183
6184           Version 4 may require GDB 7.0 and -fvar-tracking-assignments for
6185           maximum benefit.
6186
6187           GCC no longer supports DWARF Version 1, which is substantially
6188           different than Version 2 and later.  For historical reasons, some
6189           other DWARF-related options such as -fno-dwarf2-cfi-asm) retain a
6190           reference to DWARF Version 2 in their names, but apply to all
6191           currently-supported versions of DWARF.
6192
6193       -gstabs
6194           Produce debugging information in stabs format (if that is
6195           supported), without GDB extensions.  This is the format used by DBX
6196           on most BSD systems.  On MIPS, Alpha and System V Release 4 systems
6197           this option produces stabs debugging output that is not understood
6198           by DBX.  On System V Release 4 systems this option requires the GNU
6199           assembler.
6200
6201       -gstabs+
6202           Produce debugging information in stabs format (if that is
6203           supported), using GNU extensions understood only by the GNU
6204           debugger (GDB).  The use of these extensions is likely to make
6205           other debuggers crash or refuse to read the program.
6206
6207       -gxcoff
6208           Produce debugging information in XCOFF format (if that is
6209           supported).  This is the format used by the DBX debugger on IBM
6210           RS/6000 systems.
6211
6212       -gxcoff+
6213           Produce debugging information in XCOFF format (if that is
6214           supported), using GNU extensions understood only by the GNU
6215           debugger (GDB).  The use of these extensions is likely to make
6216           other debuggers crash or refuse to read the program, and may cause
6217           assemblers other than the GNU assembler (GAS) to fail with an
6218           error.
6219
6220       -gvms
6221           Produce debugging information in Alpha/VMS debug format (if that is
6222           supported).  This is the format used by DEBUG on Alpha/VMS systems.
6223
6224       -glevel
6225       -ggdblevel
6226       -gstabslevel
6227       -gxcofflevel
6228       -gvmslevel
6229           Request debugging information and also use level to specify how
6230           much information.  The default level is 2.
6231
6232           Level 0 produces no debug information at all.  Thus, -g0 negates
6233           -g.
6234
6235           Level 1 produces minimal information, enough for making backtraces
6236           in parts of the program that you don't plan to debug.  This
6237           includes descriptions of functions and external variables, and line
6238           number tables, but no information about local variables.
6239
6240           Level 3 includes extra information, such as all the macro
6241           definitions present in the program.  Some debuggers support macro
6242           expansion when you use -g3.
6243
6244           If you use multiple -g options, with or without level numbers, the
6245           last such option is the one that is effective.
6246
6247           -gdwarf does not accept a concatenated debug level, to avoid
6248           confusion with -gdwarf-level.  Instead use an additional -glevel
6249           option to change the debug level for DWARF.
6250
6251       -feliminate-unused-debug-symbols
6252           Produce debugging information in stabs format (if that is
6253           supported), for only symbols that are actually used.
6254
6255       -femit-class-debug-always
6256           Instead of emitting debugging information for a C++ class in only
6257           one object file, emit it in all object files using the class.  This
6258           option should be used only with debuggers that are unable to handle
6259           the way GCC normally emits debugging information for classes
6260           because using this option increases the size of debugging
6261           information by as much as a factor of two.
6262
6263       -fno-merge-debug-strings
6264           Direct the linker to not merge together strings in the debugging
6265           information that are identical in different object files.  Merging
6266           is not supported by all assemblers or linkers.  Merging decreases
6267           the size of the debug information in the output file at the cost of
6268           increasing link processing time.  Merging is enabled by default.
6269
6270       -fdebug-prefix-map=old=new
6271           When compiling files residing in directory old, record debugging
6272           information describing them as if the files resided in directory
6273           new instead.  This can be used to replace a build-time path with an
6274           install-time path in the debug info.  It can also be used to change
6275           an absolute path to a relative path by using . for new.  This can
6276           give more reproducible builds, which are location independent, but
6277           may require an extra command to tell GDB where to find the source
6278           files. See also -ffile-prefix-map.
6279
6280       -fvar-tracking
6281           Run variable tracking pass.  It computes where variables are stored
6282           at each position in code.  Better debugging information is then
6283           generated (if the debugging information format supports this
6284           information).
6285
6286           It is enabled by default when compiling with optimization (-Os, -O,
6287           -O2, ...), debugging information (-g) and the debug info format
6288           supports it.
6289
6290       -fvar-tracking-assignments
6291           Annotate assignments to user variables early in the compilation and
6292           attempt to carry the annotations over throughout the compilation
6293           all the way to the end, in an attempt to improve debug information
6294           while optimizing.  Use of -gdwarf-4 is recommended along with it.
6295
6296           It can be enabled even if var-tracking is disabled, in which case
6297           annotations are created and maintained, but discarded at the end.
6298           By default, this flag is enabled together with -fvar-tracking,
6299           except when selective scheduling is enabled.
6300
6301       -gsplit-dwarf
6302           Separate as much DWARF debugging information as possible into a
6303           separate output file with the extension .dwo.  This option allows
6304           the build system to avoid linking files with debug information.  To
6305           be useful, this option requires a debugger capable of reading .dwo
6306           files.
6307
6308       -gdescribe-dies
6309           Add description attributes to some DWARF DIEs that have no name
6310           attribute, such as artificial variables, external references and
6311           call site parameter DIEs.
6312
6313       -gpubnames
6314           Generate DWARF ".debug_pubnames" and ".debug_pubtypes" sections.
6315
6316       -ggnu-pubnames
6317           Generate ".debug_pubnames" and ".debug_pubtypes" sections in a
6318           format suitable for conversion into a GDB index.  This option is
6319           only useful with a linker that can produce GDB index version 7.
6320
6321       -fdebug-types-section
6322           When using DWARF Version 4 or higher, type DIEs can be put into
6323           their own ".debug_types" section instead of making them part of the
6324           ".debug_info" section.  It is more efficient to put them in a
6325           separate comdat section since the linker can then remove
6326           duplicates.  But not all DWARF consumers support ".debug_types"
6327           sections yet and on some objects ".debug_types" produces larger
6328           instead of smaller debugging information.
6329
6330       -grecord-gcc-switches
6331       -gno-record-gcc-switches
6332           This switch causes the command-line options used to invoke the
6333           compiler that may affect code generation to be appended to the
6334           DW_AT_producer attribute in DWARF debugging information.  The
6335           options are concatenated with spaces separating them from each
6336           other and from the compiler version.  It is enabled by default.
6337           See also -frecord-gcc-switches for another way of storing compiler
6338           options into the object file.
6339
6340       -gstrict-dwarf
6341           Disallow using extensions of later DWARF standard version than
6342           selected with -gdwarf-version.  On most targets using non-
6343           conflicting DWARF extensions from later standard versions is
6344           allowed.
6345
6346       -gno-strict-dwarf
6347           Allow using extensions of later DWARF standard version than
6348           selected with -gdwarf-version.
6349
6350       -gas-loc-support
6351           Inform the compiler that the assembler supports ".loc" directives.
6352           It may then use them for the assembler to generate DWARF2+ line
6353           number tables.
6354
6355           This is generally desirable, because assembler-generated line-
6356           number tables are a lot more compact than those the compiler can
6357           generate itself.
6358
6359           This option will be enabled by default if, at GCC configure time,
6360           the assembler was found to support such directives.
6361
6362       -gno-as-loc-support
6363           Force GCC to generate DWARF2+ line number tables internally, if
6364           DWARF2+ line number tables are to be generated.
6365
6366       gas-locview-support
6367           Inform the compiler that the assembler supports "view" assignment
6368           and reset assertion checking in ".loc" directives.
6369
6370           This option will be enabled by default if, at GCC configure time,
6371           the assembler was found to support them.
6372
6373       gno-as-locview-support
6374           Force GCC to assign view numbers internally, if
6375           -gvariable-location-views are explicitly requested.
6376
6377       -gcolumn-info
6378       -gno-column-info
6379           Emit location column information into DWARF debugging information,
6380           rather than just file and line.  This option is enabled by default.
6381
6382       -gstatement-frontiers
6383       -gno-statement-frontiers
6384           This option causes GCC to create markers in the internal
6385           representation at the beginning of statements, and to keep them
6386           roughly in place throughout compilation, using them to guide the
6387           output of "is_stmt" markers in the line number table.  This is
6388           enabled by default when compiling with optimization (-Os, -O, -O2,
6389           ...), and outputting DWARF 2 debug information at the normal level.
6390
6391       -gvariable-location-views
6392       -gvariable-location-views=incompat5
6393       -gno-variable-location-views
6394           Augment variable location lists with progressive view numbers
6395           implied from the line number table.  This enables debug information
6396           consumers to inspect state at certain points of the program, even
6397           if no instructions associated with the corresponding source
6398           locations are present at that point.  If the assembler lacks
6399           support for view numbers in line number tables, this will cause the
6400           compiler to emit the line number table, which generally makes them
6401           somewhat less compact.  The augmented line number tables and
6402           location lists are fully backward-compatible, so they can be
6403           consumed by debug information consumers that are not aware of these
6404           augmentations, but they won't derive any benefit from them either.
6405
6406           This is enabled by default when outputting DWARF 2 debug
6407           information at the normal level, as long as there is assembler
6408           support, -fvar-tracking-assignments is enabled and -gstrict-dwarf
6409           is not.  When assembler support is not available, this may still be
6410           enabled, but it will force GCC to output internal line number
6411           tables, and if -ginternal-reset-location-views is not enabled, that
6412           will most certainly lead to silently mismatching location views.
6413
6414           There is a proposed representation for view numbers that is not
6415           backward compatible with the location list format introduced in
6416           DWARF 5, that can be enabled with
6417           -gvariable-location-views=incompat5.  This option may be removed in
6418           the future, is only provided as a reference implementation of the
6419           proposed representation.  Debug information consumers are not
6420           expected to support this extended format, and they would be
6421           rendered unable to decode location lists using it.
6422
6423       -ginternal-reset-location-views
6424       -gnointernal-reset-location-views
6425           Attempt to determine location views that can be omitted from
6426           location view lists.  This requires the compiler to have very
6427           accurate insn length estimates, which isn't always the case, and it
6428           may cause incorrect view lists to be generated silently when using
6429           an assembler that does not support location view lists.  The GNU
6430           assembler will flag any such error as a "view number mismatch".
6431           This is only enabled on ports that define a reliable estimation
6432           function.
6433
6434       -ginline-points
6435       -gno-inline-points
6436           Generate extended debug information for inlined functions.
6437           Location view tracking markers are inserted at inlined entry
6438           points, so that address and view numbers can be computed and output
6439           in debug information.  This can be enabled independently of
6440           location views, in which case the view numbers won't be output, but
6441           it can only be enabled along with statement frontiers, and it is
6442           only enabled by default if location views are enabled.
6443
6444       -gz[=type]
6445           Produce compressed debug sections in DWARF format, if that is
6446           supported.  If type is not given, the default type depends on the
6447           capabilities of the assembler and linker used.  type may be one of
6448           none (don't compress debug sections), zlib (use zlib compression in
6449           ELF gABI format), or zlib-gnu (use zlib compression in traditional
6450           GNU format).  If the linker doesn't support writing compressed
6451           debug sections, the option is rejected.  Otherwise, if the
6452           assembler does not support them, -gz is silently ignored when
6453           producing object files.
6454
6455       -femit-struct-debug-baseonly
6456           Emit debug information for struct-like types only when the base
6457           name of the compilation source file matches the base name of file
6458           in which the struct is defined.
6459
6460           This option substantially reduces the size of debugging
6461           information, but at significant potential loss in type information
6462           to the debugger.  See -femit-struct-debug-reduced for a less
6463           aggressive option.  See -femit-struct-debug-detailed for more
6464           detailed control.
6465
6466           This option works only with DWARF debug output.
6467
6468       -femit-struct-debug-reduced
6469           Emit debug information for struct-like types only when the base
6470           name of the compilation source file matches the base name of file
6471           in which the type is defined, unless the struct is a template or
6472           defined in a system header.
6473
6474           This option significantly reduces the size of debugging
6475           information, with some potential loss in type information to the
6476           debugger.  See -femit-struct-debug-baseonly for a more aggressive
6477           option.  See -femit-struct-debug-detailed for more detailed
6478           control.
6479
6480           This option works only with DWARF debug output.
6481
6482       -femit-struct-debug-detailed[=spec-list]
6483           Specify the struct-like types for which the compiler generates
6484           debug information.  The intent is to reduce duplicate struct debug
6485           information between different object files within the same program.
6486
6487           This option is a detailed version of -femit-struct-debug-reduced
6488           and -femit-struct-debug-baseonly, which serves for most needs.
6489
6490           A specification has the
6491           syntax[dir:|ind:][ord:|gen:](any|sys|base|none)
6492
6493           The optional first word limits the specification to structs that
6494           are used directly (dir:) or used indirectly (ind:).  A struct type
6495           is used directly when it is the type of a variable, member.
6496           Indirect uses arise through pointers to structs.  That is, when use
6497           of an incomplete struct is valid, the use is indirect.  An example
6498           is struct one direct; struct two * indirect;.
6499
6500           The optional second word limits the specification to ordinary
6501           structs (ord:) or generic structs (gen:).  Generic structs are a
6502           bit complicated to explain.  For C++, these are non-explicit
6503           specializations of template classes, or non-template classes within
6504           the above.  Other programming languages have generics, but
6505           -femit-struct-debug-detailed does not yet implement them.
6506
6507           The third word specifies the source files for those structs for
6508           which the compiler should emit debug information.  The values none
6509           and any have the normal meaning.  The value base means that the
6510           base of name of the file in which the type declaration appears must
6511           match the base of the name of the main compilation file.  In
6512           practice, this means that when compiling foo.c, debug information
6513           is generated for types declared in that file and foo.h, but not
6514           other header files.  The value sys means those types satisfying
6515           base or declared in system or compiler headers.
6516
6517           You may need to experiment to determine the best settings for your
6518           application.
6519
6520           The default is -femit-struct-debug-detailed=all.
6521
6522           This option works only with DWARF debug output.
6523
6524       -fno-dwarf2-cfi-asm
6525           Emit DWARF unwind info as compiler generated ".eh_frame" section
6526           instead of using GAS ".cfi_*" directives.
6527
6528       -fno-eliminate-unused-debug-types
6529           Normally, when producing DWARF output, GCC avoids producing debug
6530           symbol output for types that are nowhere used in the source file
6531           being compiled.  Sometimes it is useful to have GCC emit debugging
6532           information for all types declared in a compilation unit,
6533           regardless of whether or not they are actually used in that
6534           compilation unit, for example if, in the debugger, you want to cast
6535           a value to a type that is not actually used in your program (but is
6536           declared).  More often, however, this results in a significant
6537           amount of wasted space.
6538
6539   Options That Control Optimization
6540       These options control various sorts of optimizations.
6541
6542       Without any optimization option, the compiler's goal is to reduce the
6543       cost of compilation and to make debugging produce the expected results.
6544       Statements are independent: if you stop the program with a breakpoint
6545       between statements, you can then assign a new value to any variable or
6546       change the program counter to any other statement in the function and
6547       get exactly the results you expect from the source code.
6548
6549       Turning on optimization flags makes the compiler attempt to improve the
6550       performance and/or code size at the expense of compilation time and
6551       possibly the ability to debug the program.
6552
6553       The compiler performs optimization based on the knowledge it has of the
6554       program.  Compiling multiple files at once to a single output file mode
6555       allows the compiler to use information gained from all of the files
6556       when compiling each of them.
6557
6558       Not all optimizations are controlled directly by a flag.  Only
6559       optimizations that have a flag are listed in this section.
6560
6561       Most optimizations are completely disabled at -O0 or if an -O level is
6562       not set on the command line, even if individual optimization flags are
6563       specified.  Similarly, -Og suppresses many optimization passes.
6564
6565       Depending on the target and how GCC was configured, a slightly
6566       different set of optimizations may be enabled at each -O level than
6567       those listed here.  You can invoke GCC with -Q --help=optimizers to
6568       find out the exact set of optimizations that are enabled at each level.
6569
6570       -O
6571       -O1 Optimize.  Optimizing compilation takes somewhat more time, and a
6572           lot more memory for a large function.
6573
6574           With -O, the compiler tries to reduce code size and execution time,
6575           without performing any optimizations that take a great deal of
6576           compilation time.
6577
6578           -O turns on the following optimization flags:
6579
6580           -fauto-inc-dec -fbranch-count-reg -fcombine-stack-adjustments
6581           -fcompare-elim -fcprop-registers -fdce -fdefer-pop -fdelayed-branch
6582           -fdse -fforward-propagate -fguess-branch-probability
6583           -fif-conversion -fif-conversion2 -finline-functions-called-once
6584           -fipa-profile -fipa-pure-const -fipa-reference
6585           -fipa-reference-addressable -fmerge-constants
6586           -fmove-loop-invariants -fomit-frame-pointer -freorder-blocks
6587           -fshrink-wrap -fshrink-wrap-separate -fsplit-wide-types
6588           -fssa-backprop -fssa-phiopt -ftree-bit-ccp -ftree-ccp -ftree-ch
6589           -ftree-coalesce-vars -ftree-copy-prop -ftree-dce
6590           -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre
6591           -ftree-phiprop -ftree-pta -ftree-scev-cprop -ftree-sink -ftree-slsr
6592           -ftree-sra -ftree-ter -funit-at-a-time
6593
6594       -O2 Optimize even more.  GCC performs nearly all supported
6595           optimizations that do not involve a space-speed tradeoff.  As
6596           compared to -O, this option increases both compilation time and the
6597           performance of the generated code.
6598
6599           -O2 turns on all optimization flags specified by -O.  It also turns
6600           on the following optimization flags:
6601
6602           -falign-functions  -falign-jumps -falign-labels  -falign-loops
6603           -fcaller-saves -fcode-hoisting -fcrossjumping -fcse-follow-jumps
6604           -fcse-skip-blocks -fdelete-null-pointer-checks -fdevirtualize
6605           -fdevirtualize-speculatively -fexpensive-optimizations -fgcse
6606           -fgcse-lm -fhoist-adjacent-loads -finline-small-functions
6607           -findirect-inlining -fipa-bit-cp  -fipa-cp  -fipa-icf -fipa-ra
6608           -fipa-sra  -fipa-vrp -fisolate-erroneous-paths-dereference
6609           -flra-remat -foptimize-sibling-calls -foptimize-strlen
6610           -fpartial-inlining -fpeephole2 -freorder-blocks-algorithm=stc
6611           -freorder-blocks-and-partition  -freorder-functions
6612           -frerun-cse-after-loop -fschedule-insns  -fschedule-insns2
6613           -fsched-interblock  -fsched-spec -fstore-merging -fstrict-aliasing
6614           -fthread-jumps -ftree-builtin-call-dce -ftree-pre
6615           -ftree-switch-conversion  -ftree-tail-merge -ftree-vrp
6616
6617           Please note the warning under -fgcse about invoking -O2 on programs
6618           that use computed gotos.
6619
6620       -O3 Optimize yet more.  -O3 turns on all optimizations specified by -O2
6621           and also turns on the following optimization flags:
6622
6623           -fgcse-after-reload -finline-functions -fipa-cp-clone
6624           -floop-interchange -floop-unroll-and-jam -fpeel-loops
6625           -fpredictive-commoning -fsplit-paths
6626           -ftree-loop-distribute-patterns -ftree-loop-distribution
6627           -ftree-loop-vectorize -ftree-partial-pre -ftree-slp-vectorize
6628           -funswitch-loops -fvect-cost-model -fversion-loops-for-strides
6629
6630       -O0 Reduce compilation time and make debugging produce the expected
6631           results.  This is the default.
6632
6633       -Os Optimize for size.  -Os enables all -O2 optimizations except those
6634           that often increase code size:
6635
6636           -falign-functions  -falign-jumps -falign-labels  -falign-loops
6637           -fprefetch-loop-arrays  -freorder-blocks-algorithm=stc
6638
6639           It also enables -finline-functions, causes the compiler to tune for
6640           code size rather than execution speed, and performs further
6641           optimizations designed to reduce code size.
6642
6643       -Ofast
6644           Disregard strict standards compliance.  -Ofast enables all -O3
6645           optimizations.  It also enables optimizations that are not valid
6646           for all standard-compliant programs.  It turns on -ffast-math and
6647           the Fortran-specific -fstack-arrays, unless -fmax-stack-var-size is
6648           specified, and -fno-protect-parens.
6649
6650       -Og Optimize debugging experience.  -Og should be the optimization
6651           level of choice for the standard edit-compile-debug cycle, offering
6652           a reasonable level of optimization while maintaining fast
6653           compilation and a good debugging experience.  It is a better choice
6654           than -O0 for producing debuggable code because some compiler passes
6655           that collect debug information are disabled at -O0.
6656
6657           Like -O0, -Og completely disables a number of optimization passes
6658           so that individual options controlling them have no effect.
6659           Otherwise -Og enables all -O1 optimization flags except for those
6660           that may interfere with debugging:
6661
6662           -fbranch-count-reg  -fdelayed-branch -fif-conversion
6663           -fif-conversion2 -finline-functions-called-once
6664           -fmove-loop-invariants  -fssa-phiopt -ftree-bit-ccp  -ftree-pta
6665           -ftree-sra
6666
6667       If you use multiple -O options, with or without level numbers, the last
6668       such option is the one that is effective.
6669
6670       Options of the form -fflag specify machine-independent flags.  Most
6671       flags have both positive and negative forms; the negative form of -ffoo
6672       is -fno-foo.  In the table below, only one of the forms is listed---the
6673       one you typically use.  You can figure out the other form by either
6674       removing no- or adding it.
6675
6676       The following options control specific optimizations.  They are either
6677       activated by -O options or are related to ones that are.  You can use
6678       the following flags in the rare cases when "fine-tuning" of
6679       optimizations to be performed is desired.
6680
6681       -fno-defer-pop
6682           For machines that must pop arguments after a function call, always
6683           pop the arguments as soon as each function returns.  At levels -O1
6684           and higher, -fdefer-pop is the default; this allows the compiler to
6685           let arguments accumulate on the stack for several function calls
6686           and pop them all at once.
6687
6688       -fforward-propagate
6689           Perform a forward propagation pass on RTL.  The pass tries to
6690           combine two instructions and checks if the result can be
6691           simplified.  If loop unrolling is active, two passes are performed
6692           and the second is scheduled after loop unrolling.
6693
6694           This option is enabled by default at optimization levels -O, -O2,
6695           -O3, -Os.
6696
6697       -ffp-contract=style
6698           -ffp-contract=off disables floating-point expression contraction.
6699           -ffp-contract=fast enables floating-point expression contraction
6700           such as forming of fused multiply-add operations if the target has
6701           native support for them.  -ffp-contract=on enables floating-point
6702           expression contraction if allowed by the language standard.  This
6703           is currently not implemented and treated equal to
6704           -ffp-contract=off.
6705
6706           The default is -ffp-contract=fast.
6707
6708       -fomit-frame-pointer
6709           Omit the frame pointer in functions that don't need one.  This
6710           avoids the instructions to save, set up and restore the frame
6711           pointer; on many targets it also makes an extra register available.
6712
6713           On some targets this flag has no effect because the standard
6714           calling sequence always uses a frame pointer, so it cannot be
6715           omitted.
6716
6717           Note that -fno-omit-frame-pointer doesn't guarantee the frame
6718           pointer is used in all functions.  Several targets always omit the
6719           frame pointer in leaf functions.
6720
6721           Enabled by default at -O and higher.
6722
6723       -foptimize-sibling-calls
6724           Optimize sibling and tail recursive calls.
6725
6726           Enabled at levels -O2, -O3, -Os.
6727
6728       -foptimize-strlen
6729           Optimize various standard C string functions (e.g. "strlen",
6730           "strchr" or "strcpy") and their "_FORTIFY_SOURCE" counterparts into
6731           faster alternatives.
6732
6733           Enabled at levels -O2, -O3.
6734
6735       -fno-inline
6736           Do not expand any functions inline apart from those marked with the
6737           "always_inline" attribute.  This is the default when not
6738           optimizing.
6739
6740           Single functions can be exempted from inlining by marking them with
6741           the "noinline" attribute.
6742
6743       -finline-small-functions
6744           Integrate functions into their callers when their body is smaller
6745           than expected function call code (so overall size of program gets
6746           smaller).  The compiler heuristically decides which functions are
6747           simple enough to be worth integrating in this way.  This inlining
6748           applies to all functions, even those not declared inline.
6749
6750           Enabled at levels -O2, -O3, -Os.
6751
6752       -findirect-inlining
6753           Inline also indirect calls that are discovered to be known at
6754           compile time thanks to previous inlining.  This option has any
6755           effect only when inlining itself is turned on by the
6756           -finline-functions or -finline-small-functions options.
6757
6758           Enabled at levels -O2, -O3, -Os.
6759
6760       -finline-functions
6761           Consider all functions for inlining, even if they are not declared
6762           inline.  The compiler heuristically decides which functions are
6763           worth integrating in this way.
6764
6765           If all calls to a given function are integrated, and the function
6766           is declared "static", then the function is normally not output as
6767           assembler code in its own right.
6768
6769           Enabled at levels -O3, -Os.  Also enabled by -fprofile-use and
6770           -fauto-profile.
6771
6772       -finline-functions-called-once
6773           Consider all "static" functions called once for inlining into their
6774           caller even if they are not marked "inline".  If a call to a given
6775           function is integrated, then the function is not output as
6776           assembler code in its own right.
6777
6778           Enabled at levels -O1, -O2, -O3 and -Os, but not -Og.
6779
6780       -fearly-inlining
6781           Inline functions marked by "always_inline" and functions whose body
6782           seems smaller than the function call overhead early before doing
6783           -fprofile-generate instrumentation and real inlining pass.  Doing
6784           so makes profiling significantly cheaper and usually inlining
6785           faster on programs having large chains of nested wrapper functions.
6786
6787           Enabled by default.
6788
6789       -fipa-sra
6790           Perform interprocedural scalar replacement of aggregates, removal
6791           of unused parameters and replacement of parameters passed by
6792           reference by parameters passed by value.
6793
6794           Enabled at levels -O2, -O3 and -Os.
6795
6796       -finline-limit=n
6797           By default, GCC limits the size of functions that can be inlined.
6798           This flag allows coarse control of this limit.  n is the size of
6799           functions that can be inlined in number of pseudo instructions.
6800
6801           Inlining is actually controlled by a number of parameters, which
6802           may be specified individually by using --param name=value.  The
6803           -finline-limit=n option sets some of these parameters as follows:
6804
6805           max-inline-insns-single
6806               is set to n/2.
6807
6808           max-inline-insns-auto
6809               is set to n/2.
6810
6811           See below for a documentation of the individual parameters
6812           controlling inlining and for the defaults of these parameters.
6813
6814           Note: there may be no value to -finline-limit that results in
6815           default behavior.
6816
6817           Note: pseudo instruction represents, in this particular context, an
6818           abstract measurement of function's size.  In no way does it
6819           represent a count of assembly instructions and as such its exact
6820           meaning might change from one release to an another.
6821
6822       -fno-keep-inline-dllexport
6823           This is a more fine-grained version of -fkeep-inline-functions,
6824           which applies only to functions that are declared using the
6825           "dllexport" attribute or declspec.
6826
6827       -fkeep-inline-functions
6828           In C, emit "static" functions that are declared "inline" into the
6829           object file, even if the function has been inlined into all of its
6830           callers.  This switch does not affect functions using the "extern
6831           inline" extension in GNU C90.  In C++, emit any and all inline
6832           functions into the object file.
6833
6834       -fkeep-static-functions
6835           Emit "static" functions into the object file, even if the function
6836           is never used.
6837
6838       -fkeep-static-consts
6839           Emit variables declared "static const" when optimization isn't
6840           turned on, even if the variables aren't referenced.
6841
6842           GCC enables this option by default.  If you want to force the
6843           compiler to check if a variable is referenced, regardless of
6844           whether or not optimization is turned on, use the
6845           -fno-keep-static-consts option.
6846
6847       -fmerge-constants
6848           Attempt to merge identical constants (string constants and
6849           floating-point constants) across compilation units.
6850
6851           This option is the default for optimized compilation if the
6852           assembler and linker support it.  Use -fno-merge-constants to
6853           inhibit this behavior.
6854
6855           Enabled at levels -O, -O2, -O3, -Os.
6856
6857       -fmerge-all-constants
6858           Attempt to merge identical constants and identical variables.
6859
6860           This option implies -fmerge-constants.  In addition to
6861           -fmerge-constants this considers e.g. even constant initialized
6862           arrays or initialized constant variables with integral or floating-
6863           point types.  Languages like C or C++ require each variable,
6864           including multiple instances of the same variable in recursive
6865           calls, to have distinct locations, so using this option results in
6866           non-conforming behavior.
6867
6868       -fmodulo-sched
6869           Perform swing modulo scheduling immediately before the first
6870           scheduling pass.  This pass looks at innermost loops and reorders
6871           their instructions by overlapping different iterations.
6872
6873       -fmodulo-sched-allow-regmoves
6874           Perform more aggressive SMS-based modulo scheduling with register
6875           moves allowed.  By setting this flag certain anti-dependences edges
6876           are deleted, which triggers the generation of reg-moves based on
6877           the life-range analysis.  This option is effective only with
6878           -fmodulo-sched enabled.
6879
6880       -fno-branch-count-reg
6881           Disable the optimization pass that scans for opportunities to use
6882           "decrement and branch" instructions on a count register instead of
6883           instruction sequences that decrement a register, compare it against
6884           zero, and then branch based upon the result.  This option is only
6885           meaningful on architectures that support such instructions, which
6886           include x86, PowerPC, IA-64 and S/390.  Note that the
6887           -fno-branch-count-reg option doesn't remove the decrement and
6888           branch instructions from the generated instruction stream
6889           introduced by other optimization passes.
6890
6891           The default is -fbranch-count-reg at -O1 and higher, except for
6892           -Og.
6893
6894       -fno-function-cse
6895           Do not put function addresses in registers; make each instruction
6896           that calls a constant function contain the function's address
6897           explicitly.
6898
6899           This option results in less efficient code, but some strange hacks
6900           that alter the assembler output may be confused by the
6901           optimizations performed when this option is not used.
6902
6903           The default is -ffunction-cse
6904
6905       -fno-zero-initialized-in-bss
6906           If the target supports a BSS section, GCC by default puts variables
6907           that are initialized to zero into BSS.  This can save space in the
6908           resulting code.
6909
6910           This option turns off this behavior because some programs
6911           explicitly rely on variables going to the data section---e.g., so
6912           that the resulting executable can find the beginning of that
6913           section and/or make assumptions based on that.
6914
6915           The default is -fzero-initialized-in-bss.
6916
6917       -fthread-jumps
6918           Perform optimizations that check to see if a jump branches to a
6919           location where another comparison subsumed by the first is found.
6920           If so, the first branch is redirected to either the destination of
6921           the second branch or a point immediately following it, depending on
6922           whether the condition is known to be true or false.
6923
6924           Enabled at levels -O2, -O3, -Os.
6925
6926       -fsplit-wide-types
6927           When using a type that occupies multiple registers, such as "long
6928           long" on a 32-bit system, split the registers apart and allocate
6929           them independently.  This normally generates better code for those
6930           types, but may make debugging more difficult.
6931
6932           Enabled at levels -O, -O2, -O3, -Os.
6933
6934       -fcse-follow-jumps
6935           In common subexpression elimination (CSE), scan through jump
6936           instructions when the target of the jump is not reached by any
6937           other path.  For example, when CSE encounters an "if" statement
6938           with an "else" clause, CSE follows the jump when the condition
6939           tested is false.
6940
6941           Enabled at levels -O2, -O3, -Os.
6942
6943       -fcse-skip-blocks
6944           This is similar to -fcse-follow-jumps, but causes CSE to follow
6945           jumps that conditionally skip over blocks.  When CSE encounters a
6946           simple "if" statement with no else clause, -fcse-skip-blocks causes
6947           CSE to follow the jump around the body of the "if".
6948
6949           Enabled at levels -O2, -O3, -Os.
6950
6951       -frerun-cse-after-loop
6952           Re-run common subexpression elimination after loop optimizations
6953           are performed.
6954
6955           Enabled at levels -O2, -O3, -Os.
6956
6957       -fgcse
6958           Perform a global common subexpression elimination pass.  This pass
6959           also performs global constant and copy propagation.
6960
6961           Note: When compiling a program using computed gotos, a GCC
6962           extension, you may get better run-time performance if you disable
6963           the global common subexpression elimination pass by adding
6964           -fno-gcse to the command line.
6965
6966           Enabled at levels -O2, -O3, -Os.
6967
6968       -fgcse-lm
6969           When -fgcse-lm is enabled, global common subexpression elimination
6970           attempts to move loads that are only killed by stores into
6971           themselves.  This allows a loop containing a load/store sequence to
6972           be changed to a load outside the loop, and a copy/store within the
6973           loop.
6974
6975           Enabled by default when -fgcse is enabled.
6976
6977       -fgcse-sm
6978           When -fgcse-sm is enabled, a store motion pass is run after global
6979           common subexpression elimination.  This pass attempts to move
6980           stores out of loops.  When used in conjunction with -fgcse-lm,
6981           loops containing a load/store sequence can be changed to a load
6982           before the loop and a store after the loop.
6983
6984           Not enabled at any optimization level.
6985
6986       -fgcse-las
6987           When -fgcse-las is enabled, the global common subexpression
6988           elimination pass eliminates redundant loads that come after stores
6989           to the same memory location (both partial and full redundancies).
6990
6991           Not enabled at any optimization level.
6992
6993       -fgcse-after-reload
6994           When -fgcse-after-reload is enabled, a redundant load elimination
6995           pass is performed after reload.  The purpose of this pass is to
6996           clean up redundant spilling.
6997
6998           Enabled by -fprofile-use and -fauto-profile.
6999
7000       -faggressive-loop-optimizations
7001           This option tells the loop optimizer to use language constraints to
7002           derive bounds for the number of iterations of a loop.  This assumes
7003           that loop code does not invoke undefined behavior by for example
7004           causing signed integer overflows or out-of-bound array accesses.
7005           The bounds for the number of iterations of a loop are used to guide
7006           loop unrolling and peeling and loop exit test optimizations.  This
7007           option is enabled by default.
7008
7009       -funconstrained-commons
7010           This option tells the compiler that variables declared in common
7011           blocks (e.g. Fortran) may later be overridden with longer trailing
7012           arrays. This prevents certain optimizations that depend on knowing
7013           the array bounds.
7014
7015       -fcrossjumping
7016           Perform cross-jumping transformation.  This transformation unifies
7017           equivalent code and saves code size.  The resulting code may or may
7018           not perform better than without cross-jumping.
7019
7020           Enabled at levels -O2, -O3, -Os.
7021
7022       -fauto-inc-dec
7023           Combine increments or decrements of addresses with memory accesses.
7024           This pass is always skipped on architectures that do not have
7025           instructions to support this.  Enabled by default at -O and higher
7026           on architectures that support this.
7027
7028       -fdce
7029           Perform dead code elimination (DCE) on RTL.  Enabled by default at
7030           -O and higher.
7031
7032       -fdse
7033           Perform dead store elimination (DSE) on RTL.  Enabled by default at
7034           -O and higher.
7035
7036       -fif-conversion
7037           Attempt to transform conditional jumps into branch-less
7038           equivalents.  This includes use of conditional moves, min, max, set
7039           flags and abs instructions, and some tricks doable by standard
7040           arithmetics.  The use of conditional execution on chips where it is
7041           available is controlled by -fif-conversion2.
7042
7043           Enabled at levels -O, -O2, -O3, -Os, but not with -Og.
7044
7045       -fif-conversion2
7046           Use conditional execution (where available) to transform
7047           conditional jumps into branch-less equivalents.
7048
7049           Enabled at levels -O, -O2, -O3, -Os, but not with -Og.
7050
7051       -fdeclone-ctor-dtor
7052           The C++ ABI requires multiple entry points for constructors and
7053           destructors: one for a base subobject, one for a complete object,
7054           and one for a virtual destructor that calls operator delete
7055           afterwards.  For a hierarchy with virtual bases, the base and
7056           complete variants are clones, which means two copies of the
7057           function.  With this option, the base and complete variants are
7058           changed to be thunks that call a common implementation.
7059
7060           Enabled by -Os.
7061
7062       -fdelete-null-pointer-checks
7063           Assume that programs cannot safely dereference null pointers, and
7064           that no code or data element resides at address zero.  This option
7065           enables simple constant folding optimizations at all optimization
7066           levels.  In addition, other optimization passes in GCC use this
7067           flag to control global dataflow analyses that eliminate useless
7068           checks for null pointers; these assume that a memory access to
7069           address zero always results in a trap, so that if a pointer is
7070           checked after it has already been dereferenced, it cannot be null.
7071
7072           Note however that in some environments this assumption is not true.
7073           Use -fno-delete-null-pointer-checks to disable this optimization
7074           for programs that depend on that behavior.
7075
7076           This option is enabled by default on most targets.  On Nios II ELF,
7077           it defaults to off.  On AVR, CR16, and MSP430, this option is
7078           completely disabled.
7079
7080           Passes that use the dataflow information are enabled independently
7081           at different optimization levels.
7082
7083       -fdevirtualize
7084           Attempt to convert calls to virtual functions to direct calls.
7085           This is done both within a procedure and interprocedurally as part
7086           of indirect inlining (-findirect-inlining) and interprocedural
7087           constant propagation (-fipa-cp).  Enabled at levels -O2, -O3, -Os.
7088
7089       -fdevirtualize-speculatively
7090           Attempt to convert calls to virtual functions to speculative direct
7091           calls.  Based on the analysis of the type inheritance graph,
7092           determine for a given call the set of likely targets. If the set is
7093           small, preferably of size 1, change the call into a conditional
7094           deciding between direct and indirect calls.  The speculative calls
7095           enable more optimizations, such as inlining.  When they seem
7096           useless after further optimization, they are converted back into
7097           original form.
7098
7099       -fdevirtualize-at-ltrans
7100           Stream extra information needed for aggressive devirtualization
7101           when running the link-time optimizer in local transformation mode.
7102           This option enables more devirtualization but significantly
7103           increases the size of streamed data. For this reason it is disabled
7104           by default.
7105
7106       -fexpensive-optimizations
7107           Perform a number of minor optimizations that are relatively
7108           expensive.
7109
7110           Enabled at levels -O2, -O3, -Os.
7111
7112       -free
7113           Attempt to remove redundant extension instructions.  This is
7114           especially helpful for the x86-64 architecture, which implicitly
7115           zero-extends in 64-bit registers after writing to their lower
7116           32-bit half.
7117
7118           Enabled for Alpha, AArch64 and x86 at levels -O2, -O3, -Os.
7119
7120       -fno-lifetime-dse
7121           In C++ the value of an object is only affected by changes within
7122           its lifetime: when the constructor begins, the object has an
7123           indeterminate value, and any changes during the lifetime of the
7124           object are dead when the object is destroyed.  Normally dead store
7125           elimination will take advantage of this; if your code relies on the
7126           value of the object storage persisting beyond the lifetime of the
7127           object, you can use this flag to disable this optimization.  To
7128           preserve stores before the constructor starts (e.g. because your
7129           operator new clears the object storage) but still treat the object
7130           as dead after the destructor you, can use -flifetime-dse=1.  The
7131           default behavior can be explicitly selected with -flifetime-dse=2.
7132           -flifetime-dse=0 is equivalent to -fno-lifetime-dse.
7133
7134       -flive-range-shrinkage
7135           Attempt to decrease register pressure through register live range
7136           shrinkage.  This is helpful for fast processors with small or
7137           moderate size register sets.
7138
7139       -fira-algorithm=algorithm
7140           Use the specified coloring algorithm for the integrated register
7141           allocator.  The algorithm argument can be priority, which specifies
7142           Chow's priority coloring, or CB, which specifies Chaitin-Briggs
7143           coloring.  Chaitin-Briggs coloring is not implemented for all
7144           architectures, but for those targets that do support it, it is the
7145           default because it generates better code.
7146
7147       -fira-region=region
7148           Use specified regions for the integrated register allocator.  The
7149           region argument should be one of the following:
7150
7151           all Use all loops as register allocation regions.  This can give
7152               the best results for machines with a small and/or irregular
7153               register set.
7154
7155           mixed
7156               Use all loops except for loops with small register pressure as
7157               the regions.  This value usually gives the best results in most
7158               cases and for most architectures, and is enabled by default
7159               when compiling with optimization for speed (-O, -O2, ...).
7160
7161           one Use all functions as a single region.  This typically results
7162               in the smallest code size, and is enabled by default for -Os or
7163               -O0.
7164
7165       -fira-hoist-pressure
7166           Use IRA to evaluate register pressure in the code hoisting pass for
7167           decisions to hoist expressions.  This option usually results in
7168           smaller code, but it can slow the compiler down.
7169
7170           This option is enabled at level -Os for all targets.
7171
7172       -fira-loop-pressure
7173           Use IRA to evaluate register pressure in loops for decisions to
7174           move loop invariants.  This option usually results in generation of
7175           faster and smaller code on machines with large register files (>=
7176           32 registers), but it can slow the compiler down.
7177
7178           This option is enabled at level -O3 for some targets.
7179
7180       -fno-ira-share-save-slots
7181           Disable sharing of stack slots used for saving call-used hard
7182           registers living through a call.  Each hard register gets a
7183           separate stack slot, and as a result function stack frames are
7184           larger.
7185
7186       -fno-ira-share-spill-slots
7187           Disable sharing of stack slots allocated for pseudo-registers.
7188           Each pseudo-register that does not get a hard register gets a
7189           separate stack slot, and as a result function stack frames are
7190           larger.
7191
7192       -flra-remat
7193           Enable CFG-sensitive rematerialization in LRA.  Instead of loading
7194           values of spilled pseudos, LRA tries to rematerialize (recalculate)
7195           values if it is profitable.
7196
7197           Enabled at levels -O2, -O3, -Os.
7198
7199       -fdelayed-branch
7200           If supported for the target machine, attempt to reorder
7201           instructions to exploit instruction slots available after delayed
7202           branch instructions.
7203
7204           Enabled at levels -O, -O2, -O3, -Os, but not at -Og.
7205
7206       -fschedule-insns
7207           If supported for the target machine, attempt to reorder
7208           instructions to eliminate execution stalls due to required data
7209           being unavailable.  This helps machines that have slow floating
7210           point or memory load instructions by allowing other instructions to
7211           be issued until the result of the load or floating-point
7212           instruction is required.
7213
7214           Enabled at levels -O2, -O3.
7215
7216       -fschedule-insns2
7217           Similar to -fschedule-insns, but requests an additional pass of
7218           instruction scheduling after register allocation has been done.
7219           This is especially useful on machines with a relatively small
7220           number of registers and where memory load instructions take more
7221           than one cycle.
7222
7223           Enabled at levels -O2, -O3, -Os.
7224
7225       -fno-sched-interblock
7226           Disable instruction scheduling across basic blocks, which is
7227           normally enabled when scheduling before register allocation, i.e.
7228           with -fschedule-insns or at -O2 or higher.
7229
7230       -fno-sched-spec
7231           Disable speculative motion of non-load instructions, which is
7232           normally enabled when scheduling before register allocation, i.e.
7233           with -fschedule-insns or at -O2 or higher.
7234
7235       -fsched-pressure
7236           Enable register pressure sensitive insn scheduling before register
7237           allocation.  This only makes sense when scheduling before register
7238           allocation is enabled, i.e. with -fschedule-insns or at -O2 or
7239           higher.  Usage of this option can improve the generated code and
7240           decrease its size by preventing register pressure increase above
7241           the number of available hard registers and subsequent spills in
7242           register allocation.
7243
7244       -fsched-spec-load
7245           Allow speculative motion of some load instructions.  This only
7246           makes sense when scheduling before register allocation, i.e. with
7247           -fschedule-insns or at -O2 or higher.
7248
7249       -fsched-spec-load-dangerous
7250           Allow speculative motion of more load instructions.  This only
7251           makes sense when scheduling before register allocation, i.e. with
7252           -fschedule-insns or at -O2 or higher.
7253
7254       -fsched-stalled-insns
7255       -fsched-stalled-insns=n
7256           Define how many insns (if any) can be moved prematurely from the
7257           queue of stalled insns into the ready list during the second
7258           scheduling pass.  -fno-sched-stalled-insns means that no insns are
7259           moved prematurely, -fsched-stalled-insns=0 means there is no limit
7260           on how many queued insns can be moved prematurely.
7261           -fsched-stalled-insns without a value is equivalent to
7262           -fsched-stalled-insns=1.
7263
7264       -fsched-stalled-insns-dep
7265       -fsched-stalled-insns-dep=n
7266           Define how many insn groups (cycles) are examined for a dependency
7267           on a stalled insn that is a candidate for premature removal from
7268           the queue of stalled insns.  This has an effect only during the
7269           second scheduling pass, and only if -fsched-stalled-insns is used.
7270           -fno-sched-stalled-insns-dep is equivalent to
7271           -fsched-stalled-insns-dep=0.  -fsched-stalled-insns-dep without a
7272           value is equivalent to -fsched-stalled-insns-dep=1.
7273
7274       -fsched2-use-superblocks
7275           When scheduling after register allocation, use superblock
7276           scheduling.  This allows motion across basic block boundaries,
7277           resulting in faster schedules.  This option is experimental, as not
7278           all machine descriptions used by GCC model the CPU closely enough
7279           to avoid unreliable results from the algorithm.
7280
7281           This only makes sense when scheduling after register allocation,
7282           i.e. with -fschedule-insns2 or at -O2 or higher.
7283
7284       -fsched-group-heuristic
7285           Enable the group heuristic in the scheduler.  This heuristic favors
7286           the instruction that belongs to a schedule group.  This is enabled
7287           by default when scheduling is enabled, i.e. with -fschedule-insns
7288           or -fschedule-insns2 or at -O2 or higher.
7289
7290       -fsched-critical-path-heuristic
7291           Enable the critical-path heuristic in the scheduler.  This
7292           heuristic favors instructions on the critical path.  This is
7293           enabled by default when scheduling is enabled, i.e. with
7294           -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
7295
7296       -fsched-spec-insn-heuristic
7297           Enable the speculative instruction heuristic in the scheduler.
7298           This heuristic favors speculative instructions with greater
7299           dependency weakness.  This is enabled by default when scheduling is
7300           enabled, i.e.  with -fschedule-insns or -fschedule-insns2 or at -O2
7301           or higher.
7302
7303       -fsched-rank-heuristic
7304           Enable the rank heuristic in the scheduler.  This heuristic favors
7305           the instruction belonging to a basic block with greater size or
7306           frequency.  This is enabled by default when scheduling is enabled,
7307           i.e.  with -fschedule-insns or -fschedule-insns2 or at -O2 or
7308           higher.
7309
7310       -fsched-last-insn-heuristic
7311           Enable the last-instruction heuristic in the scheduler.  This
7312           heuristic favors the instruction that is less dependent on the last
7313           instruction scheduled.  This is enabled by default when scheduling
7314           is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at
7315           -O2 or higher.
7316
7317       -fsched-dep-count-heuristic
7318           Enable the dependent-count heuristic in the scheduler.  This
7319           heuristic favors the instruction that has more instructions
7320           depending on it.  This is enabled by default when scheduling is
7321           enabled, i.e.  with -fschedule-insns or -fschedule-insns2 or at -O2
7322           or higher.
7323
7324       -freschedule-modulo-scheduled-loops
7325           Modulo scheduling is performed before traditional scheduling.  If a
7326           loop is modulo scheduled, later scheduling passes may change its
7327           schedule.  Use this option to control that behavior.
7328
7329       -fselective-scheduling
7330           Schedule instructions using selective scheduling algorithm.
7331           Selective scheduling runs instead of the first scheduler pass.
7332
7333       -fselective-scheduling2
7334           Schedule instructions using selective scheduling algorithm.
7335           Selective scheduling runs instead of the second scheduler pass.
7336
7337       -fsel-sched-pipelining
7338           Enable software pipelining of innermost loops during selective
7339           scheduling.  This option has no effect unless one of
7340           -fselective-scheduling or -fselective-scheduling2 is turned on.
7341
7342       -fsel-sched-pipelining-outer-loops
7343           When pipelining loops during selective scheduling, also pipeline
7344           outer loops.  This option has no effect unless
7345           -fsel-sched-pipelining is turned on.
7346
7347       -fsemantic-interposition
7348           Some object formats, like ELF, allow interposing of symbols by the
7349           dynamic linker.  This means that for symbols exported from the DSO,
7350           the compiler cannot perform interprocedural propagation, inlining
7351           and other optimizations in anticipation that the function or
7352           variable in question may change. While this feature is useful, for
7353           example, to rewrite memory allocation functions by a debugging
7354           implementation, it is expensive in the terms of code quality.  With
7355           -fno-semantic-interposition the compiler assumes that if
7356           interposition happens for functions the overwriting function will
7357           have precisely the same semantics (and side effects).  Similarly if
7358           interposition happens for variables, the constructor of the
7359           variable will be the same. The flag has no effect for functions
7360           explicitly declared inline (where it is never allowed for
7361           interposition to change semantics) and for symbols explicitly
7362           declared weak.
7363
7364       -fshrink-wrap
7365           Emit function prologues only before parts of the function that need
7366           it, rather than at the top of the function.  This flag is enabled
7367           by default at -O and higher.
7368
7369       -fshrink-wrap-separate
7370           Shrink-wrap separate parts of the prologue and epilogue separately,
7371           so that those parts are only executed when needed.  This option is
7372           on by default, but has no effect unless -fshrink-wrap is also
7373           turned on and the target supports this.
7374
7375       -fcaller-saves
7376           Enable allocation of values to registers that are clobbered by
7377           function calls, by emitting extra instructions to save and restore
7378           the registers around such calls.  Such allocation is done only when
7379           it seems to result in better code.
7380
7381           This option is always enabled by default on certain machines,
7382           usually those which have no call-preserved registers to use
7383           instead.
7384
7385           Enabled at levels -O2, -O3, -Os.
7386
7387       -fcombine-stack-adjustments
7388           Tracks stack adjustments (pushes and pops) and stack memory
7389           references and then tries to find ways to combine them.
7390
7391           Enabled by default at -O1 and higher.
7392
7393       -fipa-ra
7394           Use caller save registers for allocation if those registers are not
7395           used by any called function.  In that case it is not necessary to
7396           save and restore them around calls.  This is only possible if
7397           called functions are part of same compilation unit as current
7398           function and they are compiled before it.
7399
7400           Enabled at levels -O2, -O3, -Os, however the option is disabled if
7401           generated code will be instrumented for profiling (-p, or -pg) or
7402           if callee's register usage cannot be known exactly (this happens on
7403           targets that do not expose prologues and epilogues in RTL).
7404
7405       -fconserve-stack
7406           Attempt to minimize stack usage.  The compiler attempts to use less
7407           stack space, even if that makes the program slower.  This option
7408           implies setting the large-stack-frame parameter to 100 and the
7409           large-stack-frame-growth parameter to 400.
7410
7411       -ftree-reassoc
7412           Perform reassociation on trees.  This flag is enabled by default at
7413           -O and higher.
7414
7415       -fcode-hoisting
7416           Perform code hoisting.  Code hoisting tries to move the evaluation
7417           of expressions executed on all paths to the function exit as early
7418           as possible.  This is especially useful as a code size
7419           optimization, but it often helps for code speed as well.  This flag
7420           is enabled by default at -O2 and higher.
7421
7422       -ftree-pre
7423           Perform partial redundancy elimination (PRE) on trees.  This flag
7424           is enabled by default at -O2 and -O3.
7425
7426       -ftree-partial-pre
7427           Make partial redundancy elimination (PRE) more aggressive.  This
7428           flag is enabled by default at -O3.
7429
7430       -ftree-forwprop
7431           Perform forward propagation on trees.  This flag is enabled by
7432           default at -O and higher.
7433
7434       -ftree-fre
7435           Perform full redundancy elimination (FRE) on trees.  The difference
7436           between FRE and PRE is that FRE only considers expressions that are
7437           computed on all paths leading to the redundant computation.  This
7438           analysis is faster than PRE, though it exposes fewer redundancies.
7439           This flag is enabled by default at -O and higher.
7440
7441       -ftree-phiprop
7442           Perform hoisting of loads from conditional pointers on trees.  This
7443           pass is enabled by default at -O and higher.
7444
7445       -fhoist-adjacent-loads
7446           Speculatively hoist loads from both branches of an if-then-else if
7447           the loads are from adjacent locations in the same structure and the
7448           target architecture has a conditional move instruction.  This flag
7449           is enabled by default at -O2 and higher.
7450
7451       -ftree-copy-prop
7452           Perform copy propagation on trees.  This pass eliminates
7453           unnecessary copy operations.  This flag is enabled by default at -O
7454           and higher.
7455
7456       -fipa-pure-const
7457           Discover which functions are pure or constant.  Enabled by default
7458           at -O and higher.
7459
7460       -fipa-reference
7461           Discover which static variables do not escape the compilation unit.
7462           Enabled by default at -O and higher.
7463
7464       -fipa-reference-addressable
7465           Discover read-only, write-only and non-addressable static
7466           variables.  Enabled by default at -O and higher.
7467
7468       -fipa-stack-alignment
7469           Reduce stack alignment on call sites if possible.  Enabled by
7470           default.
7471
7472       -fipa-pta
7473           Perform interprocedural pointer analysis and interprocedural
7474           modification and reference analysis.  This option can cause
7475           excessive memory and compile-time usage on large compilation units.
7476           It is not enabled by default at any optimization level.
7477
7478       -fipa-profile
7479           Perform interprocedural profile propagation.  The functions called
7480           only from cold functions are marked as cold. Also functions
7481           executed once (such as "cold", "noreturn", static constructors or
7482           destructors) are identified. Cold functions and loop less parts of
7483           functions executed once are then optimized for size.  Enabled by
7484           default at -O and higher.
7485
7486       -fipa-cp
7487           Perform interprocedural constant propagation.  This optimization
7488           analyzes the program to determine when values passed to functions
7489           are constants and then optimizes accordingly.  This optimization
7490           can substantially increase performance if the application has
7491           constants passed to functions.  This flag is enabled by default at
7492           -O2, -Os and -O3.  It is also enabled by -fprofile-use and
7493           -fauto-profile.
7494
7495       -fipa-cp-clone
7496           Perform function cloning to make interprocedural constant
7497           propagation stronger.  When enabled, interprocedural constant
7498           propagation performs function cloning when externally visible
7499           function can be called with constant arguments.  Because this
7500           optimization can create multiple copies of functions, it may
7501           significantly increase code size (see --param
7502           ipcp-unit-growth=value).  This flag is enabled by default at -O3.
7503           It is also enabled by -fprofile-use and -fauto-profile.
7504
7505       -fipa-bit-cp
7506           When enabled, perform interprocedural bitwise constant propagation.
7507           This flag is enabled by default at -O2 and by -fprofile-use and
7508           -fauto-profile.  It requires that -fipa-cp is enabled.
7509
7510       -fipa-vrp
7511           When enabled, perform interprocedural propagation of value ranges.
7512           This flag is enabled by default at -O2. It requires that -fipa-cp
7513           is enabled.
7514
7515       -fipa-icf
7516           Perform Identical Code Folding for functions and read-only
7517           variables.  The optimization reduces code size and may disturb
7518           unwind stacks by replacing a function by equivalent one with a
7519           different name. The optimization works more effectively with link-
7520           time optimization enabled.
7521
7522           Although the behavior is similar to the Gold Linker's ICF
7523           optimization, GCC ICF works on different levels and thus the
7524           optimizations are not same - there are equivalences that are found
7525           only by GCC and equivalences found only by Gold.
7526
7527           This flag is enabled by default at -O2 and -Os.
7528
7529       -flive-patching=level
7530           Control GCC's optimizations to produce output suitable for live-
7531           patching.
7532
7533           If the compiler's optimization uses a function's body or
7534           information extracted from its body to optimize/change another
7535           function, the latter is called an impacted function of the former.
7536           If a function is patched, its impacted functions should be patched
7537           too.
7538
7539           The impacted functions are determined by the compiler's
7540           interprocedural optimizations.  For example, a caller is impacted
7541           when inlining a function into its caller, cloning a function and
7542           changing its caller to call this new clone, or extracting a
7543           function's pureness/constness information to optimize its direct or
7544           indirect callers, etc.
7545
7546           Usually, the more IPA optimizations enabled, the larger the number
7547           of impacted functions for each function.  In order to control the
7548           number of impacted functions and more easily compute the list of
7549           impacted function, IPA optimizations can be partially enabled at
7550           two different levels.
7551
7552           The level argument should be one of the following:
7553
7554           inline-clone
7555               Only enable inlining and cloning optimizations, which includes
7556               inlining, cloning, interprocedural scalar replacement of
7557               aggregates and partial inlining.  As a result, when patching a
7558               function, all its callers and its clones' callers are impacted,
7559               therefore need to be patched as well.
7560
7561               -flive-patching=inline-clone disables the following
7562               optimization flags: -fwhole-program  -fipa-pta  -fipa-reference
7563               -fipa-ra -fipa-icf  -fipa-icf-functions  -fipa-icf-variables
7564               -fipa-bit-cp  -fipa-vrp  -fipa-pure-const
7565               -fipa-reference-addressable -fipa-stack-alignment
7566
7567           inline-only-static
7568               Only enable inlining of static functions.  As a result, when
7569               patching a static function, all its callers are impacted and so
7570               need to be patched as well.
7571
7572               In addition to all the flags that -flive-patching=inline-clone
7573               disables, -flive-patching=inline-only-static disables the
7574               following additional optimization flags: -fipa-cp-clone
7575               -fipa-sra  -fpartial-inlining  -fipa-cp
7576
7577           When -flive-patching is specified without any value, the default
7578           value is inline-clone.
7579
7580           This flag is disabled by default.
7581
7582           Note that -flive-patching is not supported with link-time
7583           optimization (-flto).
7584
7585       -fisolate-erroneous-paths-dereference
7586           Detect paths that trigger erroneous or undefined behavior due to
7587           dereferencing a null pointer.  Isolate those paths from the main
7588           control flow and turn the statement with erroneous or undefined
7589           behavior into a trap.  This flag is enabled by default at -O2 and
7590           higher and depends on -fdelete-null-pointer-checks also being
7591           enabled.
7592
7593       -fisolate-erroneous-paths-attribute
7594           Detect paths that trigger erroneous or undefined behavior due to a
7595           null value being used in a way forbidden by a "returns_nonnull" or
7596           "nonnull" attribute.  Isolate those paths from the main control
7597           flow and turn the statement with erroneous or undefined behavior
7598           into a trap.  This is not currently enabled, but may be enabled by
7599           -O2 in the future.
7600
7601       -ftree-sink
7602           Perform forward store motion on trees.  This flag is enabled by
7603           default at -O and higher.
7604
7605       -ftree-bit-ccp
7606           Perform sparse conditional bit constant propagation on trees and
7607           propagate pointer alignment information.  This pass only operates
7608           on local scalar variables and is enabled by default at -O1 and
7609           higher, except for -Og.  It requires that -ftree-ccp is enabled.
7610
7611       -ftree-ccp
7612           Perform sparse conditional constant propagation (CCP) on trees.
7613           This pass only operates on local scalar variables and is enabled by
7614           default at -O and higher.
7615
7616       -fssa-backprop
7617           Propagate information about uses of a value up the definition chain
7618           in order to simplify the definitions.  For example, this pass
7619           strips sign operations if the sign of a value never matters.  The
7620           flag is enabled by default at -O and higher.
7621
7622       -fssa-phiopt
7623           Perform pattern matching on SSA PHI nodes to optimize conditional
7624           code.  This pass is enabled by default at -O1 and higher, except
7625           for -Og.
7626
7627       -ftree-switch-conversion
7628           Perform conversion of simple initializations in a switch to
7629           initializations from a scalar array.  This flag is enabled by
7630           default at -O2 and higher.
7631
7632       -ftree-tail-merge
7633           Look for identical code sequences.  When found, replace one with a
7634           jump to the other.  This optimization is known as tail merging or
7635           cross jumping.  This flag is enabled by default at -O2 and higher.
7636           The compilation time in this pass can be limited using max-tail-
7637           merge-comparisons parameter and max-tail-merge-iterations
7638           parameter.
7639
7640       -ftree-dce
7641           Perform dead code elimination (DCE) on trees.  This flag is enabled
7642           by default at -O and higher.
7643
7644       -ftree-builtin-call-dce
7645           Perform conditional dead code elimination (DCE) for calls to built-
7646           in functions that may set "errno" but are otherwise free of side
7647           effects.  This flag is enabled by default at -O2 and higher if -Os
7648           is not also specified.
7649
7650       -ftree-dominator-opts
7651           Perform a variety of simple scalar cleanups (constant/copy
7652           propagation, redundancy elimination, range propagation and
7653           expression simplification) based on a dominator tree traversal.
7654           This also performs jump threading (to reduce jumps to jumps). This
7655           flag is enabled by default at -O and higher.
7656
7657       -ftree-dse
7658           Perform dead store elimination (DSE) on trees.  A dead store is a
7659           store into a memory location that is later overwritten by another
7660           store without any intervening loads.  In this case the earlier
7661           store can be deleted.  This flag is enabled by default at -O and
7662           higher.
7663
7664       -ftree-ch
7665           Perform loop header copying on trees.  This is beneficial since it
7666           increases effectiveness of code motion optimizations.  It also
7667           saves one jump.  This flag is enabled by default at -O and higher.
7668           It is not enabled for -Os, since it usually increases code size.
7669
7670       -ftree-loop-optimize
7671           Perform loop optimizations on trees.  This flag is enabled by
7672           default at -O and higher.
7673
7674       -ftree-loop-linear
7675       -floop-strip-mine
7676       -floop-block
7677           Perform loop nest optimizations.  Same as -floop-nest-optimize.  To
7678           use this code transformation, GCC has to be configured with
7679           --with-isl to enable the Graphite loop transformation
7680           infrastructure.
7681
7682       -fgraphite-identity
7683           Enable the identity transformation for graphite.  For every SCoP we
7684           generate the polyhedral representation and transform it back to
7685           gimple.  Using -fgraphite-identity we can check the costs or
7686           benefits of the GIMPLE -> GRAPHITE -> GIMPLE transformation.  Some
7687           minimal optimizations are also performed by the code generator isl,
7688           like index splitting and dead code elimination in loops.
7689
7690       -floop-nest-optimize
7691           Enable the isl based loop nest optimizer.  This is a generic loop
7692           nest optimizer based on the Pluto optimization algorithms.  It
7693           calculates a loop structure optimized for data-locality and
7694           parallelism.  This option is experimental.
7695
7696       -floop-parallelize-all
7697           Use the Graphite data dependence analysis to identify loops that
7698           can be parallelized.  Parallelize all the loops that can be
7699           analyzed to not contain loop carried dependences without checking
7700           that it is profitable to parallelize the loops.
7701
7702       -ftree-coalesce-vars
7703           While transforming the program out of the SSA representation,
7704           attempt to reduce copying by coalescing versions of different user-
7705           defined variables, instead of just compiler temporaries.  This may
7706           severely limit the ability to debug an optimized program compiled
7707           with -fno-var-tracking-assignments.  In the negated form, this flag
7708           prevents SSA coalescing of user variables.  This option is enabled
7709           by default if optimization is enabled, and it does very little
7710           otherwise.
7711
7712       -ftree-loop-if-convert
7713           Attempt to transform conditional jumps in the innermost loops to
7714           branch-less equivalents.  The intent is to remove control-flow from
7715           the innermost loops in order to improve the ability of the
7716           vectorization pass to handle these loops.  This is enabled by
7717           default if vectorization is enabled.
7718
7719       -ftree-loop-distribution
7720           Perform loop distribution.  This flag can improve cache performance
7721           on big loop bodies and allow further loop optimizations, like
7722           parallelization or vectorization, to take place.  For example, the
7723           loop
7724
7725                   DO I = 1, N
7726                     A(I) = B(I) + C
7727                     D(I) = E(I) * F
7728                   ENDDO
7729
7730           is transformed to
7731
7732                   DO I = 1, N
7733                      A(I) = B(I) + C
7734                   ENDDO
7735                   DO I = 1, N
7736                      D(I) = E(I) * F
7737                   ENDDO
7738
7739           This flag is enabled by default at -O3.  It is also enabled by
7740           -fprofile-use and -fauto-profile.
7741
7742       -ftree-loop-distribute-patterns
7743           Perform loop distribution of patterns that can be code generated
7744           with calls to a library.  This flag is enabled by default at -O3,
7745           and by -fprofile-use and -fauto-profile.
7746
7747           This pass distributes the initialization loops and generates a call
7748           to memset zero.  For example, the loop
7749
7750                   DO I = 1, N
7751                     A(I) = 0
7752                     B(I) = A(I) + I
7753                   ENDDO
7754
7755           is transformed to
7756
7757                   DO I = 1, N
7758                      A(I) = 0
7759                   ENDDO
7760                   DO I = 1, N
7761                      B(I) = A(I) + I
7762                   ENDDO
7763
7764           and the initialization loop is transformed into a call to memset
7765           zero.  This flag is enabled by default at -O3.  It is also enabled
7766           by -fprofile-use and -fauto-profile.
7767
7768       -floop-interchange
7769           Perform loop interchange outside of graphite.  This flag can
7770           improve cache performance on loop nest and allow further loop
7771           optimizations, like vectorization, to take place.  For example, the
7772           loop
7773
7774                   for (int i = 0; i < N; i++)
7775                     for (int j = 0; j < N; j++)
7776                       for (int k = 0; k < N; k++)
7777                         c[i][j] = c[i][j] + a[i][k]*b[k][j];
7778
7779           is transformed to
7780
7781                   for (int i = 0; i < N; i++)
7782                     for (int k = 0; k < N; k++)
7783                       for (int j = 0; j < N; j++)
7784                         c[i][j] = c[i][j] + a[i][k]*b[k][j];
7785
7786           This flag is enabled by default at -O3.  It is also enabled by
7787           -fprofile-use and -fauto-profile.
7788
7789       -floop-unroll-and-jam
7790           Apply unroll and jam transformations on feasible loops.  In a loop
7791           nest this unrolls the outer loop by some factor and fuses the
7792           resulting multiple inner loops.  This flag is enabled by default at
7793           -O3.  It is also enabled by -fprofile-use and -fauto-profile.
7794
7795       -ftree-loop-im
7796           Perform loop invariant motion on trees.  This pass moves only
7797           invariants that are hard to handle at RTL level (function calls,
7798           operations that expand to nontrivial sequences of insns).  With
7799           -funswitch-loops it also moves operands of conditions that are
7800           invariant out of the loop, so that we can use just trivial
7801           invariantness analysis in loop unswitching.  The pass also includes
7802           store motion.
7803
7804       -ftree-loop-ivcanon
7805           Create a canonical counter for number of iterations in loops for
7806           which determining number of iterations requires complicated
7807           analysis.  Later optimizations then may determine the number
7808           easily.  Useful especially in connection with unrolling.
7809
7810       -ftree-scev-cprop
7811           Perform final value replacement.  If a variable is modified in a
7812           loop in such a way that its value when exiting the loop can be
7813           determined using only its initial value and the number of loop
7814           iterations, replace uses of the final value by such a computation,
7815           provided it is sufficiently cheap.  This reduces data dependencies
7816           and may allow further simplifications.  Enabled by default at -O
7817           and higher.
7818
7819       -fivopts
7820           Perform induction variable optimizations (strength reduction,
7821           induction variable merging and induction variable elimination) on
7822           trees.
7823
7824       -ftree-parallelize-loops=n
7825           Parallelize loops, i.e., split their iteration space to run in n
7826           threads.  This is only possible for loops whose iterations are
7827           independent and can be arbitrarily reordered.  The optimization is
7828           only profitable on multiprocessor machines, for loops that are CPU-
7829           intensive, rather than constrained e.g. by memory bandwidth.  This
7830           option implies -pthread, and thus is only supported on targets that
7831           have support for -pthread.
7832
7833       -ftree-pta
7834           Perform function-local points-to analysis on trees.  This flag is
7835           enabled by default at -O1 and higher, except for -Og.
7836
7837       -ftree-sra
7838           Perform scalar replacement of aggregates.  This pass replaces
7839           structure references with scalars to prevent committing structures
7840           to memory too early.  This flag is enabled by default at -O1 and
7841           higher, except for -Og.
7842
7843       -fstore-merging
7844           Perform merging of narrow stores to consecutive memory addresses.
7845           This pass merges contiguous stores of immediate values narrower
7846           than a word into fewer wider stores to reduce the number of
7847           instructions.  This is enabled by default at -O2 and higher as well
7848           as -Os.
7849
7850       -ftree-ter
7851           Perform temporary expression replacement during the SSA->normal
7852           phase.  Single use/single def temporaries are replaced at their use
7853           location with their defining expression.  This results in non-
7854           GIMPLE code, but gives the expanders much more complex trees to
7855           work on resulting in better RTL generation.  This is enabled by
7856           default at -O and higher.
7857
7858       -ftree-slsr
7859           Perform straight-line strength reduction on trees.  This recognizes
7860           related expressions involving multiplications and replaces them by
7861           less expensive calculations when possible.  This is enabled by
7862           default at -O and higher.
7863
7864       -ftree-vectorize
7865           Perform vectorization on trees. This flag enables
7866           -ftree-loop-vectorize and -ftree-slp-vectorize if not explicitly
7867           specified.
7868
7869       -ftree-loop-vectorize
7870           Perform loop vectorization on trees. This flag is enabled by
7871           default at -O3 and by -ftree-vectorize, -fprofile-use, and
7872           -fauto-profile.
7873
7874       -ftree-slp-vectorize
7875           Perform basic block vectorization on trees. This flag is enabled by
7876           default at -O3 and by -ftree-vectorize, -fprofile-use, and
7877           -fauto-profile.
7878
7879       -fvect-cost-model=model
7880           Alter the cost model used for vectorization.  The model argument
7881           should be one of unlimited, dynamic or cheap.  With the unlimited
7882           model the vectorized code-path is assumed to be profitable while
7883           with the dynamic model a runtime check guards the vectorized code-
7884           path to enable it only for iteration counts that will likely
7885           execute faster than when executing the original scalar loop.  The
7886           cheap model disables vectorization of loops where doing so would be
7887           cost prohibitive for example due to required runtime checks for
7888           data dependence or alignment but otherwise is equal to the dynamic
7889           model.  The default cost model depends on other optimization flags
7890           and is either dynamic or cheap.
7891
7892       -fsimd-cost-model=model
7893           Alter the cost model used for vectorization of loops marked with
7894           the OpenMP simd directive.  The model argument should be one of
7895           unlimited, dynamic, cheap.  All values of model have the same
7896           meaning as described in -fvect-cost-model and by default a cost
7897           model defined with -fvect-cost-model is used.
7898
7899       -ftree-vrp
7900           Perform Value Range Propagation on trees.  This is similar to the
7901           constant propagation pass, but instead of values, ranges of values
7902           are propagated.  This allows the optimizers to remove unnecessary
7903           range checks like array bound checks and null pointer checks.  This
7904           is enabled by default at -O2 and higher.  Null pointer check
7905           elimination is only done if -fdelete-null-pointer-checks is
7906           enabled.
7907
7908       -fsplit-paths
7909           Split paths leading to loop backedges.  This can improve dead code
7910           elimination and common subexpression elimination.  This is enabled
7911           by default at -O3 and above.
7912
7913       -fsplit-ivs-in-unroller
7914           Enables expression of values of induction variables in later
7915           iterations of the unrolled loop using the value in the first
7916           iteration.  This breaks long dependency chains, thus improving
7917           efficiency of the scheduling passes.
7918
7919           A combination of -fweb and CSE is often sufficient to obtain the
7920           same effect.  However, that is not reliable in cases where the loop
7921           body is more complicated than a single basic block.  It also does
7922           not work at all on some architectures due to restrictions in the
7923           CSE pass.
7924
7925           This optimization is enabled by default.
7926
7927       -fvariable-expansion-in-unroller
7928           With this option, the compiler creates multiple copies of some
7929           local variables when unrolling a loop, which can result in superior
7930           code.
7931
7932       -fpartial-inlining
7933           Inline parts of functions.  This option has any effect only when
7934           inlining itself is turned on by the -finline-functions or
7935           -finline-small-functions options.
7936
7937           Enabled at levels -O2, -O3, -Os.
7938
7939       -fpredictive-commoning
7940           Perform predictive commoning optimization, i.e., reusing
7941           computations (especially memory loads and stores) performed in
7942           previous iterations of loops.
7943
7944           This option is enabled at level -O3.  It is also enabled by
7945           -fprofile-use and -fauto-profile.
7946
7947       -fprefetch-loop-arrays
7948           If supported by the target machine, generate instructions to
7949           prefetch memory to improve the performance of loops that access
7950           large arrays.
7951
7952           This option may generate better or worse code; results are highly
7953           dependent on the structure of loops within the source code.
7954
7955           Disabled at level -Os.
7956
7957       -fno-printf-return-value
7958           Do not substitute constants for known return value of formatted
7959           output functions such as "sprintf", "snprintf", "vsprintf", and
7960           "vsnprintf" (but not "printf" of "fprintf").  This transformation
7961           allows GCC to optimize or even eliminate branches based on the
7962           known return value of these functions called with arguments that
7963           are either constant, or whose values are known to be in a range
7964           that makes determining the exact return value possible.  For
7965           example, when -fprintf-return-value is in effect, both the branch
7966           and the body of the "if" statement (but not the call to "snprint")
7967           can be optimized away when "i" is a 32-bit or smaller integer
7968           because the return value is guaranteed to be at most 8.
7969
7970                   char buf[9];
7971                   if (snprintf (buf, "%08x", i) >= sizeof buf)
7972                     ...
7973
7974           The -fprintf-return-value option relies on other optimizations and
7975           yields best results with -O2 and above.  It works in tandem with
7976           the -Wformat-overflow and -Wformat-truncation options.  The
7977           -fprintf-return-value option is enabled by default.
7978
7979       -fno-peephole
7980       -fno-peephole2
7981           Disable any machine-specific peephole optimizations.  The
7982           difference between -fno-peephole and -fno-peephole2 is in how they
7983           are implemented in the compiler; some targets use one, some use the
7984           other, a few use both.
7985
7986           -fpeephole is enabled by default.  -fpeephole2 enabled at levels
7987           -O2, -O3, -Os.
7988
7989       -fno-guess-branch-probability
7990           Do not guess branch probabilities using heuristics.
7991
7992           GCC uses heuristics to guess branch probabilities if they are not
7993           provided by profiling feedback (-fprofile-arcs).  These heuristics
7994           are based on the control flow graph.  If some branch probabilities
7995           are specified by "__builtin_expect", then the heuristics are used
7996           to guess branch probabilities for the rest of the control flow
7997           graph, taking the "__builtin_expect" info into account.  The
7998           interactions between the heuristics and "__builtin_expect" can be
7999           complex, and in some cases, it may be useful to disable the
8000           heuristics so that the effects of "__builtin_expect" are easier to
8001           understand.
8002
8003           It is also possible to specify expected probability of the
8004           expression with "__builtin_expect_with_probability" built-in
8005           function.
8006
8007           The default is -fguess-branch-probability at levels -O, -O2, -O3,
8008           -Os.
8009
8010       -freorder-blocks
8011           Reorder basic blocks in the compiled function in order to reduce
8012           number of taken branches and improve code locality.
8013
8014           Enabled at levels -O, -O2, -O3, -Os.
8015
8016       -freorder-blocks-algorithm=algorithm
8017           Use the specified algorithm for basic block reordering.  The
8018           algorithm argument can be simple, which does not increase code size
8019           (except sometimes due to secondary effects like alignment), or stc,
8020           the "software trace cache" algorithm, which tries to put all often
8021           executed code together, minimizing the number of branches executed
8022           by making extra copies of code.
8023
8024           The default is simple at levels -O, -Os, and stc at levels -O2,
8025           -O3.
8026
8027       -freorder-blocks-and-partition
8028           In addition to reordering basic blocks in the compiled function, in
8029           order to reduce number of taken branches, partitions hot and cold
8030           basic blocks into separate sections of the assembly and .o files,
8031           to improve paging and cache locality performance.
8032
8033           This optimization is automatically turned off in the presence of
8034           exception handling or unwind tables (on targets using
8035           setjump/longjump or target specific scheme), for linkonce sections,
8036           for functions with a user-defined section attribute and on any
8037           architecture that does not support named sections.  When
8038           -fsplit-stack is used this option is not enabled by default (to
8039           avoid linker errors), but may be enabled explicitly (if using a
8040           working linker).
8041
8042           Enabled for x86 at levels -O2, -O3, -Os.
8043
8044       -freorder-functions
8045           Reorder functions in the object file in order to improve code
8046           locality.  This is implemented by using special subsections
8047           ".text.hot" for most frequently executed functions and
8048           ".text.unlikely" for unlikely executed functions.  Reordering is
8049           done by the linker so object file format must support named
8050           sections and linker must place them in a reasonable way.
8051
8052           This option isn't effective unless you either provide profile
8053           feedback (see -fprofile-arcs for details) or manually annotate
8054           functions with "hot" or "cold" attributes.
8055
8056           Enabled at levels -O2, -O3, -Os.
8057
8058       -fstrict-aliasing
8059           Allow the compiler to assume the strictest aliasing rules
8060           applicable to the language being compiled.  For C (and C++), this
8061           activates optimizations based on the type of expressions.  In
8062           particular, an object of one type is assumed never to reside at the
8063           same address as an object of a different type, unless the types are
8064           almost the same.  For example, an "unsigned int" can alias an
8065           "int", but not a "void*" or a "double".  A character type may alias
8066           any other type.
8067
8068           Pay special attention to code like this:
8069
8070                   union a_union {
8071                     int i;
8072                     double d;
8073                   };
8074
8075                   int f() {
8076                     union a_union t;
8077                     t.d = 3.0;
8078                     return t.i;
8079                   }
8080
8081           The practice of reading from a different union member than the one
8082           most recently written to (called "type-punning") is common.  Even
8083           with -fstrict-aliasing, type-punning is allowed, provided the
8084           memory is accessed through the union type.  So, the code above
8085           works as expected.    However, this code might not:
8086
8087                   int f() {
8088                     union a_union t;
8089                     int* ip;
8090                     t.d = 3.0;
8091                     ip = &t.i;
8092                     return *ip;
8093                   }
8094
8095           Similarly, access by taking the address, casting the resulting
8096           pointer and dereferencing the result has undefined behavior, even
8097           if the cast uses a union type, e.g.:
8098
8099                   int f() {
8100                     double d = 3.0;
8101                     return ((union a_union *) &d)->i;
8102                   }
8103
8104           The -fstrict-aliasing option is enabled at levels -O2, -O3, -Os.
8105
8106       -falign-functions
8107       -falign-functions=n
8108       -falign-functions=n:m
8109       -falign-functions=n:m:n2
8110       -falign-functions=n:m:n2:m2
8111           Align the start of functions to the next power-of-two greater than
8112           n, skipping up to m-1 bytes.  This ensures that at least the first
8113           m bytes of the function can be fetched by the CPU without crossing
8114           an n-byte alignment boundary.
8115
8116           If m is not specified, it defaults to n.
8117
8118           Examples: -falign-functions=32 aligns functions to the next 32-byte
8119           boundary, -falign-functions=24 aligns to the next 32-byte boundary
8120           only if this can be done by skipping 23 bytes or less,
8121           -falign-functions=32:7 aligns to the next 32-byte boundary only if
8122           this can be done by skipping 6 bytes or less.
8123
8124           The second pair of n2:m2 values allows you to specify a secondary
8125           alignment: -falign-functions=64:7:32:3 aligns to the next 64-byte
8126           boundary if this can be done by skipping 6 bytes or less, otherwise
8127           aligns to the next 32-byte boundary if this can be done by skipping
8128           2 bytes or less.  If m2 is not specified, it defaults to n2.
8129
8130           Some assemblers only support this flag when n is a power of two; in
8131           that case, it is rounded up.
8132
8133           -fno-align-functions and -falign-functions=1 are equivalent and
8134           mean that functions are not aligned.
8135
8136           If n is not specified or is zero, use a machine-dependent default.
8137           The maximum allowed n option value is 65536.
8138
8139           Enabled at levels -O2, -O3.
8140
8141       -flimit-function-alignment
8142           If this option is enabled, the compiler tries to avoid
8143           unnecessarily overaligning functions. It attempts to instruct the
8144           assembler to align by the amount specified by -falign-functions,
8145           but not to skip more bytes than the size of the function.
8146
8147       -falign-labels
8148       -falign-labels=n
8149       -falign-labels=n:m
8150       -falign-labels=n:m:n2
8151       -falign-labels=n:m:n2:m2
8152           Align all branch targets to a power-of-two boundary.
8153
8154           Parameters of this option are analogous to the -falign-functions
8155           option.  -fno-align-labels and -falign-labels=1 are equivalent and
8156           mean that labels are not aligned.
8157
8158           If -falign-loops or -falign-jumps are applicable and are greater
8159           than this value, then their values are used instead.
8160
8161           If n is not specified or is zero, use a machine-dependent default
8162           which is very likely to be 1, meaning no alignment.  The maximum
8163           allowed n option value is 65536.
8164
8165           Enabled at levels -O2, -O3.
8166
8167       -falign-loops
8168       -falign-loops=n
8169       -falign-loops=n:m
8170       -falign-loops=n:m:n2
8171       -falign-loops=n:m:n2:m2
8172           Align loops to a power-of-two boundary.  If the loops are executed
8173           many times, this makes up for any execution of the dummy padding
8174           instructions.
8175
8176           Parameters of this option are analogous to the -falign-functions
8177           option.  -fno-align-loops and -falign-loops=1 are equivalent and
8178           mean that loops are not aligned.  The maximum allowed n option
8179           value is 65536.
8180
8181           If n is not specified or is zero, use a machine-dependent default.
8182
8183           Enabled at levels -O2, -O3.
8184
8185       -falign-jumps
8186       -falign-jumps=n
8187       -falign-jumps=n:m
8188       -falign-jumps=n:m:n2
8189       -falign-jumps=n:m:n2:m2
8190           Align branch targets to a power-of-two boundary, for branch targets
8191           where the targets can only be reached by jumping.  In this case, no
8192           dummy operations need be executed.
8193
8194           Parameters of this option are analogous to the -falign-functions
8195           option.  -fno-align-jumps and -falign-jumps=1 are equivalent and
8196           mean that loops are not aligned.
8197
8198           If n is not specified or is zero, use a machine-dependent default.
8199           The maximum allowed n option value is 65536.
8200
8201           Enabled at levels -O2, -O3.
8202
8203       -funit-at-a-time
8204           This option is left for compatibility reasons. -funit-at-a-time has
8205           no effect, while -fno-unit-at-a-time implies -fno-toplevel-reorder
8206           and -fno-section-anchors.
8207
8208           Enabled by default.
8209
8210       -fno-toplevel-reorder
8211           Do not reorder top-level functions, variables, and "asm"
8212           statements.  Output them in the same order that they appear in the
8213           input file.  When this option is used, unreferenced static
8214           variables are not removed.  This option is intended to support
8215           existing code that relies on a particular ordering.  For new code,
8216           it is better to use attributes when possible.
8217
8218           -ftoplevel-reorder is the default at -O1 and higher, and also at
8219           -O0 if -fsection-anchors is explicitly requested.  Additionally
8220           -fno-toplevel-reorder implies -fno-section-anchors.
8221
8222       -fweb
8223           Constructs webs as commonly used for register allocation purposes
8224           and assign each web individual pseudo register.  This allows the
8225           register allocation pass to operate on pseudos directly, but also
8226           strengthens several other optimization passes, such as CSE, loop
8227           optimizer and trivial dead code remover.  It can, however, make
8228           debugging impossible, since variables no longer stay in a "home
8229           register".
8230
8231           Enabled by default with -funroll-loops.
8232
8233       -fwhole-program
8234           Assume that the current compilation unit represents the whole
8235           program being compiled.  All public functions and variables with
8236           the exception of "main" and those merged by attribute
8237           "externally_visible" become static functions and in effect are
8238           optimized more aggressively by interprocedural optimizers.
8239
8240           This option should not be used in combination with -flto.  Instead
8241           relying on a linker plugin should provide safer and more precise
8242           information.
8243
8244       -flto[=n]
8245           This option runs the standard link-time optimizer.  When invoked
8246           with source code, it generates GIMPLE (one of GCC's internal
8247           representations) and writes it to special ELF sections in the
8248           object file.  When the object files are linked together, all the
8249           function bodies are read from these ELF sections and instantiated
8250           as if they had been part of the same translation unit.
8251
8252           To use the link-time optimizer, -flto and optimization options
8253           should be specified at compile time and during the final link.  It
8254           is recommended that you compile all the files participating in the
8255           same link with the same options and also specify those options at
8256           link time.  For example:
8257
8258                   gcc -c -O2 -flto foo.c
8259                   gcc -c -O2 -flto bar.c
8260                   gcc -o myprog -flto -O2 foo.o bar.o
8261
8262           The first two invocations to GCC save a bytecode representation of
8263           GIMPLE into special ELF sections inside foo.o and bar.o.  The final
8264           invocation reads the GIMPLE bytecode from foo.o and bar.o, merges
8265           the two files into a single internal image, and compiles the result
8266           as usual.  Since both foo.o and bar.o are merged into a single
8267           image, this causes all the interprocedural analyses and
8268           optimizations in GCC to work across the two files as if they were a
8269           single one.  This means, for example, that the inliner is able to
8270           inline functions in bar.o into functions in foo.o and vice-versa.
8271
8272           Another (simpler) way to enable link-time optimization is:
8273
8274                   gcc -o myprog -flto -O2 foo.c bar.c
8275
8276           The above generates bytecode for foo.c and bar.c, merges them
8277           together into a single GIMPLE representation and optimizes them as
8278           usual to produce myprog.
8279
8280           The important thing to keep in mind is that to enable link-time
8281           optimizations you need to use the GCC driver to perform the link
8282           step.  GCC automatically performs link-time optimization if any of
8283           the objects involved were compiled with the -flto command-line
8284           option.  You can always override the automatic decision to do link-
8285           time optimization by passing -fno-lto to the link command.
8286
8287           To make whole program optimization effective, it is necessary to
8288           make certain whole program assumptions.  The compiler needs to know
8289           what functions and variables can be accessed by libraries and
8290           runtime outside of the link-time optimized unit.  When supported by
8291           the linker, the linker plugin (see -fuse-linker-plugin) passes
8292           information to the compiler about used and externally visible
8293           symbols.  When the linker plugin is not available, -fwhole-program
8294           should be used to allow the compiler to make these assumptions,
8295           which leads to more aggressive optimization decisions.
8296
8297           When a file is compiled with -flto without -fuse-linker-plugin, the
8298           generated object file is larger than a regular object file because
8299           it contains GIMPLE bytecodes and the usual final code (see
8300           -ffat-lto-objects.  This means that object files with LTO
8301           information can be linked as normal object files; if -fno-lto is
8302           passed to the linker, no interprocedural optimizations are applied.
8303           Note that when -fno-fat-lto-objects is enabled the compile stage is
8304           faster but you cannot perform a regular, non-LTO link on them.
8305
8306           When producing the final binary, GCC only applies link-time
8307           optimizations to those files that contain bytecode.  Therefore, you
8308           can mix and match object files and libraries with GIMPLE bytecodes
8309           and final object code.  GCC automatically selects which files to
8310           optimize in LTO mode and which files to link without further
8311           processing.
8312
8313           Generally, options specified at link time override those specified
8314           at compile time, although in some cases GCC attempts to infer link-
8315           time options from the settings used to compile the input files.
8316
8317           If you do not specify an optimization level option -O at link time,
8318           then GCC uses the highest optimization level used when compiling
8319           the object files.  Note that it is generally ineffective to specify
8320           an optimization level option only at link time and not at compile
8321           time, for two reasons.  First, compiling without optimization
8322           suppresses compiler passes that gather information needed for
8323           effective optimization at link time.  Second, some early
8324           optimization passes can be performed only at compile time and not
8325           at link time.
8326
8327           There are some code generation flags preserved by GCC when
8328           generating bytecodes, as they need to be used during the final
8329           link.  Currently, the following options and their settings are
8330           taken from the first object file that explicitly specifies them:
8331           -fPIC, -fpic, -fpie, -fcommon, -fexceptions, -fnon-call-exceptions,
8332           -fgnu-tm and all the -m target flags.
8333
8334           Certain ABI-changing flags are required to match in all compilation
8335           units, and trying to override this at link time with a conflicting
8336           value is ignored.  This includes options such as
8337           -freg-struct-return and -fpcc-struct-return.
8338
8339           Other options such as -ffp-contract, -fno-strict-overflow, -fwrapv,
8340           -fno-trapv or -fno-strict-aliasing are passed through to the link
8341           stage and merged conservatively for conflicting translation units.
8342           Specifically -fno-strict-overflow, -fwrapv and -fno-trapv take
8343           precedence; and for example -ffp-contract=off takes precedence over
8344           -ffp-contract=fast.  You can override them at link time.
8345
8346           If LTO encounters objects with C linkage declared with incompatible
8347           types in separate translation units to be linked together
8348           (undefined behavior according to ISO C99 6.2.7), a non-fatal
8349           diagnostic may be issued.  The behavior is still undefined at run
8350           time.  Similar diagnostics may be raised for other languages.
8351
8352           Another feature of LTO is that it is possible to apply
8353           interprocedural optimizations on files written in different
8354           languages:
8355
8356                   gcc -c -flto foo.c
8357                   g++ -c -flto bar.cc
8358                   gfortran -c -flto baz.f90
8359                   g++ -o myprog -flto -O3 foo.o bar.o baz.o -lgfortran
8360
8361           Notice that the final link is done with g++ to get the C++ runtime
8362           libraries and -lgfortran is added to get the Fortran runtime
8363           libraries.  In general, when mixing languages in LTO mode, you
8364           should use the same link command options as when mixing languages
8365           in a regular (non-LTO) compilation.
8366
8367           If object files containing GIMPLE bytecode are stored in a library
8368           archive, say libfoo.a, it is possible to extract and use them in an
8369           LTO link if you are using a linker with plugin support.  To create
8370           static libraries suitable for LTO, use gcc-ar and gcc-ranlib
8371           instead of ar and ranlib; to show the symbols of object files with
8372           GIMPLE bytecode, use gcc-nm.  Those commands require that ar,
8373           ranlib and nm have been compiled with plugin support.  At link
8374           time, use the flag -fuse-linker-plugin to ensure that the library
8375           participates in the LTO optimization process:
8376
8377                   gcc -o myprog -O2 -flto -fuse-linker-plugin a.o b.o -lfoo
8378
8379           With the linker plugin enabled, the linker extracts the needed
8380           GIMPLE files from libfoo.a and passes them on to the running GCC to
8381           make them part of the aggregated GIMPLE image to be optimized.
8382
8383           If you are not using a linker with plugin support and/or do not
8384           enable the linker plugin, then the objects inside libfoo.a are
8385           extracted and linked as usual, but they do not participate in the
8386           LTO optimization process.  In order to make a static library
8387           suitable for both LTO optimization and usual linkage, compile its
8388           object files with -flto -ffat-lto-objects.
8389
8390           Link-time optimizations do not require the presence of the whole
8391           program to operate.  If the program does not require any symbols to
8392           be exported, it is possible to combine -flto and -fwhole-program to
8393           allow the interprocedural optimizers to use more aggressive
8394           assumptions which may lead to improved optimization opportunities.
8395           Use of -fwhole-program is not needed when linker plugin is active
8396           (see -fuse-linker-plugin).
8397
8398           The current implementation of LTO makes no attempt to generate
8399           bytecode that is portable between different types of hosts.  The
8400           bytecode files are versioned and there is a strict version check,
8401           so bytecode files generated in one version of GCC do not work with
8402           an older or newer version of GCC.
8403
8404           Link-time optimization does not work well with generation of
8405           debugging information on systems other than those using a
8406           combination of ELF and DWARF.
8407
8408           If you specify the optional n, the optimization and code generation
8409           done at link time is executed in parallel using n parallel jobs by
8410           utilizing an installed make program.  The environment variable MAKE
8411           may be used to override the program used.  The default value for n
8412           is 1.
8413
8414           You can also specify -flto=jobserver to use GNU make's job server
8415           mode to determine the number of parallel jobs. This is useful when
8416           the Makefile calling GCC is already executing in parallel.  You
8417           must prepend a + to the command recipe in the parent Makefile for
8418           this to work.  This option likely only works if MAKE is GNU make.
8419
8420       -flto-partition=alg
8421           Specify the partitioning algorithm used by the link-time optimizer.
8422           The value is either 1to1 to specify a partitioning mirroring the
8423           original source files or balanced to specify partitioning into
8424           equally sized chunks (whenever possible) or max to create new
8425           partition for every symbol where possible.  Specifying none as an
8426           algorithm disables partitioning and streaming completely.  The
8427           default value is balanced. While 1to1 can be used as an workaround
8428           for various code ordering issues, the max partitioning is intended
8429           for internal testing only.  The value one specifies that exactly
8430           one partition should be used while the value none bypasses
8431           partitioning and executes the link-time optimization step directly
8432           from the WPA phase.
8433
8434       -flto-odr-type-merging
8435           Enable streaming of mangled types names of C++ types and their
8436           unification at link time.  This increases size of LTO object files,
8437           but enables diagnostics about One Definition Rule violations.
8438
8439       -flto-compression-level=n
8440           This option specifies the level of compression used for
8441           intermediate language written to LTO object files, and is only
8442           meaningful in conjunction with LTO mode (-flto).  Valid values are
8443           0 (no compression) to 9 (maximum compression).  Values outside this
8444           range are clamped to either 0 or 9.  If the option is not given, a
8445           default balanced compression setting is used.
8446
8447       -fuse-linker-plugin
8448           Enables the use of a linker plugin during link-time optimization.
8449           This option relies on plugin support in the linker, which is
8450           available in gold or in GNU ld 2.21 or newer.
8451
8452           This option enables the extraction of object files with GIMPLE
8453           bytecode out of library archives. This improves the quality of
8454           optimization by exposing more code to the link-time optimizer.
8455           This information specifies what symbols can be accessed externally
8456           (by non-LTO object or during dynamic linking).  Resulting code
8457           quality improvements on binaries (and shared libraries that use
8458           hidden visibility) are similar to -fwhole-program.  See -flto for a
8459           description of the effect of this flag and how to use it.
8460
8461           This option is enabled by default when LTO support in GCC is
8462           enabled and GCC was configured for use with a linker supporting
8463           plugins (GNU ld 2.21 or newer or gold).
8464
8465       -ffat-lto-objects
8466           Fat LTO objects are object files that contain both the intermediate
8467           language and the object code. This makes them usable for both LTO
8468           linking and normal linking. This option is effective only when
8469           compiling with -flto and is ignored at link time.
8470
8471           -fno-fat-lto-objects improves compilation time over plain LTO, but
8472           requires the complete toolchain to be aware of LTO. It requires a
8473           linker with linker plugin support for basic functionality.
8474           Additionally, nm, ar and ranlib need to support linker plugins to
8475           allow a full-featured build environment (capable of building static
8476           libraries etc).  GCC provides the gcc-ar, gcc-nm, gcc-ranlib
8477           wrappers to pass the right options to these tools. With non fat LTO
8478           makefiles need to be modified to use them.
8479
8480           Note that modern binutils provide plugin auto-load mechanism.
8481           Installing the linker plugin into $libdir/bfd-plugins has the same
8482           effect as usage of the command wrappers (gcc-ar, gcc-nm and gcc-
8483           ranlib).
8484
8485           The default is -fno-fat-lto-objects on targets with linker plugin
8486           support.
8487
8488       -fcompare-elim
8489           After register allocation and post-register allocation instruction
8490           splitting, identify arithmetic instructions that compute processor
8491           flags similar to a comparison operation based on that arithmetic.
8492           If possible, eliminate the explicit comparison operation.
8493
8494           This pass only applies to certain targets that cannot explicitly
8495           represent the comparison operation before register allocation is
8496           complete.
8497
8498           Enabled at levels -O, -O2, -O3, -Os.
8499
8500       -fcprop-registers
8501           After register allocation and post-register allocation instruction
8502           splitting, perform a copy-propagation pass to try to reduce
8503           scheduling dependencies and occasionally eliminate the copy.
8504
8505           Enabled at levels -O, -O2, -O3, -Os.
8506
8507       -fprofile-correction
8508           Profiles collected using an instrumented binary for multi-threaded
8509           programs may be inconsistent due to missed counter updates. When
8510           this option is specified, GCC uses heuristics to correct or smooth
8511           out such inconsistencies. By default, GCC emits an error message
8512           when an inconsistent profile is detected.
8513
8514           This option is enabled by -fauto-profile.
8515
8516       -fprofile-use
8517       -fprofile-use=path
8518           Enable profile feedback-directed optimizations, and the following
8519           optimizations, many of which are generally profitable only with
8520           profile feedback available:
8521
8522           -fbranch-probabilities  -fprofile-values -funroll-loops
8523           -fpeel-loops  -ftracer  -fvpt -finline-functions  -fipa-cp
8524           -fipa-cp-clone  -fipa-bit-cp -fpredictive-commoning  -fsplit-loops
8525           -funswitch-loops -fgcse-after-reload  -ftree-loop-vectorize
8526           -ftree-slp-vectorize -fvect-cost-model=dynamic
8527           -ftree-loop-distribute-patterns -fprofile-reorder-functions
8528
8529           Before you can use this option, you must first generate profiling
8530           information.
8531
8532           By default, GCC emits an error message if the feedback profiles do
8533           not match the source code.  This error can be turned into a warning
8534           by using -Wno-error=coverage-mismatch.  Note this may result in
8535           poorly optimized code.  Additionally, by default, GCC also emits a
8536           warning message if the feedback profiles do not exist (see
8537           -Wmissing-profile).
8538
8539           If path is specified, GCC looks at the path to find the profile
8540           feedback data files. See -fprofile-dir.
8541
8542       -fauto-profile
8543       -fauto-profile=path
8544           Enable sampling-based feedback-directed optimizations, and the
8545           following optimizations, many of which are generally profitable
8546           only with profile feedback available:
8547
8548           -fbranch-probabilities  -fprofile-values -funroll-loops
8549           -fpeel-loops  -ftracer  -fvpt -finline-functions  -fipa-cp
8550           -fipa-cp-clone  -fipa-bit-cp -fpredictive-commoning  -fsplit-loops
8551           -funswitch-loops -fgcse-after-reload  -ftree-loop-vectorize
8552           -ftree-slp-vectorize -fvect-cost-model=dynamic
8553           -ftree-loop-distribute-patterns -fprofile-correction
8554
8555           path is the name of a file containing AutoFDO profile information.
8556           If omitted, it defaults to fbdata.afdo in the current directory.
8557
8558           Producing an AutoFDO profile data file requires running your
8559           program with the perf utility on a supported GNU/Linux target
8560           system.  For more information, see <https://perf.wiki.kernel.org/>.
8561
8562           E.g.
8563
8564                   perf record -e br_inst_retired:near_taken -b -o perf.data \
8565                       -- your_program
8566
8567           Then use the create_gcov tool to convert the raw profile data to a
8568           format that can be used by GCC.  You must also supply the
8569           unstripped binary for your program to this tool.  See
8570           <https://github.com/google/autofdo>.
8571
8572           E.g.
8573
8574                   create_gcov --binary=your_program.unstripped --profile=perf.data \
8575                       --gcov=profile.afdo
8576
8577       The following options control compiler behavior regarding floating-
8578       point arithmetic.  These options trade off between speed and
8579       correctness.  All must be specifically enabled.
8580
8581       -ffloat-store
8582           Do not store floating-point variables in registers, and inhibit
8583           other options that might change whether a floating-point value is
8584           taken from a register or memory.
8585
8586           This option prevents undesirable excess precision on machines such
8587           as the 68000 where the floating registers (of the 68881) keep more
8588           precision than a "double" is supposed to have.  Similarly for the
8589           x86 architecture.  For most programs, the excess precision does
8590           only good, but a few programs rely on the precise definition of
8591           IEEE floating point.  Use -ffloat-store for such programs, after
8592           modifying them to store all pertinent intermediate computations
8593           into variables.
8594
8595       -fexcess-precision=style
8596           This option allows further control over excess precision on
8597           machines where floating-point operations occur in a format with
8598           more precision or range than the IEEE standard and interchange
8599           floating-point types.  By default, -fexcess-precision=fast is in
8600           effect; this means that operations may be carried out in a wider
8601           precision than the types specified in the source if that would
8602           result in faster code, and it is unpredictable when rounding to the
8603           types specified in the source code takes place.  When compiling C,
8604           if -fexcess-precision=standard is specified then excess precision
8605           follows the rules specified in ISO C99; in particular, both casts
8606           and assignments cause values to be rounded to their semantic types
8607           (whereas -ffloat-store only affects assignments).  This option is
8608           enabled by default for C if a strict conformance option such as
8609           -std=c99 is used.  -ffast-math enables -fexcess-precision=fast by
8610           default regardless of whether a strict conformance option is used.
8611
8612           -fexcess-precision=standard is not implemented for languages other
8613           than C.  On the x86, it has no effect if -mfpmath=sse or
8614           -mfpmath=sse+387 is specified; in the former case, IEEE semantics
8615           apply without excess precision, and in the latter, rounding is
8616           unpredictable.
8617
8618       -ffast-math
8619           Sets the options -fno-math-errno, -funsafe-math-optimizations,
8620           -ffinite-math-only, -fno-rounding-math, -fno-signaling-nans,
8621           -fcx-limited-range and -fexcess-precision=fast.
8622
8623           This option causes the preprocessor macro "__FAST_MATH__" to be
8624           defined.
8625
8626           This option is not turned on by any -O option besides -Ofast since
8627           it can result in incorrect output for programs that depend on an
8628           exact implementation of IEEE or ISO rules/specifications for math
8629           functions. It may, however, yield faster code for programs that do
8630           not require the guarantees of these specifications.
8631
8632       -fno-math-errno
8633           Do not set "errno" after calling math functions that are executed
8634           with a single instruction, e.g., "sqrt".  A program that relies on
8635           IEEE exceptions for math error handling may want to use this flag
8636           for speed while maintaining IEEE arithmetic compatibility.
8637
8638           This option is not turned on by any -O option since it can result
8639           in incorrect output for programs that depend on an exact
8640           implementation of IEEE or ISO rules/specifications for math
8641           functions. It may, however, yield faster code for programs that do
8642           not require the guarantees of these specifications.
8643
8644           The default is -fmath-errno.
8645
8646           On Darwin systems, the math library never sets "errno".  There is
8647           therefore no reason for the compiler to consider the possibility
8648           that it might, and -fno-math-errno is the default.
8649
8650       -funsafe-math-optimizations
8651           Allow optimizations for floating-point arithmetic that (a) assume
8652           that arguments and results are valid and (b) may violate IEEE or
8653           ANSI standards.  When used at link time, it may include libraries
8654           or startup files that change the default FPU control word or other
8655           similar optimizations.
8656
8657           This option is not turned on by any -O option since it can result
8658           in incorrect output for programs that depend on an exact
8659           implementation of IEEE or ISO rules/specifications for math
8660           functions. It may, however, yield faster code for programs that do
8661           not require the guarantees of these specifications.  Enables
8662           -fno-signed-zeros, -fno-trapping-math, -fassociative-math and
8663           -freciprocal-math.
8664
8665           The default is -fno-unsafe-math-optimizations.
8666
8667       -fassociative-math
8668           Allow re-association of operands in series of floating-point
8669           operations.  This violates the ISO C and C++ language standard by
8670           possibly changing computation result.  NOTE: re-ordering may change
8671           the sign of zero as well as ignore NaNs and inhibit or create
8672           underflow or overflow (and thus cannot be used on code that relies
8673           on rounding behavior like "(x + 2**52) - 2**52".  May also reorder
8674           floating-point comparisons and thus may not be used when ordered
8675           comparisons are required.  This option requires that both
8676           -fno-signed-zeros and -fno-trapping-math be in effect.  Moreover,
8677           it doesn't make much sense with -frounding-math. For Fortran the
8678           option is automatically enabled when both -fno-signed-zeros and
8679           -fno-trapping-math are in effect.
8680
8681           The default is -fno-associative-math.
8682
8683       -freciprocal-math
8684           Allow the reciprocal of a value to be used instead of dividing by
8685           the value if this enables optimizations.  For example "x / y" can
8686           be replaced with "x * (1/y)", which is useful if "(1/y)" is subject
8687           to common subexpression elimination.  Note that this loses
8688           precision and increases the number of flops operating on the value.
8689
8690           The default is -fno-reciprocal-math.
8691
8692       -ffinite-math-only
8693           Allow optimizations for floating-point arithmetic that assume that
8694           arguments and results are not NaNs or +-Infs.
8695
8696           This option is not turned on by any -O option since it can result
8697           in incorrect output for programs that depend on an exact
8698           implementation of IEEE or ISO rules/specifications for math
8699           functions. It may, however, yield faster code for programs that do
8700           not require the guarantees of these specifications.
8701
8702           The default is -fno-finite-math-only.
8703
8704       -fno-signed-zeros
8705           Allow optimizations for floating-point arithmetic that ignore the
8706           signedness of zero.  IEEE arithmetic specifies the behavior of
8707           distinct +0.0 and -0.0 values, which then prohibits simplification
8708           of expressions such as x+0.0 or 0.0*x (even with
8709           -ffinite-math-only).  This option implies that the sign of a zero
8710           result isn't significant.
8711
8712           The default is -fsigned-zeros.
8713
8714       -fno-trapping-math
8715           Compile code assuming that floating-point operations cannot
8716           generate user-visible traps.  These traps include division by zero,
8717           overflow, underflow, inexact result and invalid operation.  This
8718           option requires that -fno-signaling-nans be in effect.  Setting
8719           this option may allow faster code if one relies on "non-stop" IEEE
8720           arithmetic, for example.
8721
8722           This option should never be turned on by any -O option since it can
8723           result in incorrect output for programs that depend on an exact
8724           implementation of IEEE or ISO rules/specifications for math
8725           functions.
8726
8727           The default is -ftrapping-math.
8728
8729       -frounding-math
8730           Disable transformations and optimizations that assume default
8731           floating-point rounding behavior.  This is round-to-zero for all
8732           floating point to integer conversions, and round-to-nearest for all
8733           other arithmetic truncations.  This option should be specified for
8734           programs that change the FP rounding mode dynamically, or that may
8735           be executed with a non-default rounding mode.  This option disables
8736           constant folding of floating-point expressions at compile time
8737           (which may be affected by rounding mode) and arithmetic
8738           transformations that are unsafe in the presence of sign-dependent
8739           rounding modes.
8740
8741           The default is -fno-rounding-math.
8742
8743           This option is experimental and does not currently guarantee to
8744           disable all GCC optimizations that are affected by rounding mode.
8745           Future versions of GCC may provide finer control of this setting
8746           using C99's "FENV_ACCESS" pragma.  This command-line option will be
8747           used to specify the default state for "FENV_ACCESS".
8748
8749       -fsignaling-nans
8750           Compile code assuming that IEEE signaling NaNs may generate user-
8751           visible traps during floating-point operations.  Setting this
8752           option disables optimizations that may change the number of
8753           exceptions visible with signaling NaNs.  This option implies
8754           -ftrapping-math.
8755
8756           This option causes the preprocessor macro "__SUPPORT_SNAN__" to be
8757           defined.
8758
8759           The default is -fno-signaling-nans.
8760
8761           This option is experimental and does not currently guarantee to
8762           disable all GCC optimizations that affect signaling NaN behavior.
8763
8764       -fno-fp-int-builtin-inexact
8765           Do not allow the built-in functions "ceil", "floor", "round" and
8766           "trunc", and their "float" and "long double" variants, to generate
8767           code that raises the "inexact" floating-point exception for
8768           noninteger arguments.  ISO C99 and C11 allow these functions to
8769           raise the "inexact" exception, but ISO/IEC TS 18661-1:2014, the C
8770           bindings to IEEE 754-2008, does not allow these functions to do so.
8771
8772           The default is -ffp-int-builtin-inexact, allowing the exception to
8773           be raised.  This option does nothing unless -ftrapping-math is in
8774           effect.
8775
8776           Even if -fno-fp-int-builtin-inexact is used, if the functions
8777           generate a call to a library function then the "inexact" exception
8778           may be raised if the library implementation does not follow TS
8779           18661.
8780
8781       -fsingle-precision-constant
8782           Treat floating-point constants as single precision instead of
8783           implicitly converting them to double-precision constants.
8784
8785       -fcx-limited-range
8786           When enabled, this option states that a range reduction step is not
8787           needed when performing complex division.  Also, there is no
8788           checking whether the result of a complex multiplication or division
8789           is "NaN + I*NaN", with an attempt to rescue the situation in that
8790           case.  The default is -fno-cx-limited-range, but is enabled by
8791           -ffast-math.
8792
8793           This option controls the default setting of the ISO C99
8794           "CX_LIMITED_RANGE" pragma.  Nevertheless, the option applies to all
8795           languages.
8796
8797       -fcx-fortran-rules
8798           Complex multiplication and division follow Fortran rules.  Range
8799           reduction is done as part of complex division, but there is no
8800           checking whether the result of a complex multiplication or division
8801           is "NaN + I*NaN", with an attempt to rescue the situation in that
8802           case.
8803
8804           The default is -fno-cx-fortran-rules.
8805
8806       The following options control optimizations that may improve
8807       performance, but are not enabled by any -O options.  This section
8808       includes experimental options that may produce broken code.
8809
8810       -fbranch-probabilities
8811           After running a program compiled with -fprofile-arcs, you can
8812           compile it a second time using -fbranch-probabilities, to improve
8813           optimizations based on the number of times each branch was taken.
8814           When a program compiled with -fprofile-arcs exits, it saves arc
8815           execution counts to a file called sourcename.gcda for each source
8816           file.  The information in this data file is very dependent on the
8817           structure of the generated code, so you must use the same source
8818           code and the same optimization options for both compilations.
8819
8820           With -fbranch-probabilities, GCC puts a REG_BR_PROB note on each
8821           JUMP_INSN and CALL_INSN.  These can be used to improve
8822           optimization.  Currently, they are only used in one place: in
8823           reorg.c, instead of guessing which path a branch is most likely to
8824           take, the REG_BR_PROB values are used to exactly determine which
8825           path is taken more often.
8826
8827           Enabled by -fprofile-use and -fauto-profile.
8828
8829       -fprofile-values
8830           If combined with -fprofile-arcs, it adds code so that some data
8831           about values of expressions in the program is gathered.
8832
8833           With -fbranch-probabilities, it reads back the data gathered from
8834           profiling values of expressions for usage in optimizations.
8835
8836           Enabled by -fprofile-generate, -fprofile-use, and -fauto-profile.
8837
8838       -fprofile-reorder-functions
8839           Function reordering based on profile instrumentation collects first
8840           time of execution of a function and orders these functions in
8841           ascending order.
8842
8843           Enabled with -fprofile-use.
8844
8845       -fvpt
8846           If combined with -fprofile-arcs, this option instructs the compiler
8847           to add code to gather information about values of expressions.
8848
8849           With -fbranch-probabilities, it reads back the data gathered and
8850           actually performs the optimizations based on them.  Currently the
8851           optimizations include specialization of division operations using
8852           the knowledge about the value of the denominator.
8853
8854           Enabled with -fprofile-use and -fauto-profile.
8855
8856       -frename-registers
8857           Attempt to avoid false dependencies in scheduled code by making use
8858           of registers left over after register allocation.  This
8859           optimization most benefits processors with lots of registers.
8860           Depending on the debug information format adopted by the target,
8861           however, it can make debugging impossible, since variables no
8862           longer stay in a "home register".
8863
8864           Enabled by default with -funroll-loops.
8865
8866       -fschedule-fusion
8867           Performs a target dependent pass over the instruction stream to
8868           schedule instructions of same type together because target machine
8869           can execute them more efficiently if they are adjacent to each
8870           other in the instruction flow.
8871
8872           Enabled at levels -O2, -O3, -Os.
8873
8874       -ftracer
8875           Perform tail duplication to enlarge superblock size.  This
8876           transformation simplifies the control flow of the function allowing
8877           other optimizations to do a better job.
8878
8879           Enabled by -fprofile-use and -fauto-profile.
8880
8881       -funroll-loops
8882           Unroll loops whose number of iterations can be determined at
8883           compile time or upon entry to the loop.  -funroll-loops implies
8884           -frerun-cse-after-loop, -fweb and -frename-registers.  It also
8885           turns on complete loop peeling (i.e. complete removal of loops with
8886           a small constant number of iterations).  This option makes code
8887           larger, and may or may not make it run faster.
8888
8889           Enabled by -fprofile-use and -fauto-profile.
8890
8891       -funroll-all-loops
8892           Unroll all loops, even if their number of iterations is uncertain
8893           when the loop is entered.  This usually makes programs run more
8894           slowly.  -funroll-all-loops implies the same options as
8895           -funroll-loops.
8896
8897       -fpeel-loops
8898           Peels loops for which there is enough information that they do not
8899           roll much (from profile feedback or static analysis).  It also
8900           turns on complete loop peeling (i.e. complete removal of loops with
8901           small constant number of iterations).
8902
8903           Enabled by -O3, -fprofile-use, and -fauto-profile.
8904
8905       -fmove-loop-invariants
8906           Enables the loop invariant motion pass in the RTL loop optimizer.
8907           Enabled at level -O1 and higher, except for -Og.
8908
8909       -fsplit-loops
8910           Split a loop into two if it contains a condition that's always true
8911           for one side of the iteration space and false for the other.
8912
8913           Enabled by -fprofile-use and -fauto-profile.
8914
8915       -funswitch-loops
8916           Move branches with loop invariant conditions out of the loop, with
8917           duplicates of the loop on both branches (modified according to
8918           result of the condition).
8919
8920           Enabled by -fprofile-use and -fauto-profile.
8921
8922       -fversion-loops-for-strides
8923           If a loop iterates over an array with a variable stride, create
8924           another version of the loop that assumes the stride is always one.
8925           For example:
8926
8927                   for (int i = 0; i < n; ++i)
8928                     x[i * stride] = ...;
8929
8930           becomes:
8931
8932                   if (stride == 1)
8933                     for (int i = 0; i < n; ++i)
8934                       x[i] = ...;
8935                   else
8936                     for (int i = 0; i < n; ++i)
8937                       x[i * stride] = ...;
8938
8939           This is particularly useful for assumed-shape arrays in Fortran
8940           where (for example) it allows better vectorization assuming
8941           contiguous accesses.  This flag is enabled by default at -O3.  It
8942           is also enabled by -fprofile-use and -fauto-profile.
8943
8944       -ffunction-sections
8945       -fdata-sections
8946           Place each function or data item into its own section in the output
8947           file if the target supports arbitrary sections.  The name of the
8948           function or the name of the data item determines the section's name
8949           in the output file.
8950
8951           Use these options on systems where the linker can perform
8952           optimizations to improve locality of reference in the instruction
8953           space.  Most systems using the ELF object format have linkers with
8954           such optimizations.  On AIX, the linker rearranges sections
8955           (CSECTs) based on the call graph.  The performance impact varies.
8956
8957           Together with a linker garbage collection (linker --gc-sections
8958           option) these options may lead to smaller statically-linked
8959           executables (after stripping).
8960
8961           On ELF/DWARF systems these options do not degenerate the quality of
8962           the debug information.  There could be issues with other object
8963           files/debug info formats.
8964
8965           Only use these options when there are significant benefits from
8966           doing so.  When you specify these options, the assembler and linker
8967           create larger object and executable files and are also slower.
8968           These options affect code generation.  They prevent optimizations
8969           by the compiler and assembler using relative locations inside a
8970           translation unit since the locations are unknown until link time.
8971           An example of such an optimization is relaxing calls to short call
8972           instructions.
8973
8974       -fbranch-target-load-optimize
8975           Perform branch target register load optimization before prologue /
8976           epilogue threading.  The use of target registers can typically be
8977           exposed only during reload, thus hoisting loads out of loops and
8978           doing inter-block scheduling needs a separate optimization pass.
8979
8980       -fbranch-target-load-optimize2
8981           Perform branch target register load optimization after prologue /
8982           epilogue threading.
8983
8984       -fbtr-bb-exclusive
8985           When performing branch target register load optimization, don't
8986           reuse branch target registers within any basic block.
8987
8988       -fstdarg-opt
8989           Optimize the prologue of variadic argument functions with respect
8990           to usage of those arguments.
8991
8992       -fsection-anchors
8993           Try to reduce the number of symbolic address calculations by using
8994           shared "anchor" symbols to address nearby objects.  This
8995           transformation can help to reduce the number of GOT entries and GOT
8996           accesses on some targets.
8997
8998           For example, the implementation of the following function "foo":
8999
9000                   static int a, b, c;
9001                   int foo (void) { return a + b + c; }
9002
9003           usually calculates the addresses of all three variables, but if you
9004           compile it with -fsection-anchors, it accesses the variables from a
9005           common anchor point instead.  The effect is similar to the
9006           following pseudocode (which isn't valid C):
9007
9008                   int foo (void)
9009                   {
9010                     register int *xr = &x;
9011                     return xr[&a - &x] + xr[&b - &x] + xr[&c - &x];
9012                   }
9013
9014           Not all targets support this option.
9015
9016       --param name=value
9017           In some places, GCC uses various constants to control the amount of
9018           optimization that is done.  For example, GCC does not inline
9019           functions that contain more than a certain number of instructions.
9020           You can control some of these constants on the command line using
9021           the --param option.
9022
9023           The names of specific parameters, and the meaning of the values,
9024           are tied to the internals of the compiler, and are subject to
9025           change without notice in future releases.
9026
9027           In order to get minimal, maximal and default value of a parameter,
9028           one can use --help=param -Q options.
9029
9030           In each case, the value is an integer.  The allowable choices for
9031           name are:
9032
9033           predictable-branch-outcome
9034               When branch is predicted to be taken with probability lower
9035               than this threshold (in percent), then it is considered well
9036               predictable.
9037
9038           max-rtl-if-conversion-insns
9039               RTL if-conversion tries to remove conditional branches around a
9040               block and replace them with conditionally executed
9041               instructions.  This parameter gives the maximum number of
9042               instructions in a block which should be considered for if-
9043               conversion.  The compiler will also use other heuristics to
9044               decide whether if-conversion is likely to be profitable.
9045
9046           max-rtl-if-conversion-predictable-cost
9047           max-rtl-if-conversion-unpredictable-cost
9048               RTL if-conversion will try to remove conditional branches
9049               around a block and replace them with conditionally executed
9050               instructions.  These parameters give the maximum permissible
9051               cost for the sequence that would be generated by if-conversion
9052               depending on whether the branch is statically determined to be
9053               predictable or not.  The units for this parameter are the same
9054               as those for the GCC internal seq_cost metric.  The compiler
9055               will try to provide a reasonable default for this parameter
9056               using the BRANCH_COST target macro.
9057
9058           max-crossjump-edges
9059               The maximum number of incoming edges to consider for cross-
9060               jumping.  The algorithm used by -fcrossjumping is O(N^2) in the
9061               number of edges incoming to each block.  Increasing values mean
9062               more aggressive optimization, making the compilation time
9063               increase with probably small improvement in executable size.
9064
9065           min-crossjump-insns
9066               The minimum number of instructions that must be matched at the
9067               end of two blocks before cross-jumping is performed on them.
9068               This value is ignored in the case where all instructions in the
9069               block being cross-jumped from are matched.
9070
9071           max-grow-copy-bb-insns
9072               The maximum code size expansion factor when copying basic
9073               blocks instead of jumping.  The expansion is relative to a jump
9074               instruction.
9075
9076           max-goto-duplication-insns
9077               The maximum number of instructions to duplicate to a block that
9078               jumps to a computed goto.  To avoid O(N^2) behavior in a number
9079               of passes, GCC factors computed gotos early in the compilation
9080               process, and unfactors them as late as possible.  Only computed
9081               jumps at the end of a basic blocks with no more than max-goto-
9082               duplication-insns are unfactored.
9083
9084           max-delay-slot-insn-search
9085               The maximum number of instructions to consider when looking for
9086               an instruction to fill a delay slot.  If more than this
9087               arbitrary number of instructions are searched, the time savings
9088               from filling the delay slot are minimal, so stop searching.
9089               Increasing values mean more aggressive optimization, making the
9090               compilation time increase with probably small improvement in
9091               execution time.
9092
9093           max-delay-slot-live-search
9094               When trying to fill delay slots, the maximum number of
9095               instructions to consider when searching for a block with valid
9096               live register information.  Increasing this arbitrarily chosen
9097               value means more aggressive optimization, increasing the
9098               compilation time.  This parameter should be removed when the
9099               delay slot code is rewritten to maintain the control-flow
9100               graph.
9101
9102           max-gcse-memory
9103               The approximate maximum amount of memory that can be allocated
9104               in order to perform the global common subexpression elimination
9105               optimization.  If more memory than specified is required, the
9106               optimization is not done.
9107
9108           max-gcse-insertion-ratio
9109               If the ratio of expression insertions to deletions is larger
9110               than this value for any expression, then RTL PRE inserts or
9111               removes the expression and thus leaves partially redundant
9112               computations in the instruction stream.
9113
9114           max-pending-list-length
9115               The maximum number of pending dependencies scheduling allows
9116               before flushing the current state and starting over.  Large
9117               functions with few branches or calls can create excessively
9118               large lists which needlessly consume memory and resources.
9119
9120           max-modulo-backtrack-attempts
9121               The maximum number of backtrack attempts the scheduler should
9122               make when modulo scheduling a loop.  Larger values can
9123               exponentially increase compilation time.
9124
9125           max-inline-insns-single
9126               Several parameters control the tree inliner used in GCC.  This
9127               number sets the maximum number of instructions (counted in
9128               GCC's internal representation) in a single function that the
9129               tree inliner considers for inlining.  This only affects
9130               functions declared inline and methods implemented in a class
9131               declaration (C++).
9132
9133           max-inline-insns-auto
9134               When you use -finline-functions (included in -O3), a lot of
9135               functions that would otherwise not be considered for inlining
9136               by the compiler are investigated.  To those functions, a
9137               different (more restrictive) limit compared to functions
9138               declared inline can be applied.
9139
9140           max-inline-insns-small
9141               This is bound applied to calls which are considered relevant
9142               with -finline-small-functions.
9143
9144           max-inline-insns-size
9145               This is bound applied to calls which are optimized for size.
9146               Small growth may be desirable to anticipate optimization
9147               oppurtunities exposed by inlining.
9148
9149           uninlined-function-insns
9150               Number of instructions accounted by inliner for function
9151               overhead such as function prologue and epilogue.
9152
9153           uninlined-function-time
9154               Extra time accounted by inliner for function overhead such as
9155               time needed to execute function prologue and epilogue
9156
9157           uninlined-thunk-insns
9158           uninlined-thunk-time
9159               Same as --param uninlined-function-insns and --param uninlined-
9160               function-time but applied to function thunks
9161
9162           inline-min-speedup
9163               When estimated performance improvement of caller + callee
9164               runtime exceeds this threshold (in percent), the function can
9165               be inlined regardless of the limit on --param max-inline-insns-
9166               single and --param max-inline-insns-auto.
9167
9168           large-function-insns
9169               The limit specifying really large functions.  For functions
9170               larger than this limit after inlining, inlining is constrained
9171               by --param large-function-growth.  This parameter is useful
9172               primarily to avoid extreme compilation time caused by non-
9173               linear algorithms used by the back end.
9174
9175           large-function-growth
9176               Specifies maximal growth of large function caused by inlining
9177               in percents.  For example, parameter value 100 limits large
9178               function growth to 2.0 times the original size.
9179
9180           large-unit-insns
9181               The limit specifying large translation unit.  Growth caused by
9182               inlining of units larger than this limit is limited by --param
9183               inline-unit-growth.  For small units this might be too tight.
9184               For example, consider a unit consisting of function A that is
9185               inline and B that just calls A three times.  If B is small
9186               relative to A, the growth of unit is 300\% and yet such
9187               inlining is very sane.  For very large units consisting of
9188               small inlineable functions, however, the overall unit growth
9189               limit is needed to avoid exponential explosion of code size.
9190               Thus for smaller units, the size is increased to --param large-
9191               unit-insns before applying --param inline-unit-growth.
9192
9193           inline-unit-growth
9194               Specifies maximal overall growth of the compilation unit caused
9195               by inlining.  For example, parameter value 20 limits unit
9196               growth to 1.2 times the original size. Cold functions (either
9197               marked cold via an attribute or by profile feedback) are not
9198               accounted into the unit size.
9199
9200           ipcp-unit-growth
9201               Specifies maximal overall growth of the compilation unit caused
9202               by interprocedural constant propagation.  For example,
9203               parameter value 10 limits unit growth to 1.1 times the original
9204               size.
9205
9206           large-stack-frame
9207               The limit specifying large stack frames.  While inlining the
9208               algorithm is trying to not grow past this limit too much.
9209
9210           large-stack-frame-growth
9211               Specifies maximal growth of large stack frames caused by
9212               inlining in percents.  For example, parameter value 1000 limits
9213               large stack frame growth to 11 times the original size.
9214
9215           max-inline-insns-recursive
9216           max-inline-insns-recursive-auto
9217               Specifies the maximum number of instructions an out-of-line
9218               copy of a self-recursive inline function can grow into by
9219               performing recursive inlining.
9220
9221               --param max-inline-insns-recursive applies to functions
9222               declared inline.  For functions not declared inline, recursive
9223               inlining happens only when -finline-functions (included in -O3)
9224               is enabled; --param max-inline-insns-recursive-auto applies
9225               instead.
9226
9227           max-inline-recursive-depth
9228           max-inline-recursive-depth-auto
9229               Specifies the maximum recursion depth used for recursive
9230               inlining.
9231
9232               --param max-inline-recursive-depth applies to functions
9233               declared inline.  For functions not declared inline, recursive
9234               inlining happens only when -finline-functions (included in -O3)
9235               is enabled; --param max-inline-recursive-depth-auto applies
9236               instead.
9237
9238           min-inline-recursive-probability
9239               Recursive inlining is profitable only for function having deep
9240               recursion in average and can hurt for function having little
9241               recursion depth by increasing the prologue size or complexity
9242               of function body to other optimizers.
9243
9244               When profile feedback is available (see -fprofile-generate) the
9245               actual recursion depth can be guessed from the probability that
9246               function recurses via a given call expression.  This parameter
9247               limits inlining only to call expressions whose probability
9248               exceeds the given threshold (in percents).
9249
9250           early-inlining-insns
9251               Specify growth that the early inliner can make.  In effect it
9252               increases the amount of inlining for code having a large
9253               abstraction penalty.
9254
9255           max-early-inliner-iterations
9256               Limit of iterations of the early inliner.  This basically
9257               bounds the number of nested indirect calls the early inliner
9258               can resolve.  Deeper chains are still handled by late inlining.
9259
9260           comdat-sharing-probability
9261               Probability (in percent) that C++ inline function with comdat
9262               visibility are shared across multiple compilation units.
9263
9264           profile-func-internal-id
9265               A parameter to control whether to use function internal id in
9266               profile database lookup. If the value is 0, the compiler uses
9267               an id that is based on function assembler name and filename,
9268               which makes old profile data more tolerant to source changes
9269               such as function reordering etc.
9270
9271           min-vect-loop-bound
9272               The minimum number of iterations under which loops are not
9273               vectorized when -ftree-vectorize is used.  The number of
9274               iterations after vectorization needs to be greater than the
9275               value specified by this option to allow vectorization.
9276
9277           gcse-cost-distance-ratio
9278               Scaling factor in calculation of maximum distance an expression
9279               can be moved by GCSE optimizations.  This is currently
9280               supported only in the code hoisting pass.  The bigger the
9281               ratio, the more aggressive code hoisting is with simple
9282               expressions, i.e., the expressions that have cost less than
9283               gcse-unrestricted-cost.  Specifying 0 disables hoisting of
9284               simple expressions.
9285
9286           gcse-unrestricted-cost
9287               Cost, roughly measured as the cost of a single typical machine
9288               instruction, at which GCSE optimizations do not constrain the
9289               distance an expression can travel.  This is currently supported
9290               only in the code hoisting pass.  The lesser the cost, the more
9291               aggressive code hoisting is.  Specifying 0 allows all
9292               expressions to travel unrestricted distances.
9293
9294           max-hoist-depth
9295               The depth of search in the dominator tree for expressions to
9296               hoist.  This is used to avoid quadratic behavior in hoisting
9297               algorithm.  The value of 0 does not limit on the search, but
9298               may slow down compilation of huge functions.
9299
9300           max-tail-merge-comparisons
9301               The maximum amount of similar bbs to compare a bb with.  This
9302               is used to avoid quadratic behavior in tree tail merging.
9303
9304           max-tail-merge-iterations
9305               The maximum amount of iterations of the pass over the function.
9306               This is used to limit compilation time in tree tail merging.
9307
9308           store-merging-allow-unaligned
9309               Allow the store merging pass to introduce unaligned stores if
9310               it is legal to do so.
9311
9312           max-stores-to-merge
9313               The maximum number of stores to attempt to merge into wider
9314               stores in the store merging pass.
9315
9316           max-unrolled-insns
9317               The maximum number of instructions that a loop may have to be
9318               unrolled.  If a loop is unrolled, this parameter also
9319               determines how many times the loop code is unrolled.
9320
9321           max-average-unrolled-insns
9322               The maximum number of instructions biased by probabilities of
9323               their execution that a loop may have to be unrolled.  If a loop
9324               is unrolled, this parameter also determines how many times the
9325               loop code is unrolled.
9326
9327           max-unroll-times
9328               The maximum number of unrollings of a single loop.
9329
9330           max-peeled-insns
9331               The maximum number of instructions that a loop may have to be
9332               peeled.  If a loop is peeled, this parameter also determines
9333               how many times the loop code is peeled.
9334
9335           max-peel-times
9336               The maximum number of peelings of a single loop.
9337
9338           max-peel-branches
9339               The maximum number of branches on the hot path through the
9340               peeled sequence.
9341
9342           max-completely-peeled-insns
9343               The maximum number of insns of a completely peeled loop.
9344
9345           max-completely-peel-times
9346               The maximum number of iterations of a loop to be suitable for
9347               complete peeling.
9348
9349           max-completely-peel-loop-nest-depth
9350               The maximum depth of a loop nest suitable for complete peeling.
9351
9352           max-unswitch-insns
9353               The maximum number of insns of an unswitched loop.
9354
9355           max-unswitch-level
9356               The maximum number of branches unswitched in a single loop.
9357
9358           lim-expensive
9359               The minimum cost of an expensive expression in the loop
9360               invariant motion.
9361
9362           iv-consider-all-candidates-bound
9363               Bound on number of candidates for induction variables, below
9364               which all candidates are considered for each use in induction
9365               variable optimizations.  If there are more candidates than
9366               this, only the most relevant ones are considered to avoid
9367               quadratic time complexity.
9368
9369           iv-max-considered-uses
9370               The induction variable optimizations give up on loops that
9371               contain more induction variable uses.
9372
9373           iv-always-prune-cand-set-bound
9374               If the number of candidates in the set is smaller than this
9375               value, always try to remove unnecessary ivs from the set when
9376               adding a new one.
9377
9378           avg-loop-niter
9379               Average number of iterations of a loop.
9380
9381           dse-max-object-size
9382               Maximum size (in bytes) of objects tracked bytewise by dead
9383               store elimination.  Larger values may result in larger
9384               compilation times.
9385
9386           dse-max-alias-queries-per-store
9387               Maximum number of queries into the alias oracle per store.
9388               Larger values result in larger compilation times and may result
9389               in more removed dead stores.
9390
9391           scev-max-expr-size
9392               Bound on size of expressions used in the scalar evolutions
9393               analyzer.  Large expressions slow the analyzer.
9394
9395           scev-max-expr-complexity
9396               Bound on the complexity of the expressions in the scalar
9397               evolutions analyzer.  Complex expressions slow the analyzer.
9398
9399           max-tree-if-conversion-phi-args
9400               Maximum number of arguments in a PHI supported by TREE if
9401               conversion unless the loop is marked with simd pragma.
9402
9403           vect-max-version-for-alignment-checks
9404               The maximum number of run-time checks that can be performed
9405               when doing loop versioning for alignment in the vectorizer.
9406
9407           vect-max-version-for-alias-checks
9408               The maximum number of run-time checks that can be performed
9409               when doing loop versioning for alias in the vectorizer.
9410
9411           vect-max-peeling-for-alignment
9412               The maximum number of loop peels to enhance access alignment
9413               for vectorizer. Value -1 means no limit.
9414
9415           max-iterations-to-track
9416               The maximum number of iterations of a loop the brute-force
9417               algorithm for analysis of the number of iterations of the loop
9418               tries to evaluate.
9419
9420           hot-bb-count-ws-permille
9421               A basic block profile count is considered hot if it contributes
9422               to the given permillage (i.e. 0...1000) of the entire profiled
9423               execution.
9424
9425           hot-bb-frequency-fraction
9426               Select fraction of the entry block frequency of executions of
9427               basic block in function given basic block needs to have to be
9428               considered hot.
9429
9430           max-predicted-iterations
9431               The maximum number of loop iterations we predict statically.
9432               This is useful in cases where a function contains a single loop
9433               with known bound and another loop with unknown bound.  The
9434               known number of iterations is predicted correctly, while the
9435               unknown number of iterations average to roughly 10.  This means
9436               that the loop without bounds appears artificially cold relative
9437               to the other one.
9438
9439           builtin-expect-probability
9440               Control the probability of the expression having the specified
9441               value. This parameter takes a percentage (i.e. 0 ... 100) as
9442               input.
9443
9444           builtin-string-cmp-inline-length
9445               The maximum length of a constant string for a builtin string
9446               cmp call eligible for inlining.
9447
9448           align-threshold
9449               Select fraction of the maximal frequency of executions of a
9450               basic block in a function to align the basic block.
9451
9452           align-loop-iterations
9453               A loop expected to iterate at least the selected number of
9454               iterations is aligned.
9455
9456           tracer-dynamic-coverage
9457           tracer-dynamic-coverage-feedback
9458               This value is used to limit superblock formation once the given
9459               percentage of executed instructions is covered.  This limits
9460               unnecessary code size expansion.
9461
9462               The tracer-dynamic-coverage-feedback parameter is used only
9463               when profile feedback is available.  The real profiles (as
9464               opposed to statically estimated ones) are much less balanced
9465               allowing the threshold to be larger value.
9466
9467           tracer-max-code-growth
9468               Stop tail duplication once code growth has reached given
9469               percentage.  This is a rather artificial limit, as most of the
9470               duplicates are eliminated later in cross jumping, so it may be
9471               set to much higher values than is the desired code growth.
9472
9473           tracer-min-branch-ratio
9474               Stop reverse growth when the reverse probability of best edge
9475               is less than this threshold (in percent).
9476
9477           tracer-min-branch-probability
9478           tracer-min-branch-probability-feedback
9479               Stop forward growth if the best edge has probability lower than
9480               this threshold.
9481
9482               Similarly to tracer-dynamic-coverage two parameters are
9483               provided.  tracer-min-branch-probability-feedback is used for
9484               compilation with profile feedback and tracer-min-branch-
9485               probability compilation without.  The value for compilation
9486               with profile feedback needs to be more conservative (higher) in
9487               order to make tracer effective.
9488
9489           stack-clash-protection-guard-size
9490               Specify the size of the operating system provided stack guard
9491               as 2 raised to num bytes.  Higher values may reduce the number
9492               of explicit probes, but a value larger than the operating
9493               system provided guard will leave code vulnerable to stack clash
9494               style attacks.
9495
9496           stack-clash-protection-probe-interval
9497               Stack clash protection involves probing stack space as it is
9498               allocated.  This param controls the maximum distance between
9499               probes into the stack as 2 raised to num bytes.  Higher values
9500               may reduce the number of explicit probes, but a value larger
9501               than the operating system provided guard will leave code
9502               vulnerable to stack clash style attacks.
9503
9504           max-cse-path-length
9505               The maximum number of basic blocks on path that CSE considers.
9506
9507           max-cse-insns
9508               The maximum number of instructions CSE processes before
9509               flushing.
9510
9511           ggc-min-expand
9512               GCC uses a garbage collector to manage its own memory
9513               allocation.  This parameter specifies the minimum percentage by
9514               which the garbage collector's heap should be allowed to expand
9515               between collections.  Tuning this may improve compilation
9516               speed; it has no effect on code generation.
9517
9518               The default is 30% + 70% * (RAM/1GB) with an upper bound of
9519               100% when RAM >= 1GB.  If "getrlimit" is available, the notion
9520               of "RAM" is the smallest of actual RAM and "RLIMIT_DATA" or
9521               "RLIMIT_AS".  If GCC is not able to calculate RAM on a
9522               particular platform, the lower bound of 30% is used.  Setting
9523               this parameter and ggc-min-heapsize to zero causes a full
9524               collection to occur at every opportunity.  This is extremely
9525               slow, but can be useful for debugging.
9526
9527           ggc-min-heapsize
9528               Minimum size of the garbage collector's heap before it begins
9529               bothering to collect garbage.  The first collection occurs
9530               after the heap expands by ggc-min-expand% beyond ggc-min-
9531               heapsize.  Again, tuning this may improve compilation speed,
9532               and has no effect on code generation.
9533
9534               The default is the smaller of RAM/8, RLIMIT_RSS, or a limit
9535               that tries to ensure that RLIMIT_DATA or RLIMIT_AS are not
9536               exceeded, but with a lower bound of 4096 (four megabytes) and
9537               an upper bound of 131072 (128 megabytes).  If GCC is not able
9538               to calculate RAM on a particular platform, the lower bound is
9539               used.  Setting this parameter very large effectively disables
9540               garbage collection.  Setting this parameter and ggc-min-expand
9541               to zero causes a full collection to occur at every opportunity.
9542
9543           max-reload-search-insns
9544               The maximum number of instruction reload should look backward
9545               for equivalent register.  Increasing values mean more
9546               aggressive optimization, making the compilation time increase
9547               with probably slightly better performance.
9548
9549           max-cselib-memory-locations
9550               The maximum number of memory locations cselib should take into
9551               account.  Increasing values mean more aggressive optimization,
9552               making the compilation time increase with probably slightly
9553               better performance.
9554
9555           max-sched-ready-insns
9556               The maximum number of instructions ready to be issued the
9557               scheduler should consider at any given time during the first
9558               scheduling pass.  Increasing values mean more thorough
9559               searches, making the compilation time increase with probably
9560               little benefit.
9561
9562           max-sched-region-blocks
9563               The maximum number of blocks in a region to be considered for
9564               interblock scheduling.
9565
9566           max-pipeline-region-blocks
9567               The maximum number of blocks in a region to be considered for
9568               pipelining in the selective scheduler.
9569
9570           max-sched-region-insns
9571               The maximum number of insns in a region to be considered for
9572               interblock scheduling.
9573
9574           max-pipeline-region-insns
9575               The maximum number of insns in a region to be considered for
9576               pipelining in the selective scheduler.
9577
9578           min-spec-prob
9579               The minimum probability (in percents) of reaching a source
9580               block for interblock speculative scheduling.
9581
9582           max-sched-extend-regions-iters
9583               The maximum number of iterations through CFG to extend regions.
9584               A value of 0 disables region extensions.
9585
9586           max-sched-insn-conflict-delay
9587               The maximum conflict delay for an insn to be considered for
9588               speculative motion.
9589
9590           sched-spec-prob-cutoff
9591               The minimal probability of speculation success (in percents),
9592               so that speculative insns are scheduled.
9593
9594           sched-state-edge-prob-cutoff
9595               The minimum probability an edge must have for the scheduler to
9596               save its state across it.
9597
9598           sched-mem-true-dep-cost
9599               Minimal distance (in CPU cycles) between store and load
9600               targeting same memory locations.
9601
9602           selsched-max-lookahead
9603               The maximum size of the lookahead window of selective
9604               scheduling.  It is a depth of search for available
9605               instructions.
9606
9607           selsched-max-sched-times
9608               The maximum number of times that an instruction is scheduled
9609               during selective scheduling.  This is the limit on the number
9610               of iterations through which the instruction may be pipelined.
9611
9612           selsched-insns-to-rename
9613               The maximum number of best instructions in the ready list that
9614               are considered for renaming in the selective scheduler.
9615
9616           sms-min-sc
9617               The minimum value of stage count that swing modulo scheduler
9618               generates.
9619
9620           max-last-value-rtl
9621               The maximum size measured as number of RTLs that can be
9622               recorded in an expression in combiner for a pseudo register as
9623               last known value of that register.
9624
9625           max-combine-insns
9626               The maximum number of instructions the RTL combiner tries to
9627               combine.
9628
9629           integer-share-limit
9630               Small integer constants can use a shared data structure,
9631               reducing the compiler's memory usage and increasing its speed.
9632               This sets the maximum value of a shared integer constant.
9633
9634           ssp-buffer-size
9635               The minimum size of buffers (i.e. arrays) that receive stack
9636               smashing protection when -fstack-protection is used.
9637
9638           min-size-for-stack-sharing
9639               The minimum size of variables taking part in stack slot sharing
9640               when not optimizing.
9641
9642           max-jump-thread-duplication-stmts
9643               Maximum number of statements allowed in a block that needs to
9644               be duplicated when threading jumps.
9645
9646           max-fields-for-field-sensitive
9647               Maximum number of fields in a structure treated in a field
9648               sensitive manner during pointer analysis.
9649
9650           prefetch-latency
9651               Estimate on average number of instructions that are executed
9652               before prefetch finishes.  The distance prefetched ahead is
9653               proportional to this constant.  Increasing this number may also
9654               lead to less streams being prefetched (see simultaneous-
9655               prefetches).
9656
9657           simultaneous-prefetches
9658               Maximum number of prefetches that can run at the same time.
9659
9660           l1-cache-line-size
9661               The size of cache line in L1 data cache, in bytes.
9662
9663           l1-cache-size
9664               The size of L1 data cache, in kilobytes.
9665
9666           l2-cache-size
9667               The size of L2 data cache, in kilobytes.
9668
9669           prefetch-dynamic-strides
9670               Whether the loop array prefetch pass should issue software
9671               prefetch hints for strides that are non-constant.  In some
9672               cases this may be beneficial, though the fact the stride is
9673               non-constant may make it hard to predict when there is clear
9674               benefit to issuing these hints.
9675
9676               Set to 1 if the prefetch hints should be issued for non-
9677               constant strides.  Set to 0 if prefetch hints should be issued
9678               only for strides that are known to be constant and below
9679               prefetch-minimum-stride.
9680
9681           prefetch-minimum-stride
9682               Minimum constant stride, in bytes, to start using prefetch
9683               hints for.  If the stride is less than this threshold, prefetch
9684               hints will not be issued.
9685
9686               This setting is useful for processors that have hardware
9687               prefetchers, in which case there may be conflicts between the
9688               hardware prefetchers and the software prefetchers.  If the
9689               hardware prefetchers have a maximum stride they can handle, it
9690               should be used here to improve the use of software prefetchers.
9691
9692               A value of -1 means we don't have a threshold and therefore
9693               prefetch hints can be issued for any constant stride.
9694
9695               This setting is only useful for strides that are known and
9696               constant.
9697
9698           loop-interchange-max-num-stmts
9699               The maximum number of stmts in a loop to be interchanged.
9700
9701           loop-interchange-stride-ratio
9702               The minimum ratio between stride of two loops for interchange
9703               to be profitable.
9704
9705           min-insn-to-prefetch-ratio
9706               The minimum ratio between the number of instructions and the
9707               number of prefetches to enable prefetching in a loop.
9708
9709           prefetch-min-insn-to-mem-ratio
9710               The minimum ratio between the number of instructions and the
9711               number of memory references to enable prefetching in a loop.
9712
9713           use-canonical-types
9714               Whether the compiler should use the "canonical" type system.
9715               Should always be 1, which uses a more efficient internal
9716               mechanism for comparing types in C++ and Objective-C++.
9717               However, if bugs in the canonical type system are causing
9718               compilation failures, set this value to 0 to disable canonical
9719               types.
9720
9721           switch-conversion-max-branch-ratio
9722               Switch initialization conversion refuses to create arrays that
9723               are bigger than switch-conversion-max-branch-ratio times the
9724               number of branches in the switch.
9725
9726           max-partial-antic-length
9727               Maximum length of the partial antic set computed during the
9728               tree partial redundancy elimination optimization (-ftree-pre)
9729               when optimizing at -O3 and above.  For some sorts of source
9730               code the enhanced partial redundancy elimination optimization
9731               can run away, consuming all of the memory available on the host
9732               machine.  This parameter sets a limit on the length of the sets
9733               that are computed, which prevents the runaway behavior.
9734               Setting a value of 0 for this parameter allows an unlimited set
9735               length.
9736
9737           rpo-vn-max-loop-depth
9738               Maximum loop depth that is value-numbered optimistically.  When
9739               the limit hits the innermost rpo-vn-max-loop-depth loops and
9740               the outermost loop in the loop nest are value-numbered
9741               optimistically and the remaining ones not.
9742
9743           sccvn-max-alias-queries-per-access
9744               Maximum number of alias-oracle queries we perform when looking
9745               for redundancies for loads and stores.  If this limit is hit
9746               the search is aborted and the load or store is not considered
9747               redundant.  The number of queries is algorithmically limited to
9748               the number of stores on all paths from the load to the function
9749               entry.
9750
9751           ira-max-loops-num
9752               IRA uses regional register allocation by default.  If a
9753               function contains more loops than the number given by this
9754               parameter, only at most the given number of the most
9755               frequently-executed loops form regions for regional register
9756               allocation.
9757
9758           ira-max-conflict-table-size
9759               Although IRA uses a sophisticated algorithm to compress the
9760               conflict table, the table can still require excessive amounts
9761               of memory for huge functions.  If the conflict table for a
9762               function could be more than the size in MB given by this
9763               parameter, the register allocator instead uses a faster,
9764               simpler, and lower-quality algorithm that does not require
9765               building a pseudo-register conflict table.
9766
9767           ira-loop-reserved-regs
9768               IRA can be used to evaluate more accurate register pressure in
9769               loops for decisions to move loop invariants (see -O3).  The
9770               number of available registers reserved for some other purposes
9771               is given by this parameter.  Default of the parameter is the
9772               best found from numerous experiments.
9773
9774           lra-inheritance-ebb-probability-cutoff
9775               LRA tries to reuse values reloaded in registers in subsequent
9776               insns.  This optimization is called inheritance.  EBB is used
9777               as a region to do this optimization.  The parameter defines a
9778               minimal fall-through edge probability in percentage used to add
9779               BB to inheritance EBB in LRA.  The default value was chosen
9780               from numerous runs of SPEC2000 on x86-64.
9781
9782           loop-invariant-max-bbs-in-loop
9783               Loop invariant motion can be very expensive, both in
9784               compilation time and in amount of needed compile-time memory,
9785               with very large loops.  Loops with more basic blocks than this
9786               parameter won't have loop invariant motion optimization
9787               performed on them.
9788
9789           loop-max-datarefs-for-datadeps
9790               Building data dependencies is expensive for very large loops.
9791               This parameter limits the number of data references in loops
9792               that are considered for data dependence analysis.  These large
9793               loops are no handled by the optimizations using loop data
9794               dependencies.
9795
9796           max-vartrack-size
9797               Sets a maximum number of hash table slots to use during
9798               variable tracking dataflow analysis of any function.  If this
9799               limit is exceeded with variable tracking at assignments
9800               enabled, analysis for that function is retried without it,
9801               after removing all debug insns from the function.  If the limit
9802               is exceeded even without debug insns, var tracking analysis is
9803               completely disabled for the function.  Setting the parameter to
9804               zero makes it unlimited.
9805
9806           max-vartrack-expr-depth
9807               Sets a maximum number of recursion levels when attempting to
9808               map variable names or debug temporaries to value expressions.
9809               This trades compilation time for more complete debug
9810               information.  If this is set too low, value expressions that
9811               are available and could be represented in debug information may
9812               end up not being used; setting this higher may enable the
9813               compiler to find more complex debug expressions, but compile
9814               time and memory use may grow.
9815
9816           max-debug-marker-count
9817               Sets a threshold on the number of debug markers (e.g. begin
9818               stmt markers) to avoid complexity explosion at inlining or
9819               expanding to RTL.  If a function has more such gimple stmts
9820               than the set limit, such stmts will be dropped from the inlined
9821               copy of a function, and from its RTL expansion.
9822
9823           min-nondebug-insn-uid
9824               Use uids starting at this parameter for nondebug insns.  The
9825               range below the parameter is reserved exclusively for debug
9826               insns created by -fvar-tracking-assignments, but debug insns
9827               may get (non-overlapping) uids above it if the reserved range
9828               is exhausted.
9829
9830           ipa-sra-ptr-growth-factor
9831               IPA-SRA replaces a pointer to an aggregate with one or more new
9832               parameters only when their cumulative size is less or equal to
9833               ipa-sra-ptr-growth-factor times the size of the original
9834               pointer parameter.
9835
9836           sra-max-scalarization-size-Ospeed
9837           sra-max-scalarization-size-Osize
9838               The two Scalar Reduction of Aggregates passes (SRA and IPA-SRA)
9839               aim to replace scalar parts of aggregates with uses of
9840               independent scalar variables.  These parameters control the
9841               maximum size, in storage units, of aggregate which is
9842               considered for replacement when compiling for speed (sra-max-
9843               scalarization-size-Ospeed) or size (sra-max-scalarization-size-
9844               Osize) respectively.
9845
9846           tm-max-aggregate-size
9847               When making copies of thread-local variables in a transaction,
9848               this parameter specifies the size in bytes after which
9849               variables are saved with the logging functions as opposed to
9850               save/restore code sequence pairs.  This option only applies
9851               when using -fgnu-tm.
9852
9853           graphite-max-nb-scop-params
9854               To avoid exponential effects in the Graphite loop transforms,
9855               the number of parameters in a Static Control Part (SCoP) is
9856               bounded.  A value of zero can be used to lift the bound.  A
9857               variable whose value is unknown at compilation time and defined
9858               outside a SCoP is a parameter of the SCoP.
9859
9860           loop-block-tile-size
9861               Loop blocking or strip mining transforms, enabled with
9862               -floop-block or -floop-strip-mine, strip mine each loop in the
9863               loop nest by a given number of iterations.  The strip length
9864               can be changed using the loop-block-tile-size parameter.
9865
9866           ipa-cp-value-list-size
9867               IPA-CP attempts to track all possible values and types passed
9868               to a function's parameter in order to propagate them and
9869               perform devirtualization.  ipa-cp-value-list-size is the
9870               maximum number of values and types it stores per one formal
9871               parameter of a function.
9872
9873           ipa-cp-eval-threshold
9874               IPA-CP calculates its own score of cloning profitability
9875               heuristics and performs those cloning opportunities with scores
9876               that exceed ipa-cp-eval-threshold.
9877
9878           ipa-cp-recursion-penalty
9879               Percentage penalty the recursive functions will receive when
9880               they are evaluated for cloning.
9881
9882           ipa-cp-single-call-penalty
9883               Percentage penalty functions containing a single call to
9884               another function will receive when they are evaluated for
9885               cloning.
9886
9887           ipa-max-agg-items
9888               IPA-CP is also capable to propagate a number of scalar values
9889               passed in an aggregate. ipa-max-agg-items controls the maximum
9890               number of such values per one parameter.
9891
9892           ipa-cp-loop-hint-bonus
9893               When IPA-CP determines that a cloning candidate would make the
9894               number of iterations of a loop known, it adds a bonus of ipa-
9895               cp-loop-hint-bonus to the profitability score of the candidate.
9896
9897           ipa-cp-array-index-hint-bonus
9898               When IPA-CP determines that a cloning candidate would make the
9899               index of an array access known, it adds a bonus of ipa-cp-
9900               array-index-hint-bonus to the profitability score of the
9901               candidate.
9902
9903           ipa-max-aa-steps
9904               During its analysis of function bodies, IPA-CP employs alias
9905               analysis in order to track values pointed to by function
9906               parameters.  In order not spend too much time analyzing huge
9907               functions, it gives up and consider all memory clobbered after
9908               examining ipa-max-aa-steps statements modifying memory.
9909
9910           lto-partitions
9911               Specify desired number of partitions produced during WHOPR
9912               compilation.  The number of partitions should exceed the number
9913               of CPUs used for compilation.
9914
9915           lto-min-partition
9916               Size of minimal partition for WHOPR (in estimated
9917               instructions).  This prevents expenses of splitting very small
9918               programs into too many partitions.
9919
9920           lto-max-partition
9921               Size of max partition for WHOPR (in estimated instructions).
9922               to provide an upper bound for individual size of partition.
9923               Meant to be used only with balanced partitioning.
9924
9925           lto-max-streaming-parallelism
9926               Maximal number of parallel processes used for LTO streaming.
9927
9928           cxx-max-namespaces-for-diagnostic-help
9929               The maximum number of namespaces to consult for suggestions
9930               when C++ name lookup fails for an identifier.
9931
9932           sink-frequency-threshold
9933               The maximum relative execution frequency (in percents) of the
9934               target block relative to a statement's original block to allow
9935               statement sinking of a statement.  Larger numbers result in
9936               more aggressive statement sinking.  A small positive adjustment
9937               is applied for statements with memory operands as those are
9938               even more profitable so sink.
9939
9940           max-stores-to-sink
9941               The maximum number of conditional store pairs that can be sunk.
9942               Set to 0 if either vectorization (-ftree-vectorize) or if-
9943               conversion (-ftree-loop-if-convert) is disabled.
9944
9945           allow-store-data-races
9946               Allow optimizers to introduce new data races on stores.  Set to
9947               1 to allow, otherwise to 0.
9948
9949           case-values-threshold
9950               The smallest number of different values for which it is best to
9951               use a jump-table instead of a tree of conditional branches.  If
9952               the value is 0, use the default for the machine.
9953
9954           tree-reassoc-width
9955               Set the maximum number of instructions executed in parallel in
9956               reassociated tree. This parameter overrides target dependent
9957               heuristics used by default if has non zero value.
9958
9959           sched-pressure-algorithm
9960               Choose between the two available implementations of
9961               -fsched-pressure.  Algorithm 1 is the original implementation
9962               and is the more likely to prevent instructions from being
9963               reordered.  Algorithm 2 was designed to be a compromise between
9964               the relatively conservative approach taken by algorithm 1 and
9965               the rather aggressive approach taken by the default scheduler.
9966               It relies more heavily on having a regular register file and
9967               accurate register pressure classes.  See haifa-sched.c in the
9968               GCC sources for more details.
9969
9970               The default choice depends on the target.
9971
9972           max-slsr-cand-scan
9973               Set the maximum number of existing candidates that are
9974               considered when seeking a basis for a new straight-line
9975               strength reduction candidate.
9976
9977           asan-globals
9978               Enable buffer overflow detection for global objects.  This kind
9979               of protection is enabled by default if you are using
9980               -fsanitize=address option.  To disable global objects
9981               protection use --param asan-globals=0.
9982
9983           asan-stack
9984               Enable buffer overflow detection for stack objects.  This kind
9985               of protection is enabled by default when using
9986               -fsanitize=address.  To disable stack protection use --param
9987               asan-stack=0 option.
9988
9989           asan-instrument-reads
9990               Enable buffer overflow detection for memory reads.  This kind
9991               of protection is enabled by default when using
9992               -fsanitize=address.  To disable memory reads protection use
9993               --param asan-instrument-reads=0.
9994
9995           asan-instrument-writes
9996               Enable buffer overflow detection for memory writes.  This kind
9997               of protection is enabled by default when using
9998               -fsanitize=address.  To disable memory writes protection use
9999               --param asan-instrument-writes=0 option.
10000
10001           asan-memintrin
10002               Enable detection for built-in functions.  This kind of
10003               protection is enabled by default when using -fsanitize=address.
10004               To disable built-in functions protection use --param
10005               asan-memintrin=0.
10006
10007           asan-use-after-return
10008               Enable detection of use-after-return.  This kind of protection
10009               is enabled by default when using the -fsanitize=address option.
10010               To disable it use --param asan-use-after-return=0.
10011
10012               Note: By default the check is disabled at run time.  To enable
10013               it, add "detect_stack_use_after_return=1" to the environment
10014               variable ASAN_OPTIONS.
10015
10016           asan-instrumentation-with-call-threshold
10017               If number of memory accesses in function being instrumented is
10018               greater or equal to this number, use callbacks instead of
10019               inline checks.  E.g. to disable inline code use --param
10020               asan-instrumentation-with-call-threshold=0.
10021
10022           use-after-scope-direct-emission-threshold
10023               If the size of a local variable in bytes is smaller or equal to
10024               this number, directly poison (or unpoison) shadow memory
10025               instead of using run-time callbacks.
10026
10027           max-fsm-thread-path-insns
10028               Maximum number of instructions to copy when duplicating blocks
10029               on a finite state automaton jump thread path.
10030
10031           max-fsm-thread-length
10032               Maximum number of basic blocks on a finite state automaton jump
10033               thread path.
10034
10035           max-fsm-thread-paths
10036               Maximum number of new jump thread paths to create for a finite
10037               state automaton.
10038
10039           parloops-chunk-size
10040               Chunk size of omp schedule for loops parallelized by parloops.
10041
10042           parloops-schedule
10043               Schedule type of omp schedule for loops parallelized by
10044               parloops (static, dynamic, guided, auto, runtime).
10045
10046           parloops-min-per-thread
10047               The minimum number of iterations per thread of an innermost
10048               parallelized loop for which the parallelized variant is
10049               preferred over the single threaded one.  Note that for a
10050               parallelized loop nest the minimum number of iterations of the
10051               outermost loop per thread is two.
10052
10053           max-ssa-name-query-depth
10054               Maximum depth of recursion when querying properties of SSA
10055               names in things like fold routines.  One level of recursion
10056               corresponds to following a use-def chain.
10057
10058           hsa-gen-debug-stores
10059               Enable emission of special debug stores within HSA kernels
10060               which are then read and reported by libgomp plugin.  Generation
10061               of these stores is disabled by default, use --param
10062               hsa-gen-debug-stores=1 to enable it.
10063
10064           max-speculative-devirt-maydefs
10065               The maximum number of may-defs we analyze when looking for a
10066               must-def specifying the dynamic type of an object that invokes
10067               a virtual call we may be able to devirtualize speculatively.
10068
10069           max-vrp-switch-assertions
10070               The maximum number of assertions to add along the default edge
10071               of a switch statement during VRP.
10072
10073           unroll-jam-min-percent
10074               The minimum percentage of memory references that must be
10075               optimized away for the unroll-and-jam transformation to be
10076               considered profitable.
10077
10078           unroll-jam-max-unroll
10079               The maximum number of times the outer loop should be unrolled
10080               by the unroll-and-jam transformation.
10081
10082           max-rtl-if-conversion-unpredictable-cost
10083               Maximum permissible cost for the sequence that would be
10084               generated by the RTL if-conversion pass for a branch that is
10085               considered unpredictable.
10086
10087           max-variable-expansions-in-unroller
10088               If -fvariable-expansion-in-unroller is used, the maximum number
10089               of times that an individual variable will be expanded during
10090               loop unrolling.
10091
10092           tracer-min-branch-probability-feedback
10093               Stop forward growth if the probability of best edge is less
10094               than this threshold (in percent). Used when profile feedback is
10095               available.
10096
10097           partial-inlining-entry-probability
10098               Maximum probability of the entry BB of split region (in percent
10099               relative to entry BB of the function) to make partial inlining
10100               happen.
10101
10102           max-tracked-strlens
10103               Maximum number of strings for which strlen optimization pass
10104               will track string lengths.
10105
10106           gcse-after-reload-partial-fraction
10107               The threshold ratio for performing partial redundancy
10108               elimination after reload.
10109
10110           gcse-after-reload-critical-fraction
10111               The threshold ratio of critical edges execution count that
10112               permit performing redundancy elimination after reload.
10113
10114           max-loop-header-insns
10115               The maximum number of insns in loop header duplicated by the
10116               copy loop headers pass.
10117
10118           vect-epilogues-nomask
10119               Enable loop epilogue vectorization using smaller vector size.
10120
10121           slp-max-insns-in-bb
10122               Maximum number of instructions in basic block to be considered
10123               for SLP vectorization.
10124
10125           avoid-fma-max-bits
10126               Maximum number of bits for which we avoid creating FMAs.
10127
10128           sms-loop-average-count-threshold
10129               A threshold on the average loop count considered by the swing
10130               modulo scheduler.
10131
10132           sms-dfa-history
10133               The number of cycles the swing modulo scheduler considers when
10134               checking conflicts using DFA.
10135
10136           hot-bb-count-fraction
10137               Select fraction of the maximal count of repetitions of basic
10138               block in program given basic block needs to have to be
10139               considered hot (used in non-LTO mode)
10140
10141           max-inline-insns-recursive-auto
10142               The maximum number of instructions non-inline function can grow
10143               to via recursive inlining.
10144
10145           graphite-allow-codegen-errors
10146               Whether codegen errors should be ICEs when -fchecking.
10147
10148           sms-max-ii-factor
10149               A factor for tuning the upper bound that swing modulo scheduler
10150               uses for scheduling a loop.
10151
10152           lra-max-considered-reload-pseudos
10153               The max number of reload pseudos which are considered during
10154               spilling a non-reload pseudo.
10155
10156           max-pow-sqrt-depth
10157               Maximum depth of sqrt chains to use when synthesizing
10158               exponentiation by a real constant.
10159
10160           max-dse-active-local-stores
10161               Maximum number of active local stores in RTL dead store
10162               elimination.
10163
10164           asan-instrument-allocas
10165               Enable asan allocas/VLAs protection.
10166
10167           max-iterations-computation-cost
10168               Bound on the cost of an expression to compute the number of
10169               iterations.
10170
10171           max-isl-operations
10172               Maximum number of isl operations, 0 means unlimited.
10173
10174           graphite-max-arrays-per-scop
10175               Maximum number of arrays per scop.
10176
10177           max-vartrack-reverse-op-size
10178               Max. size of loc list for which reverse ops should be added.
10179
10180           unlikely-bb-count-fraction
10181               The minimum fraction of profile runs a given basic block
10182               execution count must be not to be considered unlikely.
10183
10184           tracer-dynamic-coverage-feedback
10185               The percentage of function, weighted by execution frequency,
10186               that must be covered by trace formation.  Used when profile
10187               feedback is available.
10188
10189           max-inline-recursive-depth-auto
10190               The maximum depth of recursive inlining for non-inline
10191               functions.
10192
10193           fsm-scale-path-stmts
10194               Scale factor to apply to the number of statements in a
10195               threading path when comparing to the number of (scaled) blocks.
10196
10197           fsm-maximum-phi-arguments
10198               Maximum number of arguments a PHI may have before the FSM
10199               threader will not try to thread through its block.
10200
10201           uninit-control-dep-attempts
10202               Maximum number of nested calls to search for control
10203               dependencies during uninitialized variable analysis.
10204
10205           indir-call-topn-profile
10206               Track top N target addresses in indirect-call profile.
10207
10208           max-once-peeled-insns
10209               The maximum number of insns of a peeled loop that rolls only
10210               once.
10211
10212           sra-max-scalarization-size-Osize
10213               Maximum size, in storage units, of an aggregate which should be
10214               considered for scalarization when compiling for size.
10215
10216           fsm-scale-path-blocks
10217               Scale factor to apply to the number of blocks in a threading
10218               path when comparing to the number of (scaled) statements.
10219
10220           sched-autopref-queue-depth
10221               Hardware autoprefetcher scheduler model control flag.  Number
10222               of lookahead cycles the model looks into; at ' ' only enable
10223               instruction sorting heuristic.
10224
10225           loop-versioning-max-inner-insns
10226               The maximum number of instructions that an inner loop can have
10227               before the loop versioning pass considers it too big to copy.
10228
10229           loop-versioning-max-outer-insns
10230               The maximum number of instructions that an outer loop can have
10231               before the loop versioning pass considers it too big to copy,
10232               discounting any instructions in inner loops that directly
10233               benefit from versioning.
10234
10235           ssa-name-def-chain-limit
10236               The maximum number of SSA_NAME assignments to follow in
10237               determining a property of a variable such as its value.  This
10238               limits the number of iterations or recursive calls GCC performs
10239               when optimizing certain statements or when determining their
10240               validity prior to issuing diagnostics.
10241
10242   Program Instrumentation Options
10243       GCC supports a number of command-line options that control adding run-
10244       time instrumentation to the code it normally generates.  For example,
10245       one purpose of instrumentation is collect profiling statistics for use
10246       in finding program hot spots, code coverage analysis, or profile-guided
10247       optimizations.  Another class of program instrumentation is adding run-
10248       time checking to detect programming errors like invalid pointer
10249       dereferences or out-of-bounds array accesses, as well as deliberately
10250       hostile attacks such as stack smashing or C++ vtable hijacking.  There
10251       is also a general hook which can be used to implement other forms of
10252       tracing or function-level instrumentation for debug or program analysis
10253       purposes.
10254
10255       -p
10256       -pg Generate extra code to write profile information suitable for the
10257           analysis program prof (for -p) or gprof (for -pg).  You must use
10258           this option when compiling the source files you want data about,
10259           and you must also use it when linking.
10260
10261           You can use the function attribute "no_instrument_function" to
10262           suppress profiling of individual functions when compiling with
10263           these options.
10264
10265       -fprofile-arcs
10266           Add code so that program flow arcs are instrumented.  During
10267           execution the program records how many times each branch and call
10268           is executed and how many times it is taken or returns.  On targets
10269           that support constructors with priority support, profiling properly
10270           handles constructors, destructors and C++ constructors (and
10271           destructors) of classes which are used as a type of a global
10272           variable.
10273
10274           When the compiled program exits it saves this data to a file called
10275           auxname.gcda for each source file.  The data may be used for
10276           profile-directed optimizations (-fbranch-probabilities), or for
10277           test coverage analysis (-ftest-coverage).  Each object file's
10278           auxname is generated from the name of the output file, if
10279           explicitly specified and it is not the final executable, otherwise
10280           it is the basename of the source file.  In both cases any suffix is
10281           removed (e.g. foo.gcda for input file dir/foo.c, or dir/foo.gcda
10282           for output file specified as -o dir/foo.o).
10283
10284       --coverage
10285           This option is used to compile and link code instrumented for
10286           coverage analysis.  The option is a synonym for -fprofile-arcs
10287           -ftest-coverage (when compiling) and -lgcov (when linking).  See
10288           the documentation for those options for more details.
10289
10290           *   Compile the source files with -fprofile-arcs plus optimization
10291               and code generation options.  For test coverage analysis, use
10292               the additional -ftest-coverage option.  You do not need to
10293               profile every source file in a program.
10294
10295           *   Compile the source files additionally with -fprofile-abs-path
10296               to create absolute path names in the .gcno files.  This allows
10297               gcov to find the correct sources in projects where compilations
10298               occur with different working directories.
10299
10300           *   Link your object files with -lgcov or -fprofile-arcs (the
10301               latter implies the former).
10302
10303           *   Run the program on a representative workload to generate the
10304               arc profile information.  This may be repeated any number of
10305               times.  You can run concurrent instances of your program, and
10306               provided that the file system supports locking, the data files
10307               will be correctly updated.  Unless a strict ISO C dialect
10308               option is in effect, "fork" calls are detected and correctly
10309               handled without double counting.
10310
10311           *   For profile-directed optimizations, compile the source files
10312               again with the same optimization and code generation options
10313               plus -fbranch-probabilities.
10314
10315           *   For test coverage analysis, use gcov to produce human readable
10316               information from the .gcno and .gcda files.  Refer to the gcov
10317               documentation for further information.
10318
10319           With -fprofile-arcs, for each function of your program GCC creates
10320           a program flow graph, then finds a spanning tree for the graph.
10321           Only arcs that are not on the spanning tree have to be
10322           instrumented: the compiler adds code to count the number of times
10323           that these arcs are executed.  When an arc is the only exit or only
10324           entrance to a block, the instrumentation code can be added to the
10325           block; otherwise, a new basic block must be created to hold the
10326           instrumentation code.
10327
10328       -ftest-coverage
10329           Produce a notes file that the gcov code-coverage utility can use to
10330           show program coverage.  Each source file's note file is called
10331           auxname.gcno.  Refer to the -fprofile-arcs option above for a
10332           description of auxname and instructions on how to generate test
10333           coverage data.  Coverage data matches the source files more closely
10334           if you do not optimize.
10335
10336       -fprofile-abs-path
10337           Automatically convert relative source file names to absolute path
10338           names in the .gcno files.  This allows gcov to find the correct
10339           sources in projects where compilations occur with different working
10340           directories.
10341
10342       -fprofile-dir=path
10343           Set the directory to search for the profile data files in to path.
10344           This option affects only the profile data generated by
10345           -fprofile-generate, -ftest-coverage, -fprofile-arcs and used by
10346           -fprofile-use and -fbranch-probabilities and its related options.
10347           Both absolute and relative paths can be used.  By default, GCC uses
10348           the current directory as path, thus the profile data file appears
10349           in the same directory as the object file.  In order to prevent the
10350           file name clashing, if the object file name is not an absolute
10351           path, we mangle the absolute path of the sourcename.gcda file and
10352           use it as the file name of a .gcda file.
10353
10354           When an executable is run in a massive parallel environment, it is
10355           recommended to save profile to different folders.  That can be done
10356           with variables in path that are exported during run-time:
10357
10358           %p  process ID.
10359
10360           %q{VAR}
10361               value of environment variable VAR
10362
10363       -fprofile-generate
10364       -fprofile-generate=path
10365           Enable options usually used for instrumenting application to
10366           produce profile useful for later recompilation with profile
10367           feedback based optimization.  You must use -fprofile-generate both
10368           when compiling and when linking your program.
10369
10370           The following options are enabled: -fprofile-arcs,
10371           -fprofile-values, -finline-functions, and -fipa-bit-cp.
10372
10373           If path is specified, GCC looks at the path to find the profile
10374           feedback data files. See -fprofile-dir.
10375
10376           To optimize the program based on the collected profile information,
10377           use -fprofile-use.
10378
10379       -fprofile-update=method
10380           Alter the update method for an application instrumented for profile
10381           feedback based optimization.  The method argument should be one of
10382           single, atomic or prefer-atomic.  The first one is useful for
10383           single-threaded applications, while the second one prevents profile
10384           corruption by emitting thread-safe code.
10385
10386           Warning: When an application does not properly join all threads (or
10387           creates an detached thread), a profile file can be still corrupted.
10388
10389           Using prefer-atomic would be transformed either to atomic, when
10390           supported by a target, or to single otherwise.  The GCC driver
10391           automatically selects prefer-atomic when -pthread is present in the
10392           command line.
10393
10394       -fprofile-filter-files=regex
10395           Instrument only functions from files where names match any regular
10396           expression (separated by a semi-colon).
10397
10398           For example, -fprofile-filter-files=main.c;module.*.c will
10399           instrument only main.c and all C files starting with 'module'.
10400
10401       -fprofile-exclude-files=regex
10402           Instrument only functions from files where names do not match all
10403           the regular expressions (separated by a semi-colon).
10404
10405           For example, -fprofile-exclude-files=/usr/* will prevent
10406           instrumentation of all files that are located in /usr/ folder.
10407
10408       -fsanitize=address
10409           Enable AddressSanitizer, a fast memory error detector.  Memory
10410           access instructions are instrumented to detect out-of-bounds and
10411           use-after-free bugs.  The option enables
10412           -fsanitize-address-use-after-scope.  See
10413           <https://github.com/google/sanitizers/wiki/AddressSanitizer> for
10414           more details.  The run-time behavior can be influenced using the
10415           ASAN_OPTIONS environment variable.  When set to "help=1", the
10416           available options are shown at startup of the instrumented program.
10417           See
10418           <https://github.com/google/sanitizers/wiki/AddressSanitizerFlags#run-time-flags>
10419           for a list of supported options.  The option cannot be combined
10420           with -fsanitize=thread.
10421
10422       -fsanitize=kernel-address
10423           Enable AddressSanitizer for Linux kernel.  See
10424           <https://github.com/google/kasan/wiki> for more details.
10425
10426       -fsanitize=pointer-compare
10427           Instrument comparison operation (<, <=, >, >=) with pointer
10428           operands.  The option must be combined with either
10429           -fsanitize=kernel-address or -fsanitize=address The option cannot
10430           be combined with -fsanitize=thread.  Note: By default the check is
10431           disabled at run time.  To enable it, add
10432           "detect_invalid_pointer_pairs=2" to the environment variable
10433           ASAN_OPTIONS. Using "detect_invalid_pointer_pairs=1" detects
10434           invalid operation only when both pointers are non-null.
10435
10436       -fsanitize=pointer-subtract
10437           Instrument subtraction with pointer operands.  The option must be
10438           combined with either -fsanitize=kernel-address or
10439           -fsanitize=address The option cannot be combined with
10440           -fsanitize=thread.  Note: By default the check is disabled at run
10441           time.  To enable it, add "detect_invalid_pointer_pairs=2" to the
10442           environment variable ASAN_OPTIONS. Using
10443           "detect_invalid_pointer_pairs=1" detects invalid operation only
10444           when both pointers are non-null.
10445
10446       -fsanitize=thread
10447           Enable ThreadSanitizer, a fast data race detector.  Memory access
10448           instructions are instrumented to detect data race bugs.  See
10449           <https://github.com/google/sanitizers/wiki#threadsanitizer> for
10450           more details. The run-time behavior can be influenced using the
10451           TSAN_OPTIONS environment variable; see
10452           <https://github.com/google/sanitizers/wiki/ThreadSanitizerFlags>
10453           for a list of supported options.  The option cannot be combined
10454           with -fsanitize=address, -fsanitize=leak.
10455
10456           Note that sanitized atomic builtins cannot throw exceptions when
10457           operating on invalid memory addresses with non-call exceptions
10458           (-fnon-call-exceptions).
10459
10460       -fsanitize=leak
10461           Enable LeakSanitizer, a memory leak detector.  This option only
10462           matters for linking of executables and the executable is linked
10463           against a library that overrides "malloc" and other allocator
10464           functions.  See
10465           <https://github.com/google/sanitizers/wiki/AddressSanitizerLeakSanitizer>
10466           for more details.  The run-time behavior can be influenced using
10467           the LSAN_OPTIONS environment variable.  The option cannot be
10468           combined with -fsanitize=thread.
10469
10470       -fsanitize=undefined
10471           Enable UndefinedBehaviorSanitizer, a fast undefined behavior
10472           detector.  Various computations are instrumented to detect
10473           undefined behavior at runtime.  Current suboptions are:
10474
10475           -fsanitize=shift
10476               This option enables checking that the result of a shift
10477               operation is not undefined.  Note that what exactly is
10478               considered undefined differs slightly between C and C++, as
10479               well as between ISO C90 and C99, etc.  This option has two
10480               suboptions, -fsanitize=shift-base and
10481               -fsanitize=shift-exponent.
10482
10483           -fsanitize=shift-exponent
10484               This option enables checking that the second argument of a
10485               shift operation is not negative and is smaller than the
10486               precision of the promoted first argument.
10487
10488           -fsanitize=shift-base
10489               If the second argument of a shift operation is within range,
10490               check that the result of a shift operation is not undefined.
10491               Note that what exactly is considered undefined differs slightly
10492               between C and C++, as well as between ISO C90 and C99, etc.
10493
10494           -fsanitize=integer-divide-by-zero
10495               Detect integer division by zero as well as "INT_MIN / -1"
10496               division.
10497
10498           -fsanitize=unreachable
10499               With this option, the compiler turns the
10500               "__builtin_unreachable" call into a diagnostics message call
10501               instead.  When reaching the "__builtin_unreachable" call, the
10502               behavior is undefined.
10503
10504           -fsanitize=vla-bound
10505               This option instructs the compiler to check that the size of a
10506               variable length array is positive.
10507
10508           -fsanitize=null
10509               This option enables pointer checking.  Particularly, the
10510               application built with this option turned on will issue an
10511               error message when it tries to dereference a NULL pointer, or
10512               if a reference (possibly an rvalue reference) is bound to a
10513               NULL pointer, or if a method is invoked on an object pointed by
10514               a NULL pointer.
10515
10516           -fsanitize=return
10517               This option enables return statement checking.  Programs built
10518               with this option turned on will issue an error message when the
10519               end of a non-void function is reached without actually
10520               returning a value.  This option works in C++ only.
10521
10522           -fsanitize=signed-integer-overflow
10523               This option enables signed integer overflow checking.  We check
10524               that the result of "+", "*", and both unary and binary "-" does
10525               not overflow in the signed arithmetics.  Note, integer
10526               promotion rules must be taken into account.  That is, the
10527               following is not an overflow:
10528
10529                       signed char a = SCHAR_MAX;
10530                       a++;
10531
10532           -fsanitize=bounds
10533               This option enables instrumentation of array bounds.  Various
10534               out of bounds accesses are detected.  Flexible array members,
10535               flexible array member-like arrays, and initializers of
10536               variables with static storage are not instrumented.
10537
10538           -fsanitize=bounds-strict
10539               This option enables strict instrumentation of array bounds.
10540               Most out of bounds accesses are detected, including flexible
10541               array members and flexible array member-like arrays.
10542               Initializers of variables with static storage are not
10543               instrumented.
10544
10545           -fsanitize=alignment
10546               This option enables checking of alignment of pointers when they
10547               are dereferenced, or when a reference is bound to
10548               insufficiently aligned target, or when a method or constructor
10549               is invoked on insufficiently aligned object.
10550
10551           -fsanitize=object-size
10552               This option enables instrumentation of memory references using
10553               the "__builtin_object_size" function.  Various out of bounds
10554               pointer accesses are detected.
10555
10556           -fsanitize=float-divide-by-zero
10557               Detect floating-point division by zero.  Unlike other similar
10558               options, -fsanitize=float-divide-by-zero is not enabled by
10559               -fsanitize=undefined, since floating-point division by zero can
10560               be a legitimate way of obtaining infinities and NaNs.
10561
10562           -fsanitize=float-cast-overflow
10563               This option enables floating-point type to integer conversion
10564               checking.  We check that the result of the conversion does not
10565               overflow.  Unlike other similar options,
10566               -fsanitize=float-cast-overflow is not enabled by
10567               -fsanitize=undefined.  This option does not work well with
10568               "FE_INVALID" exceptions enabled.
10569
10570           -fsanitize=nonnull-attribute
10571               This option enables instrumentation of calls, checking whether
10572               null values are not passed to arguments marked as requiring a
10573               non-null value by the "nonnull" function attribute.
10574
10575           -fsanitize=returns-nonnull-attribute
10576               This option enables instrumentation of return statements in
10577               functions marked with "returns_nonnull" function attribute, to
10578               detect returning of null values from such functions.
10579
10580           -fsanitize=bool
10581               This option enables instrumentation of loads from bool.  If a
10582               value other than 0/1 is loaded, a run-time error is issued.
10583
10584           -fsanitize=enum
10585               This option enables instrumentation of loads from an enum type.
10586               If a value outside the range of values for the enum type is
10587               loaded, a run-time error is issued.
10588
10589           -fsanitize=vptr
10590               This option enables instrumentation of C++ member function
10591               calls, member accesses and some conversions between pointers to
10592               base and derived classes, to verify the referenced object has
10593               the correct dynamic type.
10594
10595           -fsanitize=pointer-overflow
10596               This option enables instrumentation of pointer arithmetics.  If
10597               the pointer arithmetics overflows, a run-time error is issued.
10598
10599           -fsanitize=builtin
10600               This option enables instrumentation of arguments to selected
10601               builtin functions.  If an invalid value is passed to such
10602               arguments, a run-time error is issued.  E.g. passing 0 as the
10603               argument to "__builtin_ctz" or "__builtin_clz" invokes
10604               undefined behavior and is diagnosed by this option.
10605
10606           While -ftrapv causes traps for signed overflows to be emitted,
10607           -fsanitize=undefined gives a diagnostic message.  This currently
10608           works only for the C family of languages.
10609
10610       -fno-sanitize=all
10611           This option disables all previously enabled sanitizers.
10612           -fsanitize=all is not allowed, as some sanitizers cannot be used
10613           together.
10614
10615       -fasan-shadow-offset=number
10616           This option forces GCC to use custom shadow offset in
10617           AddressSanitizer checks.  It is useful for experimenting with
10618           different shadow memory layouts in Kernel AddressSanitizer.
10619
10620       -fsanitize-sections=s1,s2,...
10621           Sanitize global variables in selected user-defined sections.  si
10622           may contain wildcards.
10623
10624       -fsanitize-recover[=opts]
10625           -fsanitize-recover= controls error recovery mode for sanitizers
10626           mentioned in comma-separated list of opts.  Enabling this option
10627           for a sanitizer component causes it to attempt to continue running
10628           the program as if no error happened.  This means multiple runtime
10629           errors can be reported in a single program run, and the exit code
10630           of the program may indicate success even when errors have been
10631           reported.  The -fno-sanitize-recover= option can be used to alter
10632           this behavior: only the first detected error is reported and
10633           program then exits with a non-zero exit code.
10634
10635           Currently this feature only works for -fsanitize=undefined (and its
10636           suboptions except for -fsanitize=unreachable and
10637           -fsanitize=return), -fsanitize=float-cast-overflow,
10638           -fsanitize=float-divide-by-zero, -fsanitize=bounds-strict,
10639           -fsanitize=kernel-address and -fsanitize=address.  For these
10640           sanitizers error recovery is turned on by default, except
10641           -fsanitize=address, for which this feature is experimental.
10642           -fsanitize-recover=all and -fno-sanitize-recover=all is also
10643           accepted, the former enables recovery for all sanitizers that
10644           support it, the latter disables recovery for all sanitizers that
10645           support it.
10646
10647           Even if a recovery mode is turned on the compiler side, it needs to
10648           be also enabled on the runtime library side, otherwise the failures
10649           are still fatal.  The runtime library defaults to "halt_on_error=0"
10650           for ThreadSanitizer and UndefinedBehaviorSanitizer, while default
10651           value for AddressSanitizer is "halt_on_error=1". This can be
10652           overridden through setting the "halt_on_error" flag in the
10653           corresponding environment variable.
10654
10655           Syntax without an explicit opts parameter is deprecated.  It is
10656           equivalent to specifying an opts list of:
10657
10658                   undefined,float-cast-overflow,float-divide-by-zero,bounds-strict
10659
10660       -fsanitize-address-use-after-scope
10661           Enable sanitization of local variables to detect use-after-scope
10662           bugs.  The option sets -fstack-reuse to none.
10663
10664       -fsanitize-undefined-trap-on-error
10665           The -fsanitize-undefined-trap-on-error option instructs the
10666           compiler to report undefined behavior using "__builtin_trap" rather
10667           than a "libubsan" library routine.  The advantage of this is that
10668           the "libubsan" library is not needed and is not linked in, so this
10669           is usable even in freestanding environments.
10670
10671       -fsanitize-coverage=trace-pc
10672           Enable coverage-guided fuzzing code instrumentation.  Inserts a
10673           call to "__sanitizer_cov_trace_pc" into every basic block.
10674
10675       -fsanitize-coverage=trace-cmp
10676           Enable dataflow guided fuzzing code instrumentation.  Inserts a
10677           call to "__sanitizer_cov_trace_cmp1", "__sanitizer_cov_trace_cmp2",
10678           "__sanitizer_cov_trace_cmp4" or "__sanitizer_cov_trace_cmp8" for
10679           integral comparison with both operands variable or
10680           "__sanitizer_cov_trace_const_cmp1",
10681           "__sanitizer_cov_trace_const_cmp2",
10682           "__sanitizer_cov_trace_const_cmp4" or
10683           "__sanitizer_cov_trace_const_cmp8" for integral comparison with one
10684           operand constant, "__sanitizer_cov_trace_cmpf" or
10685           "__sanitizer_cov_trace_cmpd" for float or double comparisons and
10686           "__sanitizer_cov_trace_switch" for switch statements.
10687
10688       -fcf-protection=[full|branch|return|none]
10689           Enable code instrumentation of control-flow transfers to increase
10690           program security by checking that target addresses of control-flow
10691           transfer instructions (such as indirect function call, function
10692           return, indirect jump) are valid.  This prevents diverting the flow
10693           of control to an unexpected target.  This is intended to protect
10694           against such threats as Return-oriented Programming (ROP), and
10695           similarly call/jmp-oriented programming (COP/JOP).
10696
10697           The value "branch" tells the compiler to implement checking of
10698           validity of control-flow transfer at the point of indirect branch
10699           instructions, i.e. call/jmp instructions.  The value "return"
10700           implements checking of validity at the point of returning from a
10701           function.  The value "full" is an alias for specifying both
10702           "branch" and "return". The value "none" turns off instrumentation.
10703
10704           The macro "__CET__" is defined when -fcf-protection is used.  The
10705           first bit of "__CET__" is set to 1 for the value "branch" and the
10706           second bit of "__CET__" is set to 1 for the "return".
10707
10708           You can also use the "nocf_check" attribute to identify which
10709           functions and calls should be skipped from instrumentation.
10710
10711           Currently the x86 GNU/Linux target provides an implementation based
10712           on Intel Control-flow Enforcement Technology (CET).
10713
10714       -fstack-protector
10715           Emit extra code to check for buffer overflows, such as stack
10716           smashing attacks.  This is done by adding a guard variable to
10717           functions with vulnerable objects.  This includes functions that
10718           call "alloca", and functions with buffers larger than 8 bytes.  The
10719           guards are initialized when a function is entered and then checked
10720           when the function exits.  If a guard check fails, an error message
10721           is printed and the program exits.
10722
10723       -fstack-protector-all
10724           Like -fstack-protector except that all functions are protected.
10725
10726       -fstack-protector-strong
10727           Like -fstack-protector but includes additional functions to be
10728           protected --- those that have local array definitions, or have
10729           references to local frame addresses.
10730
10731       -fstack-protector-explicit
10732           Like -fstack-protector but only protects those functions which have
10733           the "stack_protect" attribute.
10734
10735       -fstack-check
10736           Generate code to verify that you do not go beyond the boundary of
10737           the stack.  You should specify this flag if you are running in an
10738           environment with multiple threads, but you only rarely need to
10739           specify it in a single-threaded environment since stack overflow is
10740           automatically detected on nearly all systems if there is only one
10741           stack.
10742
10743           Note that this switch does not actually cause checking to be done;
10744           the operating system or the language runtime must do that.  The
10745           switch causes generation of code to ensure that they see the stack
10746           being extended.
10747
10748           You can additionally specify a string parameter: no means no
10749           checking, generic means force the use of old-style checking,
10750           specific means use the best checking method and is equivalent to
10751           bare -fstack-check.
10752
10753           Old-style checking is a generic mechanism that requires no specific
10754           target support in the compiler but comes with the following
10755           drawbacks:
10756
10757           1.  Modified allocation strategy for large objects: they are always
10758               allocated dynamically if their size exceeds a fixed threshold.
10759               Note this may change the semantics of some code.
10760
10761           2.  Fixed limit on the size of the static frame of functions: when
10762               it is topped by a particular function, stack checking is not
10763               reliable and a warning is issued by the compiler.
10764
10765           3.  Inefficiency: because of both the modified allocation strategy
10766               and the generic implementation, code performance is hampered.
10767
10768           Note that old-style stack checking is also the fallback method for
10769           specific if no target support has been added in the compiler.
10770
10771           -fstack-check= is designed for Ada's needs to detect infinite
10772           recursion and stack overflows.  specific is an excellent choice
10773           when compiling Ada code.  It is not generally sufficient to protect
10774           against stack-clash attacks.  To protect against those you want
10775           -fstack-clash-protection.
10776
10777       -fstack-clash-protection
10778           Generate code to prevent stack clash style attacks.  When this
10779           option is enabled, the compiler will only allocate one page of
10780           stack space at a time and each page is accessed immediately after
10781           allocation.  Thus, it prevents allocations from jumping over any
10782           stack guard page provided by the operating system.
10783
10784           Most targets do not fully support stack clash protection.  However,
10785           on those targets -fstack-clash-protection will protect dynamic
10786           stack allocations.  -fstack-clash-protection may also provide
10787           limited protection for static stack allocations if the target
10788           supports -fstack-check=specific.
10789
10790       -fstack-limit-register=reg
10791       -fstack-limit-symbol=sym
10792       -fno-stack-limit
10793           Generate code to ensure that the stack does not grow beyond a
10794           certain value, either the value of a register or the address of a
10795           symbol.  If a larger stack is required, a signal is raised at run
10796           time.  For most targets, the signal is raised before the stack
10797           overruns the boundary, so it is possible to catch the signal
10798           without taking special precautions.
10799
10800           For instance, if the stack starts at absolute address 0x80000000
10801           and grows downwards, you can use the flags
10802           -fstack-limit-symbol=__stack_limit and
10803           -Wl,--defsym,__stack_limit=0x7ffe0000 to enforce a stack limit of
10804           128KB.  Note that this may only work with the GNU linker.
10805
10806           You can locally override stack limit checking by using the
10807           "no_stack_limit" function attribute.
10808
10809       -fsplit-stack
10810           Generate code to automatically split the stack before it overflows.
10811           The resulting program has a discontiguous stack which can only
10812           overflow if the program is unable to allocate any more memory.
10813           This is most useful when running threaded programs, as it is no
10814           longer necessary to calculate a good stack size to use for each
10815           thread.  This is currently only implemented for the x86 targets
10816           running GNU/Linux.
10817
10818           When code compiled with -fsplit-stack calls code compiled without
10819           -fsplit-stack, there may not be much stack space available for the
10820           latter code to run.  If compiling all code, including library code,
10821           with -fsplit-stack is not an option, then the linker can fix up
10822           these calls so that the code compiled without -fsplit-stack always
10823           has a large stack.  Support for this is implemented in the gold
10824           linker in GNU binutils release 2.21 and later.
10825
10826       -fvtable-verify=[std|preinit|none]
10827           This option is only available when compiling C++ code.  It turns on
10828           (or off, if using -fvtable-verify=none) the security feature that
10829           verifies at run time, for every virtual call, that the vtable
10830           pointer through which the call is made is valid for the type of the
10831           object, and has not been corrupted or overwritten.  If an invalid
10832           vtable pointer is detected at run time, an error is reported and
10833           execution of the program is immediately halted.
10834
10835           This option causes run-time data structures to be built at program
10836           startup, which are used for verifying the vtable pointers.  The
10837           options std and preinit control the timing of when these data
10838           structures are built.  In both cases the data structures are built
10839           before execution reaches "main".  Using -fvtable-verify=std causes
10840           the data structures to be built after shared libraries have been
10841           loaded and initialized.  -fvtable-verify=preinit causes them to be
10842           built before shared libraries have been loaded and initialized.
10843
10844           If this option appears multiple times in the command line with
10845           different values specified, none takes highest priority over both
10846           std and preinit; preinit takes priority over std.
10847
10848       -fvtv-debug
10849           When used in conjunction with -fvtable-verify=std or
10850           -fvtable-verify=preinit, causes debug versions of the runtime
10851           functions for the vtable verification feature to be called.  This
10852           flag also causes the compiler to log information about which vtable
10853           pointers it finds for each class.  This information is written to a
10854           file named vtv_set_ptr_data.log in the directory named by the
10855           environment variable VTV_LOGS_DIR if that is defined or the current
10856           working directory otherwise.
10857
10858           Note:  This feature appends data to the log file. If you want a
10859           fresh log file, be sure to delete any existing one.
10860
10861       -fvtv-counts
10862           This is a debugging flag.  When used in conjunction with
10863           -fvtable-verify=std or -fvtable-verify=preinit, this causes the
10864           compiler to keep track of the total number of virtual calls it
10865           encounters and the number of verifications it inserts.  It also
10866           counts the number of calls to certain run-time library functions
10867           that it inserts and logs this information for each compilation
10868           unit.  The compiler writes this information to a file named
10869           vtv_count_data.log in the directory named by the environment
10870           variable VTV_LOGS_DIR if that is defined or the current working
10871           directory otherwise.  It also counts the size of the vtable pointer
10872           sets for each class, and writes this information to
10873           vtv_class_set_sizes.log in the same directory.
10874
10875           Note:  This feature appends data to the log files.  To get fresh
10876           log files, be sure to delete any existing ones.
10877
10878       -finstrument-functions
10879           Generate instrumentation calls for entry and exit to functions.
10880           Just after function entry and just before function exit, the
10881           following profiling functions are called with the address of the
10882           current function and its call site.  (On some platforms,
10883           "__builtin_return_address" does not work beyond the current
10884           function, so the call site information may not be available to the
10885           profiling functions otherwise.)
10886
10887                   void __cyg_profile_func_enter (void *this_fn,
10888                                                  void *call_site);
10889                   void __cyg_profile_func_exit  (void *this_fn,
10890                                                  void *call_site);
10891
10892           The first argument is the address of the start of the current
10893           function, which may be looked up exactly in the symbol table.
10894
10895           This instrumentation is also done for functions expanded inline in
10896           other functions.  The profiling calls indicate where, conceptually,
10897           the inline function is entered and exited.  This means that
10898           addressable versions of such functions must be available.  If all
10899           your uses of a function are expanded inline, this may mean an
10900           additional expansion of code size.  If you use "extern inline" in
10901           your C code, an addressable version of such functions must be
10902           provided.  (This is normally the case anyway, but if you get lucky
10903           and the optimizer always expands the functions inline, you might
10904           have gotten away without providing static copies.)
10905
10906           A function may be given the attribute "no_instrument_function", in
10907           which case this instrumentation is not done.  This can be used, for
10908           example, for the profiling functions listed above, high-priority
10909           interrupt routines, and any functions from which the profiling
10910           functions cannot safely be called (perhaps signal handlers, if the
10911           profiling routines generate output or allocate memory).
10912
10913       -finstrument-functions-exclude-file-list=file,file,...
10914           Set the list of functions that are excluded from instrumentation
10915           (see the description of -finstrument-functions).  If the file that
10916           contains a function definition matches with one of file, then that
10917           function is not instrumented.  The match is done on substrings: if
10918           the file parameter is a substring of the file name, it is
10919           considered to be a match.
10920
10921           For example:
10922
10923                   -finstrument-functions-exclude-file-list=/bits/stl,include/sys
10924
10925           excludes any inline function defined in files whose pathnames
10926           contain /bits/stl or include/sys.
10927
10928           If, for some reason, you want to include letter , in one of sym,
10929           write ,. For example,
10930           -finstrument-functions-exclude-file-list=',,tmp' (note the single
10931           quote surrounding the option).
10932
10933       -finstrument-functions-exclude-function-list=sym,sym,...
10934           This is similar to -finstrument-functions-exclude-file-list, but
10935           this option sets the list of function names to be excluded from
10936           instrumentation.  The function name to be matched is its user-
10937           visible name, such as "vector<int> blah(const vector<int> &)", not
10938           the internal mangled name (e.g., "_Z4blahRSt6vectorIiSaIiEE").  The
10939           match is done on substrings: if the sym parameter is a substring of
10940           the function name, it is considered to be a match.  For C99 and C++
10941           extended identifiers, the function name must be given in UTF-8, not
10942           using universal character names.
10943
10944       -fpatchable-function-entry=N[,M]
10945           Generate N NOPs right at the beginning of each function, with the
10946           function entry point before the Mth NOP.  If M is omitted, it
10947           defaults to 0 so the function entry points to the address just at
10948           the first NOP.  The NOP instructions reserve extra space which can
10949           be used to patch in any desired instrumentation at run time,
10950           provided that the code segment is writable.  The amount of space is
10951           controllable indirectly via the number of NOPs; the NOP instruction
10952           used corresponds to the instruction emitted by the internal GCC
10953           back-end interface "gen_nop".  This behavior is target-specific and
10954           may also depend on the architecture variant and/or other
10955           compilation options.
10956
10957           For run-time identification, the starting addresses of these areas,
10958           which correspond to their respective function entries minus M, are
10959           additionally collected in the "__patchable_function_entries"
10960           section of the resulting binary.
10961
10962           Note that the value of "__attribute__ ((patchable_function_entry
10963           (N,M)))" takes precedence over command-line option
10964           -fpatchable-function-entry=N,M.  This can be used to increase the
10965           area size or to remove it completely on a single function.  If
10966           "N=0", no pad location is recorded.
10967
10968           The NOP instructions are inserted at---and maybe before, depending
10969           on M---the function entry address, even before the prologue.
10970
10971   Options Controlling the Preprocessor
10972       These options control the C preprocessor, which is run on each C source
10973       file before actual compilation.
10974
10975       If you use the -E option, nothing is done except preprocessing.  Some
10976       of these options make sense only together with -E because they cause
10977       the preprocessor output to be unsuitable for actual compilation.
10978
10979       In addition to the options listed here, there are a number of options
10980       to control search paths for include files documented in Directory
10981       Options.  Options to control preprocessor diagnostics are listed in
10982       Warning Options.
10983
10984       -D name
10985           Predefine name as a macro, with definition 1.
10986
10987       -D name=definition
10988           The contents of definition are tokenized and processed as if they
10989           appeared during translation phase three in a #define directive.  In
10990           particular, the definition is truncated by embedded newline
10991           characters.
10992
10993           If you are invoking the preprocessor from a shell or shell-like
10994           program you may need to use the shell's quoting syntax to protect
10995           characters such as spaces that have a meaning in the shell syntax.
10996
10997           If you wish to define a function-like macro on the command line,
10998           write its argument list with surrounding parentheses before the
10999           equals sign (if any).  Parentheses are meaningful to most shells,
11000           so you should quote the option.  With sh and csh,
11001           -D'name(args...)=definition' works.
11002
11003           -D and -U options are processed in the order they are given on the
11004           command line.  All -imacros file and -include file options are
11005           processed after all -D and -U options.
11006
11007       -U name
11008           Cancel any previous definition of name, either built in or provided
11009           with a -D option.
11010
11011       -include file
11012           Process file as if "#include "file"" appeared as the first line of
11013           the primary source file.  However, the first directory searched for
11014           file is the preprocessor's working directory instead of the
11015           directory containing the main source file.  If not found there, it
11016           is searched for in the remainder of the "#include "..."" search
11017           chain as normal.
11018
11019           If multiple -include options are given, the files are included in
11020           the order they appear on the command line.
11021
11022       -imacros file
11023           Exactly like -include, except that any output produced by scanning
11024           file is thrown away.  Macros it defines remain defined.  This
11025           allows you to acquire all the macros from a header without also
11026           processing its declarations.
11027
11028           All files specified by -imacros are processed before all files
11029           specified by -include.
11030
11031       -undef
11032           Do not predefine any system-specific or GCC-specific macros.  The
11033           standard predefined macros remain defined.
11034
11035       -pthread
11036           Define additional macros required for using the POSIX threads
11037           library.  You should use this option consistently for both
11038           compilation and linking.  This option is supported on GNU/Linux
11039           targets, most other Unix derivatives, and also on x86 Cygwin and
11040           MinGW targets.
11041
11042       -M  Instead of outputting the result of preprocessing, output a rule
11043           suitable for make describing the dependencies of the main source
11044           file.  The preprocessor outputs one make rule containing the object
11045           file name for that source file, a colon, and the names of all the
11046           included files, including those coming from -include or -imacros
11047           command-line options.
11048
11049           Unless specified explicitly (with -MT or -MQ), the object file name
11050           consists of the name of the source file with any suffix replaced
11051           with object file suffix and with any leading directory parts
11052           removed.  If there are many included files then the rule is split
11053           into several lines using \-newline.  The rule has no commands.
11054
11055           This option does not suppress the preprocessor's debug output, such
11056           as -dM.  To avoid mixing such debug output with the dependency
11057           rules you should explicitly specify the dependency output file with
11058           -MF, or use an environment variable like DEPENDENCIES_OUTPUT.
11059           Debug output is still sent to the regular output stream as normal.
11060
11061           Passing -M to the driver implies -E, and suppresses warnings with
11062           an implicit -w.
11063
11064       -MM Like -M but do not mention header files that are found in system
11065           header directories, nor header files that are included, directly or
11066           indirectly, from such a header.
11067
11068           This implies that the choice of angle brackets or double quotes in
11069           an #include directive does not in itself determine whether that
11070           header appears in -MM dependency output.
11071
11072       -MF file
11073           When used with -M or -MM, specifies a file to write the
11074           dependencies to.  If no -MF switch is given the preprocessor sends
11075           the rules to the same place it would send preprocessed output.
11076
11077           When used with the driver options -MD or -MMD, -MF overrides the
11078           default dependency output file.
11079
11080           If file is -, then the dependencies are written to stdout.
11081
11082       -MG In conjunction with an option such as -M requesting dependency
11083           generation, -MG assumes missing header files are generated files
11084           and adds them to the dependency list without raising an error.  The
11085           dependency filename is taken directly from the "#include" directive
11086           without prepending any path.  -MG also suppresses preprocessed
11087           output, as a missing header file renders this useless.
11088
11089           This feature is used in automatic updating of makefiles.
11090
11091       -MP This option instructs CPP to add a phony target for each dependency
11092           other than the main file, causing each to depend on nothing.  These
11093           dummy rules work around errors make gives if you remove header
11094           files without updating the Makefile to match.
11095
11096           This is typical output:
11097
11098                   test.o: test.c test.h
11099
11100                   test.h:
11101
11102       -MT target
11103           Change the target of the rule emitted by dependency generation.  By
11104           default CPP takes the name of the main input file, deletes any
11105           directory components and any file suffix such as .c, and appends
11106           the platform's usual object suffix.  The result is the target.
11107
11108           An -MT option sets the target to be exactly the string you specify.
11109           If you want multiple targets, you can specify them as a single
11110           argument to -MT, or use multiple -MT options.
11111
11112           For example, -MT '$(objpfx)foo.o' might give
11113
11114                   $(objpfx)foo.o: foo.c
11115
11116       -MQ target
11117           Same as -MT, but it quotes any characters which are special to
11118           Make.  -MQ '$(objpfx)foo.o' gives
11119
11120                   $$(objpfx)foo.o: foo.c
11121
11122           The default target is automatically quoted, as if it were given
11123           with -MQ.
11124
11125       -MD -MD is equivalent to -M -MF file, except that -E is not implied.
11126           The driver determines file based on whether an -o option is given.
11127           If it is, the driver uses its argument but with a suffix of .d,
11128           otherwise it takes the name of the input file, removes any
11129           directory components and suffix, and applies a .d suffix.
11130
11131           If -MD is used in conjunction with -E, any -o switch is understood
11132           to specify the dependency output file, but if used without -E, each
11133           -o is understood to specify a target object file.
11134
11135           Since -E is not implied, -MD can be used to generate a dependency
11136           output file as a side effect of the compilation process.
11137
11138       -MMD
11139           Like -MD except mention only user header files, not system header
11140           files.
11141
11142       -fpreprocessed
11143           Indicate to the preprocessor that the input file has already been
11144           preprocessed.  This suppresses things like macro expansion,
11145           trigraph conversion, escaped newline splicing, and processing of
11146           most directives.  The preprocessor still recognizes and removes
11147           comments, so that you can pass a file preprocessed with -C to the
11148           compiler without problems.  In this mode the integrated
11149           preprocessor is little more than a tokenizer for the front ends.
11150
11151           -fpreprocessed is implicit if the input file has one of the
11152           extensions .i, .ii or .mi.  These are the extensions that GCC uses
11153           for preprocessed files created by -save-temps.
11154
11155       -fdirectives-only
11156           When preprocessing, handle directives, but do not expand macros.
11157
11158           The option's behavior depends on the -E and -fpreprocessed options.
11159
11160           With -E, preprocessing is limited to the handling of directives
11161           such as "#define", "#ifdef", and "#error".  Other preprocessor
11162           operations, such as macro expansion and trigraph conversion are not
11163           performed.  In addition, the -dD option is implicitly enabled.
11164
11165           With -fpreprocessed, predefinition of command line and most builtin
11166           macros is disabled.  Macros such as "__LINE__", which are
11167           contextually dependent, are handled normally.  This enables
11168           compilation of files previously preprocessed with "-E
11169           -fdirectives-only".
11170
11171           With both -E and -fpreprocessed, the rules for -fpreprocessed take
11172           precedence.  This enables full preprocessing of files previously
11173           preprocessed with "-E -fdirectives-only".
11174
11175       -fdollars-in-identifiers
11176           Accept $ in identifiers.
11177
11178       -fextended-identifiers
11179           Accept universal character names in identifiers.  This option is
11180           enabled by default for C99 (and later C standard versions) and C++.
11181
11182       -fno-canonical-system-headers
11183           When preprocessing, do not shorten system header paths with
11184           canonicalization.
11185
11186       -ftabstop=width
11187           Set the distance between tab stops.  This helps the preprocessor
11188           report correct column numbers in warnings or errors, even if tabs
11189           appear on the line.  If the value is less than 1 or greater than
11190           100, the option is ignored.  The default is 8.
11191
11192       -ftrack-macro-expansion[=level]
11193           Track locations of tokens across macro expansions. This allows the
11194           compiler to emit diagnostic about the current macro expansion stack
11195           when a compilation error occurs in a macro expansion. Using this
11196           option makes the preprocessor and the compiler consume more memory.
11197           The level parameter can be used to choose the level of precision of
11198           token location tracking thus decreasing the memory consumption if
11199           necessary. Value 0 of level de-activates this option. Value 1
11200           tracks tokens locations in a degraded mode for the sake of minimal
11201           memory overhead. In this mode all tokens resulting from the
11202           expansion of an argument of a function-like macro have the same
11203           location. Value 2 tracks tokens locations completely. This value is
11204           the most memory hungry.  When this option is given no argument, the
11205           default parameter value is 2.
11206
11207           Note that "-ftrack-macro-expansion=2" is activated by default.
11208
11209       -fmacro-prefix-map=old=new
11210           When preprocessing files residing in directory old, expand the
11211           "__FILE__" and "__BASE_FILE__" macros as if the files resided in
11212           directory new instead.  This can be used to change an absolute path
11213           to a relative path by using . for new which can result in more
11214           reproducible builds that are location independent.  This option
11215           also affects "__builtin_FILE()" during compilation.  See also
11216           -ffile-prefix-map.
11217
11218       -fexec-charset=charset
11219           Set the execution character set, used for string and character
11220           constants.  The default is UTF-8.  charset can be any encoding
11221           supported by the system's "iconv" library routine.
11222
11223       -fwide-exec-charset=charset
11224           Set the wide execution character set, used for wide string and
11225           character constants.  The default is UTF-32 or UTF-16, whichever
11226           corresponds to the width of "wchar_t".  As with -fexec-charset,
11227           charset can be any encoding supported by the system's "iconv"
11228           library routine; however, you will have problems with encodings
11229           that do not fit exactly in "wchar_t".
11230
11231       -finput-charset=charset
11232           Set the input character set, used for translation from the
11233           character set of the input file to the source character set used by
11234           GCC.  If the locale does not specify, or GCC cannot get this
11235           information from the locale, the default is UTF-8.  This can be
11236           overridden by either the locale or this command-line option.
11237           Currently the command-line option takes precedence if there's a
11238           conflict.  charset can be any encoding supported by the system's
11239           "iconv" library routine.
11240
11241       -fpch-deps
11242           When using precompiled headers, this flag causes the dependency-
11243           output flags to also list the files from the precompiled header's
11244           dependencies.  If not specified, only the precompiled header are
11245           listed and not the files that were used to create it, because those
11246           files are not consulted when a precompiled header is used.
11247
11248       -fpch-preprocess
11249           This option allows use of a precompiled header together with -E.
11250           It inserts a special "#pragma", "#pragma GCC pch_preprocess
11251           "filename"" in the output to mark the place where the precompiled
11252           header was found, and its filename.  When -fpreprocessed is in use,
11253           GCC recognizes this "#pragma" and loads the PCH.
11254
11255           This option is off by default, because the resulting preprocessed
11256           output is only really suitable as input to GCC.  It is switched on
11257           by -save-temps.
11258
11259           You should not write this "#pragma" in your own code, but it is
11260           safe to edit the filename if the PCH file is available in a
11261           different location.  The filename may be absolute or it may be
11262           relative to GCC's current directory.
11263
11264       -fworking-directory
11265           Enable generation of linemarkers in the preprocessor output that
11266           let the compiler know the current working directory at the time of
11267           preprocessing.  When this option is enabled, the preprocessor
11268           emits, after the initial linemarker, a second linemarker with the
11269           current working directory followed by two slashes.  GCC uses this
11270           directory, when it's present in the preprocessed input, as the
11271           directory emitted as the current working directory in some
11272           debugging information formats.  This option is implicitly enabled
11273           if debugging information is enabled, but this can be inhibited with
11274           the negated form -fno-working-directory.  If the -P flag is present
11275           in the command line, this option has no effect, since no "#line"
11276           directives are emitted whatsoever.
11277
11278       -A predicate=answer
11279           Make an assertion with the predicate predicate and answer answer.
11280           This form is preferred to the older form -A predicate(answer),
11281           which is still supported, because it does not use shell special
11282           characters.
11283
11284       -A -predicate=answer
11285           Cancel an assertion with the predicate predicate and answer answer.
11286
11287       -C  Do not discard comments.  All comments are passed through to the
11288           output file, except for comments in processed directives, which are
11289           deleted along with the directive.
11290
11291           You should be prepared for side effects when using -C; it causes
11292           the preprocessor to treat comments as tokens in their own right.
11293           For example, comments appearing at the start of what would be a
11294           directive line have the effect of turning that line into an
11295           ordinary source line, since the first token on the line is no
11296           longer a #.
11297
11298       -CC Do not discard comments, including during macro expansion.  This is
11299           like -C, except that comments contained within macros are also
11300           passed through to the output file where the macro is expanded.
11301
11302           In addition to the side effects of the -C option, the -CC option
11303           causes all C++-style comments inside a macro to be converted to
11304           C-style comments.  This is to prevent later use of that macro from
11305           inadvertently commenting out the remainder of the source line.
11306
11307           The -CC option is generally used to support lint comments.
11308
11309       -P  Inhibit generation of linemarkers in the output from the
11310           preprocessor.  This might be useful when running the preprocessor
11311           on something that is not C code, and will be sent to a program
11312           which might be confused by the linemarkers.
11313
11314       -traditional
11315       -traditional-cpp
11316           Try to imitate the behavior of pre-standard C preprocessors, as
11317           opposed to ISO C preprocessors.  See the GNU CPP manual for
11318           details.
11319
11320           Note that GCC does not otherwise attempt to emulate a pre-standard
11321           C compiler, and these options are only supported with the -E
11322           switch, or when invoking CPP explicitly.
11323
11324       -trigraphs
11325           Support ISO C trigraphs.  These are three-character sequences, all
11326           starting with ??, that are defined by ISO C to stand for single
11327           characters.  For example, ??/ stands for \, so '??/n' is a
11328           character constant for a newline.
11329
11330           The nine trigraphs and their replacements are
11331
11332                   Trigraph:       ??(  ??)  ??<  ??>  ??=  ??/  ??'  ??!  ??-
11333                   Replacement:      [    ]    {    }    #    \    ^    |    ~
11334
11335           By default, GCC ignores trigraphs, but in standard-conforming modes
11336           it converts them.  See the -std and -ansi options.
11337
11338       -remap
11339           Enable special code to work around file systems which only permit
11340           very short file names, such as MS-DOS.
11341
11342       -H  Print the name of each header file used, in addition to other
11343           normal activities.  Each name is indented to show how deep in the
11344           #include stack it is.  Precompiled header files are also printed,
11345           even if they are found to be invalid; an invalid precompiled header
11346           file is printed with ...x and a valid one with ...! .
11347
11348       -dletters
11349           Says to make debugging dumps during compilation as specified by
11350           letters.  The flags documented here are those relevant to the
11351           preprocessor.  Other letters are interpreted by the compiler
11352           proper, or reserved for future versions of GCC, and so are silently
11353           ignored.  If you specify letters whose behavior conflicts, the
11354           result is undefined.
11355
11356           -dM Instead of the normal output, generate a list of #define
11357               directives for all the macros defined during the execution of
11358               the preprocessor, including predefined macros.  This gives you
11359               a way of finding out what is predefined in your version of the
11360               preprocessor.  Assuming you have no file foo.h, the command
11361
11362                       touch foo.h; cpp -dM foo.h
11363
11364               shows all the predefined macros.
11365
11366               If you use -dM without the -E option, -dM is interpreted as a
11367               synonym for -fdump-rtl-mach.
11368
11369           -dD Like -dM except in two respects: it does not include the
11370               predefined macros, and it outputs both the #define directives
11371               and the result of preprocessing.  Both kinds of output go to
11372               the standard output file.
11373
11374           -dN Like -dD, but emit only the macro names, not their expansions.
11375
11376           -dI Output #include directives in addition to the result of
11377               preprocessing.
11378
11379           -dU Like -dD except that only macros that are expanded, or whose
11380               definedness is tested in preprocessor directives, are output;
11381               the output is delayed until the use or test of the macro; and
11382               #undef directives are also output for macros tested but
11383               undefined at the time.
11384
11385       -fdebug-cpp
11386           This option is only useful for debugging GCC.  When used from CPP
11387           or with -E, it dumps debugging information about location maps.
11388           Every token in the output is preceded by the dump of the map its
11389           location belongs to.
11390
11391           When used from GCC without -E, this option has no effect.
11392
11393       -Wp,option
11394           You can use -Wp,option to bypass the compiler driver and pass
11395           option directly through to the preprocessor.  If option contains
11396           commas, it is split into multiple options at the commas.  However,
11397           many options are modified, translated or interpreted by the
11398           compiler driver before being passed to the preprocessor, and -Wp
11399           forcibly bypasses this phase.  The preprocessor's direct interface
11400           is undocumented and subject to change, so whenever possible you
11401           should avoid using -Wp and let the driver handle the options
11402           instead.
11403
11404       -Xpreprocessor option
11405           Pass option as an option to the preprocessor.  You can use this to
11406           supply system-specific preprocessor options that GCC does not
11407           recognize.
11408
11409           If you want to pass an option that takes an argument, you must use
11410           -Xpreprocessor twice, once for the option and once for the
11411           argument.
11412
11413       -no-integrated-cpp
11414           Perform preprocessing as a separate pass before compilation.  By
11415           default, GCC performs preprocessing as an integrated part of input
11416           tokenization and parsing.  If this option is provided, the
11417           appropriate language front end (cc1, cc1plus, or cc1obj for C, C++,
11418           and Objective-C, respectively) is instead invoked twice, once for
11419           preprocessing only and once for actual compilation of the
11420           preprocessed input.  This option may be useful in conjunction with
11421           the -B or -wrapper options to specify an alternate preprocessor or
11422           perform additional processing of the program source between normal
11423           preprocessing and compilation.
11424
11425   Passing Options to the Assembler
11426       You can pass options to the assembler.
11427
11428       -Wa,option
11429           Pass option as an option to the assembler.  If option contains
11430           commas, it is split into multiple options at the commas.
11431
11432       -Xassembler option
11433           Pass option as an option to the assembler.  You can use this to
11434           supply system-specific assembler options that GCC does not
11435           recognize.
11436
11437           If you want to pass an option that takes an argument, you must use
11438           -Xassembler twice, once for the option and once for the argument.
11439
11440   Options for Linking
11441       These options come into play when the compiler links object files into
11442       an executable output file.  They are meaningless if the compiler is not
11443       doing a link step.
11444
11445       object-file-name
11446           A file name that does not end in a special recognized suffix is
11447           considered to name an object file or library.  (Object files are
11448           distinguished from libraries by the linker according to the file
11449           contents.)  If linking is done, these object files are used as
11450           input to the linker.
11451
11452       -c
11453       -S
11454       -E  If any of these options is used, then the linker is not run, and
11455           object file names should not be used as arguments.
11456
11457       -flinker-output=type
11458           This option controls the code generation of the link time
11459           optimizer.  By default the linker output is determined by the
11460           linker plugin automatically. For debugging the compiler and in the
11461           case of incremental linking to non-lto object file is desired, it
11462           may be useful to control the type manually.
11463
11464           If type is exec the code generation is configured to produce static
11465           binary. In this case -fpic and -fpie are both disabled.
11466
11467           If type is dyn the code generation is configured to produce shared
11468           library. In this case -fpic or -fPIC is preserved, but not enabled
11469           automatically.  This makes it possible to build shared libraries
11470           without position independent code on architectures this is
11471           possible, i.e. on x86.
11472
11473           If type is pie the code generation is configured to produce -fpie
11474           executable. This result in similar optimizations as exec except
11475           that -fpie is not disabled if specified at compilation time.
11476
11477           If type is rel the compiler assumes that incremental linking is
11478           done.  The sections containing intermediate code for link-time
11479           optimization are merged, pre-optimized, and output to the resulting
11480           object file. In addition, if -ffat-lto-objects is specified the
11481           binary code is produced for future non-lto linking. The object file
11482           produced by incremental linking will be smaller than a static
11483           library produced from the same object files.  At link-time the
11484           result of incremental linking will also load faster to compiler
11485           than a static library assuming that majority of objects in the
11486           library are used.
11487
11488           Finally nolto-rel configure compiler to for incremental linking
11489           where code generation is forced, final binary is produced and the
11490           intermediate code for later link-time optimization is stripped.
11491           When multiple object files are linked together the resulting code
11492           will be optimized better than with link time optimizations disabled
11493           (for example, the cross-module inlining will happen), most of
11494           benefits of whole program optimizations are however lost.
11495
11496           During the incremental link (by -r) the linker plugin will default
11497           to rel. With current interfaces to GNU Binutils it is however not
11498           possible to link incrementally LTO objects and non-LTO objects into
11499           a single mixed object file.  In the case any of object files in
11500           incremental link cannot be used for link-time optimization the
11501           linker plugin will output warning and use nolto-rel. To maintain
11502           the whole program optimization it is recommended to link such
11503           objects into static library instead. Alternatively it is possible
11504           to use H.J. Lu's binutils with support for mixed objects.
11505
11506       -fuse-ld=bfd
11507           Use the bfd linker instead of the default linker.
11508
11509       -fuse-ld=gold
11510           Use the gold linker instead of the default linker.
11511
11512       -fuse-ld=lld
11513           Use the LLVM lld linker instead of the default linker.
11514
11515       -llibrary
11516       -l library
11517           Search the library named library when linking.  (The second
11518           alternative with the library as a separate argument is only for
11519           POSIX compliance and is not recommended.)
11520
11521           The -l option is passed directly to the linker by GCC.  Refer to
11522           your linker documentation for exact details.  The general
11523           description below applies to the GNU linker.
11524
11525           The linker searches a standard list of directories for the library.
11526           The directories searched include several standard system
11527           directories plus any that you specify with -L.
11528
11529           Static libraries are archives of object files, and have file names
11530           like liblibrary.a.  Some targets also support shared libraries,
11531           which typically have names like liblibrary.so.  If both static and
11532           shared libraries are found, the linker gives preference to linking
11533           with the shared library unless the -static option is used.
11534
11535           It makes a difference where in the command you write this option;
11536           the linker searches and processes libraries and object files in the
11537           order they are specified.  Thus, foo.o -lz bar.o searches library z
11538           after file foo.o but before bar.o.  If bar.o refers to functions in
11539           z, those functions may not be loaded.
11540
11541       -lobjc
11542           You need this special case of the -l option in order to link an
11543           Objective-C or Objective-C++ program.
11544
11545       -nostartfiles
11546           Do not use the standard system startup files when linking.  The
11547           standard system libraries are used normally, unless -nostdlib,
11548           -nolibc, or -nodefaultlibs is used.
11549
11550       -nodefaultlibs
11551           Do not use the standard system libraries when linking.  Only the
11552           libraries you specify are passed to the linker, and options
11553           specifying linkage of the system libraries, such as -static-libgcc
11554           or -shared-libgcc, are ignored.  The standard startup files are
11555           used normally, unless -nostartfiles is used.
11556
11557           The compiler may generate calls to "memcmp", "memset", "memcpy" and
11558           "memmove".  These entries are usually resolved by entries in libc.
11559           These entry points should be supplied through some other mechanism
11560           when this option is specified.
11561
11562       -nolibc
11563           Do not use the C library or system libraries tightly coupled with
11564           it when linking.  Still link with the startup files, libgcc or
11565           toolchain provided language support libraries such as libgnat,
11566           libgfortran or libstdc++ unless options preventing their inclusion
11567           are used as well.  This typically removes -lc from the link command
11568           line, as well as system libraries that normally go with it and
11569           become meaningless when absence of a C library is assumed, for
11570           example -lpthread or -lm in some configurations.  This is intended
11571           for bare-board targets when there is indeed no C library available.
11572
11573       -nostdlib
11574           Do not use the standard system startup files or libraries when
11575           linking.  No startup files and only the libraries you specify are
11576           passed to the linker, and options specifying linkage of the system
11577           libraries, such as -static-libgcc or -shared-libgcc, are ignored.
11578
11579           The compiler may generate calls to "memcmp", "memset", "memcpy" and
11580           "memmove".  These entries are usually resolved by entries in libc.
11581           These entry points should be supplied through some other mechanism
11582           when this option is specified.
11583
11584           One of the standard libraries bypassed by -nostdlib and
11585           -nodefaultlibs is libgcc.a, a library of internal subroutines which
11586           GCC uses to overcome shortcomings of particular machines, or
11587           special needs for some languages.
11588
11589           In most cases, you need libgcc.a even when you want to avoid other
11590           standard libraries.  In other words, when you specify -nostdlib or
11591           -nodefaultlibs you should usually specify -lgcc as well.  This
11592           ensures that you have no unresolved references to internal GCC
11593           library subroutines.  (An example of such an internal subroutine is
11594           "__main", used to ensure C++ constructors are called.)
11595
11596       -e entry
11597       --entry=entry
11598           Specify that the program entry point is entry.  The argument is
11599           interpreted by the linker; the GNU linker accepts either a symbol
11600           name or an address.
11601
11602       -pie
11603           Produce a dynamically linked position independent executable on
11604           targets that support it.  For predictable results, you must also
11605           specify the same set of options used for compilation (-fpie, -fPIE,
11606           or model suboptions) when you specify this linker option.
11607
11608       -no-pie
11609           Don't produce a dynamically linked position independent executable.
11610
11611       -static-pie
11612           Produce a static position independent executable on targets that
11613           support it.  A static position independent executable is similar to
11614           a static executable, but can be loaded at any address without a
11615           dynamic linker.  For predictable results, you must also specify the
11616           same set of options used for compilation (-fpie, -fPIE, or model
11617           suboptions) when you specify this linker option.
11618
11619       -pthread
11620           Link with the POSIX threads library.  This option is supported on
11621           GNU/Linux targets, most other Unix derivatives, and also on x86
11622           Cygwin and MinGW targets.  On some targets this option also sets
11623           flags for the preprocessor, so it should be used consistently for
11624           both compilation and linking.
11625
11626       -r  Produce a relocatable object as output.  This is also known as
11627           partial linking.
11628
11629       -rdynamic
11630           Pass the flag -export-dynamic to the ELF linker, on targets that
11631           support it. This instructs the linker to add all symbols, not only
11632           used ones, to the dynamic symbol table. This option is needed for
11633           some uses of "dlopen" or to allow obtaining backtraces from within
11634           a program.
11635
11636       -s  Remove all symbol table and relocation information from the
11637           executable.
11638
11639       -static
11640           On systems that support dynamic linking, this overrides -pie and
11641           prevents linking with the shared libraries.  On other systems, this
11642           option has no effect.
11643
11644       -shared
11645           Produce a shared object which can then be linked with other objects
11646           to form an executable.  Not all systems support this option.  For
11647           predictable results, you must also specify the same set of options
11648           used for compilation (-fpic, -fPIC, or model suboptions) when you
11649           specify this linker option.[1]
11650
11651       -shared-libgcc
11652       -static-libgcc
11653           On systems that provide libgcc as a shared library, these options
11654           force the use of either the shared or static version, respectively.
11655           If no shared version of libgcc was built when the compiler was
11656           configured, these options have no effect.
11657
11658           There are several situations in which an application should use the
11659           shared libgcc instead of the static version.  The most common of
11660           these is when the application wishes to throw and catch exceptions
11661           across different shared libraries.  In that case, each of the
11662           libraries as well as the application itself should use the shared
11663           libgcc.
11664
11665           Therefore, the G++ driver automatically adds -shared-libgcc
11666           whenever you build a shared library or a main executable, because
11667           C++ programs typically use exceptions, so this is the right thing
11668           to do.
11669
11670           If, instead, you use the GCC driver to create shared libraries, you
11671           may find that they are not always linked with the shared libgcc.
11672           If GCC finds, at its configuration time, that you have a non-GNU
11673           linker or a GNU linker that does not support option --eh-frame-hdr,
11674           it links the shared version of libgcc into shared libraries by
11675           default.  Otherwise, it takes advantage of the linker and optimizes
11676           away the linking with the shared version of libgcc, linking with
11677           the static version of libgcc by default.  This allows exceptions to
11678           propagate through such shared libraries, without incurring
11679           relocation costs at library load time.
11680
11681           However, if a library or main executable is supposed to throw or
11682           catch exceptions, you must link it using the G++ driver, or using
11683           the option -shared-libgcc, such that it is linked with the shared
11684           libgcc.
11685
11686       -static-libasan
11687           When the -fsanitize=address option is used to link a program, the
11688           GCC driver automatically links against libasan.  If libasan is
11689           available as a shared library, and the -static option is not used,
11690           then this links against the shared version of libasan.  The
11691           -static-libasan option directs the GCC driver to link libasan
11692           statically, without necessarily linking other libraries statically.
11693
11694       -static-libtsan
11695           When the -fsanitize=thread option is used to link a program, the
11696           GCC driver automatically links against libtsan.  If libtsan is
11697           available as a shared library, and the -static option is not used,
11698           then this links against the shared version of libtsan.  The
11699           -static-libtsan option directs the GCC driver to link libtsan
11700           statically, without necessarily linking other libraries statically.
11701
11702       -static-liblsan
11703           When the -fsanitize=leak option is used to link a program, the GCC
11704           driver automatically links against liblsan.  If liblsan is
11705           available as a shared library, and the -static option is not used,
11706           then this links against the shared version of liblsan.  The
11707           -static-liblsan option directs the GCC driver to link liblsan
11708           statically, without necessarily linking other libraries statically.
11709
11710       -static-libubsan
11711           When the -fsanitize=undefined option is used to link a program, the
11712           GCC driver automatically links against libubsan.  If libubsan is
11713           available as a shared library, and the -static option is not used,
11714           then this links against the shared version of libubsan.  The
11715           -static-libubsan option directs the GCC driver to link libubsan
11716           statically, without necessarily linking other libraries statically.
11717
11718       -static-libstdc++
11719           When the g++ program is used to link a C++ program, it normally
11720           automatically links against libstdc++.  If libstdc++ is available
11721           as a shared library, and the -static option is not used, then this
11722           links against the shared version of libstdc++.  That is normally
11723           fine.  However, it is sometimes useful to freeze the version of
11724           libstdc++ used by the program without going all the way to a fully
11725           static link.  The -static-libstdc++ option directs the g++ driver
11726           to link libstdc++ statically, without necessarily linking other
11727           libraries statically.
11728
11729       -symbolic
11730           Bind references to global symbols when building a shared object.
11731           Warn about any unresolved references (unless overridden by the link
11732           editor option -Xlinker -z -Xlinker defs).  Only a few systems
11733           support this option.
11734
11735       -T script
11736           Use script as the linker script.  This option is supported by most
11737           systems using the GNU linker.  On some targets, such as bare-board
11738           targets without an operating system, the -T option may be required
11739           when linking to avoid references to undefined symbols.
11740
11741       -Xlinker option
11742           Pass option as an option to the linker.  You can use this to supply
11743           system-specific linker options that GCC does not recognize.
11744
11745           If you want to pass an option that takes a separate argument, you
11746           must use -Xlinker twice, once for the option and once for the
11747           argument.  For example, to pass -assert definitions, you must write
11748           -Xlinker -assert -Xlinker definitions.  It does not work to write
11749           -Xlinker "-assert definitions", because this passes the entire
11750           string as a single argument, which is not what the linker expects.
11751
11752           When using the GNU linker, it is usually more convenient to pass
11753           arguments to linker options using the option=value syntax than as
11754           separate arguments.  For example, you can specify -Xlinker
11755           -Map=output.map rather than -Xlinker -Map -Xlinker output.map.
11756           Other linkers may not support this syntax for command-line options.
11757
11758       -Wl,option
11759           Pass option as an option to the linker.  If option contains commas,
11760           it is split into multiple options at the commas.  You can use this
11761           syntax to pass an argument to the option.  For example,
11762           -Wl,-Map,output.map passes -Map output.map to the linker.  When
11763           using the GNU linker, you can also get the same effect with
11764           -Wl,-Map=output.map.
11765
11766       -u symbol
11767           Pretend the symbol symbol is undefined, to force linking of library
11768           modules to define it.  You can use -u multiple times with different
11769           symbols to force loading of additional library modules.
11770
11771       -z keyword
11772           -z is passed directly on to the linker along with the keyword
11773           keyword. See the section in the documentation of your linker for
11774           permitted values and their meanings.
11775
11776   Options for Directory Search
11777       These options specify directories to search for header files, for
11778       libraries and for parts of the compiler:
11779
11780       -I dir
11781       -iquote dir
11782       -isystem dir
11783       -idirafter dir
11784           Add the directory dir to the list of directories to be searched for
11785           header files during preprocessing.  If dir begins with = or
11786           $SYSROOT, then the = or $SYSROOT is replaced by the sysroot prefix;
11787           see --sysroot and -isysroot.
11788
11789           Directories specified with -iquote apply only to the quote form of
11790           the directive, "#include "file"".  Directories specified with -I,
11791           -isystem, or -idirafter apply to lookup for both the
11792           "#include "file"" and "#include <file>" directives.
11793
11794           You can specify any number or combination of these options on the
11795           command line to search for header files in several directories.
11796           The lookup order is as follows:
11797
11798           1.  For the quote form of the include directive, the directory of
11799               the current file is searched first.
11800
11801           2.  For the quote form of the include directive, the directories
11802               specified by -iquote options are searched in left-to-right
11803               order, as they appear on the command line.
11804
11805           3.  Directories specified with -I options are scanned in left-to-
11806               right order.
11807
11808           4.  Directories specified with -isystem options are scanned in
11809               left-to-right order.
11810
11811           5.  Standard system directories are scanned.
11812
11813           6.  Directories specified with -idirafter options are scanned in
11814               left-to-right order.
11815
11816           You can use -I to override a system header file, substituting your
11817           own version, since these directories are searched before the
11818           standard system header file directories.  However, you should not
11819           use this option to add directories that contain vendor-supplied
11820           system header files; use -isystem for that.
11821
11822           The -isystem and -idirafter options also mark the directory as a
11823           system directory, so that it gets the same special treatment that
11824           is applied to the standard system directories.
11825
11826           If a standard system include directory, or a directory specified
11827           with -isystem, is also specified with -I, the -I option is ignored.
11828           The directory is still searched but as a system directory at its
11829           normal position in the system include chain.  This is to ensure
11830           that GCC's procedure to fix buggy system headers and the ordering
11831           for the "#include_next" directive are not inadvertently changed.
11832           If you really need to change the search order for system
11833           directories, use the -nostdinc and/or -isystem options.
11834
11835       -I- Split the include path.  This option has been deprecated.  Please
11836           use -iquote instead for -I directories before the -I- and remove
11837           the -I- option.
11838
11839           Any directories specified with -I options before -I- are searched
11840           only for headers requested with "#include "file""; they are not
11841           searched for "#include <file>".  If additional directories are
11842           specified with -I options after the -I-, those directories are
11843           searched for all #include directives.
11844
11845           In addition, -I- inhibits the use of the directory of the current
11846           file directory as the first search directory for "#include "file"".
11847           There is no way to override this effect of -I-.
11848
11849       -iprefix prefix
11850           Specify prefix as the prefix for subsequent -iwithprefix options.
11851           If the prefix represents a directory, you should include the final
11852           /.
11853
11854       -iwithprefix dir
11855       -iwithprefixbefore dir
11856           Append dir to the prefix specified previously with -iprefix, and
11857           add the resulting directory to the include search path.
11858           -iwithprefixbefore puts it in the same place -I would; -iwithprefix
11859           puts it where -idirafter would.
11860
11861       -isysroot dir
11862           This option is like the --sysroot option, but applies only to
11863           header files (except for Darwin targets, where it applies to both
11864           header files and libraries).  See the --sysroot option for more
11865           information.
11866
11867       -imultilib dir
11868           Use dir as a subdirectory of the directory containing target-
11869           specific C++ headers.
11870
11871       -nostdinc
11872           Do not search the standard system directories for header files.
11873           Only the directories explicitly specified with -I, -iquote,
11874           -isystem, and/or -idirafter options (and the directory of the
11875           current file, if appropriate) are searched.
11876
11877       -nostdinc++
11878           Do not search for header files in the C++-specific standard
11879           directories, but do still search the other standard directories.
11880           (This option is used when building the C++ library.)
11881
11882       -iplugindir=dir
11883           Set the directory to search for plugins that are passed by
11884           -fplugin=name instead of -fplugin=path/name.so.  This option is not
11885           meant to be used by the user, but only passed by the driver.
11886
11887       -Ldir
11888           Add directory dir to the list of directories to be searched for -l.
11889
11890       -Bprefix
11891           This option specifies where to find the executables, libraries,
11892           include files, and data files of the compiler itself.
11893
11894           The compiler driver program runs one or more of the subprograms
11895           cpp, cc1, as and ld.  It tries prefix as a prefix for each program
11896           it tries to run, both with and without machine/version/ for the
11897           corresponding target machine and compiler version.
11898
11899           For each subprogram to be run, the compiler driver first tries the
11900           -B prefix, if any.  If that name is not found, or if -B is not
11901           specified, the driver tries two standard prefixes, /usr/lib/gcc/
11902           and /usr/local/lib/gcc/.  If neither of those results in a file
11903           name that is found, the unmodified program name is searched for
11904           using the directories specified in your PATH environment variable.
11905
11906           The compiler checks to see if the path provided by -B refers to a
11907           directory, and if necessary it adds a directory separator character
11908           at the end of the path.
11909
11910           -B prefixes that effectively specify directory names also apply to
11911           libraries in the linker, because the compiler translates these
11912           options into -L options for the linker.  They also apply to include
11913           files in the preprocessor, because the compiler translates these
11914           options into -isystem options for the preprocessor.  In this case,
11915           the compiler appends include to the prefix.
11916
11917           The runtime support file libgcc.a can also be searched for using
11918           the -B prefix, if needed.  If it is not found there, the two
11919           standard prefixes above are tried, and that is all.  The file is
11920           left out of the link if it is not found by those means.
11921
11922           Another way to specify a prefix much like the -B prefix is to use
11923           the environment variable GCC_EXEC_PREFIX.
11924
11925           As a special kludge, if the path provided by -B is [dir/]stageN/,
11926           where N is a number in the range 0 to 9, then it is replaced by
11927           [dir/]include.  This is to help with boot-strapping the compiler.
11928
11929       -no-canonical-prefixes
11930           Do not expand any symbolic links, resolve references to /../ or
11931           /./, or make the path absolute when generating a relative prefix.
11932
11933       --sysroot=dir
11934           Use dir as the logical root directory for headers and libraries.
11935           For example, if the compiler normally searches for headers in
11936           /usr/include and libraries in /usr/lib, it instead searches
11937           dir/usr/include and dir/usr/lib.
11938
11939           If you use both this option and the -isysroot option, then the
11940           --sysroot option applies to libraries, but the -isysroot option
11941           applies to header files.
11942
11943           The GNU linker (beginning with version 2.16) has the necessary
11944           support for this option.  If your linker does not support this
11945           option, the header file aspect of --sysroot still works, but the
11946           library aspect does not.
11947
11948       --no-sysroot-suffix
11949           For some targets, a suffix is added to the root directory specified
11950           with --sysroot, depending on the other options used, so that
11951           headers may for example be found in dir/suffix/usr/include instead
11952           of dir/usr/include.  This option disables the addition of such a
11953           suffix.
11954
11955   Options for Code Generation Conventions
11956       These machine-independent options control the interface conventions
11957       used in code generation.
11958
11959       Most of them have both positive and negative forms; the negative form
11960       of -ffoo is -fno-foo.  In the table below, only one of the forms is
11961       listed---the one that is not the default.  You can figure out the other
11962       form by either removing no- or adding it.
11963
11964       -fstack-reuse=reuse-level
11965           This option controls stack space reuse for user declared local/auto
11966           variables and compiler generated temporaries.  reuse_level can be
11967           all, named_vars, or none. all enables stack reuse for all local
11968           variables and temporaries, named_vars enables the reuse only for
11969           user defined local variables with names, and none disables stack
11970           reuse completely. The default value is all. The option is needed
11971           when the program extends the lifetime of a scoped local variable or
11972           a compiler generated temporary beyond the end point defined by the
11973           language.  When a lifetime of a variable ends, and if the variable
11974           lives in memory, the optimizing compiler has the freedom to reuse
11975           its stack space with other temporaries or scoped local variables
11976           whose live range does not overlap with it. Legacy code extending
11977           local lifetime is likely to break with the stack reuse
11978           optimization.
11979
11980           For example,
11981
11982                      int *p;
11983                      {
11984                        int local1;
11985
11986                        p = &local1;
11987                        local1 = 10;
11988                        ....
11989                      }
11990                      {
11991                         int local2;
11992                         local2 = 20;
11993                         ...
11994                      }
11995
11996                      if (*p == 10)  // out of scope use of local1
11997                        {
11998
11999                        }
12000
12001           Another example:
12002
12003                      struct A
12004                      {
12005                          A(int k) : i(k), j(k) { }
12006                          int i;
12007                          int j;
12008                      };
12009
12010                      A *ap;
12011
12012                      void foo(const A& ar)
12013                      {
12014                         ap = &ar;
12015                      }
12016
12017                      void bar()
12018                      {
12019                         foo(A(10)); // temp object's lifetime ends when foo returns
12020
12021                         {
12022                           A a(20);
12023                           ....
12024                         }
12025                         ap->i+= 10;  // ap references out of scope temp whose space
12026                                      // is reused with a. What is the value of ap->i?
12027                      }
12028
12029           The lifetime of a compiler generated temporary is well defined by
12030           the C++ standard. When a lifetime of a temporary ends, and if the
12031           temporary lives in memory, the optimizing compiler has the freedom
12032           to reuse its stack space with other temporaries or scoped local
12033           variables whose live range does not overlap with it. However some
12034           of the legacy code relies on the behavior of older compilers in
12035           which temporaries' stack space is not reused, the aggressive stack
12036           reuse can lead to runtime errors. This option is used to control
12037           the temporary stack reuse optimization.
12038
12039       -ftrapv
12040           This option generates traps for signed overflow on addition,
12041           subtraction, multiplication operations.  The options -ftrapv and
12042           -fwrapv override each other, so using -ftrapv -fwrapv on the
12043           command-line results in -fwrapv being effective.  Note that only
12044           active options override, so using -ftrapv -fwrapv -fno-wrapv on the
12045           command-line results in -ftrapv being effective.
12046
12047       -fwrapv
12048           This option instructs the compiler to assume that signed arithmetic
12049           overflow of addition, subtraction and multiplication wraps around
12050           using twos-complement representation.  This flag enables some
12051           optimizations and disables others.  The options -ftrapv and -fwrapv
12052           override each other, so using -ftrapv -fwrapv on the command-line
12053           results in -fwrapv being effective.  Note that only active options
12054           override, so using -ftrapv -fwrapv -fno-wrapv on the command-line
12055           results in -ftrapv being effective.
12056
12057       -fwrapv-pointer
12058           This option instructs the compiler to assume that pointer
12059           arithmetic overflow on addition and subtraction wraps around using
12060           twos-complement representation.  This flag disables some
12061           optimizations which assume pointer overflow is invalid.
12062
12063       -fstrict-overflow
12064           This option implies -fno-wrapv -fno-wrapv-pointer and when negated
12065           implies -fwrapv -fwrapv-pointer.
12066
12067       -fexceptions
12068           Enable exception handling.  Generates extra code needed to
12069           propagate exceptions.  For some targets, this implies GCC generates
12070           frame unwind information for all functions, which can produce
12071           significant data size overhead, although it does not affect
12072           execution.  If you do not specify this option, GCC enables it by
12073           default for languages like C++ that normally require exception
12074           handling, and disables it for languages like C that do not normally
12075           require it.  However, you may need to enable this option when
12076           compiling C code that needs to interoperate properly with exception
12077           handlers written in C++.  You may also wish to disable this option
12078           if you are compiling older C++ programs that don't use exception
12079           handling.
12080
12081       -fnon-call-exceptions
12082           Generate code that allows trapping instructions to throw
12083           exceptions.  Note that this requires platform-specific runtime
12084           support that does not exist everywhere.  Moreover, it only allows
12085           trapping instructions to throw exceptions, i.e. memory references
12086           or floating-point instructions.  It does not allow exceptions to be
12087           thrown from arbitrary signal handlers such as "SIGALRM".
12088
12089       -fdelete-dead-exceptions
12090           Consider that instructions that may throw exceptions but don't
12091           otherwise contribute to the execution of the program can be
12092           optimized away.  This option is enabled by default for the Ada
12093           front end, as permitted by the Ada language specification.
12094           Optimization passes that cause dead exceptions to be removed are
12095           enabled independently at different optimization levels.
12096
12097       -funwind-tables
12098           Similar to -fexceptions, except that it just generates any needed
12099           static data, but does not affect the generated code in any other
12100           way.  You normally do not need to enable this option; instead, a
12101           language processor that needs this handling enables it on your
12102           behalf.
12103
12104       -fasynchronous-unwind-tables
12105           Generate unwind table in DWARF format, if supported by target
12106           machine.  The table is exact at each instruction boundary, so it
12107           can be used for stack unwinding from asynchronous events (such as
12108           debugger or garbage collector).
12109
12110       -fno-gnu-unique
12111           On systems with recent GNU assembler and C library, the C++
12112           compiler uses the "STB_GNU_UNIQUE" binding to make sure that
12113           definitions of template static data members and static local
12114           variables in inline functions are unique even in the presence of
12115           "RTLD_LOCAL"; this is necessary to avoid problems with a library
12116           used by two different "RTLD_LOCAL" plugins depending on a
12117           definition in one of them and therefore disagreeing with the other
12118           one about the binding of the symbol.  But this causes "dlclose" to
12119           be ignored for affected DSOs; if your program relies on
12120           reinitialization of a DSO via "dlclose" and "dlopen", you can use
12121           -fno-gnu-unique.
12122
12123       -fpcc-struct-return
12124           Return "short" "struct" and "union" values in memory like longer
12125           ones, rather than in registers.  This convention is less efficient,
12126           but it has the advantage of allowing intercallability between GCC-
12127           compiled files and files compiled with other compilers,
12128           particularly the Portable C Compiler (pcc).
12129
12130           The precise convention for returning structures in memory depends
12131           on the target configuration macros.
12132
12133           Short structures and unions are those whose size and alignment
12134           match that of some integer type.
12135
12136           Warning: code compiled with the -fpcc-struct-return switch is not
12137           binary compatible with code compiled with the -freg-struct-return
12138           switch.  Use it to conform to a non-default application binary
12139           interface.
12140
12141       -freg-struct-return
12142           Return "struct" and "union" values in registers when possible.
12143           This is more efficient for small structures than
12144           -fpcc-struct-return.
12145
12146           If you specify neither -fpcc-struct-return nor -freg-struct-return,
12147           GCC defaults to whichever convention is standard for the target.
12148           If there is no standard convention, GCC defaults to
12149           -fpcc-struct-return, except on targets where GCC is the principal
12150           compiler.  In those cases, we can choose the standard, and we chose
12151           the more efficient register return alternative.
12152
12153           Warning: code compiled with the -freg-struct-return switch is not
12154           binary compatible with code compiled with the -fpcc-struct-return
12155           switch.  Use it to conform to a non-default application binary
12156           interface.
12157
12158       -fshort-enums
12159           Allocate to an "enum" type only as many bytes as it needs for the
12160           declared range of possible values.  Specifically, the "enum" type
12161           is equivalent to the smallest integer type that has enough room.
12162
12163           Warning: the -fshort-enums switch causes GCC to generate code that
12164           is not binary compatible with code generated without that switch.
12165           Use it to conform to a non-default application binary interface.
12166
12167       -fshort-wchar
12168           Override the underlying type for "wchar_t" to be "short unsigned
12169           int" instead of the default for the target.  This option is useful
12170           for building programs to run under WINE.
12171
12172           Warning: the -fshort-wchar switch causes GCC to generate code that
12173           is not binary compatible with code generated without that switch.
12174           Use it to conform to a non-default application binary interface.
12175
12176       -fno-common
12177           In C code, this option controls the placement of global variables
12178           defined without an initializer, known as tentative definitions in
12179           the C standard.  Tentative definitions are distinct from
12180           declarations of a variable with the "extern" keyword, which do not
12181           allocate storage.
12182
12183           Unix C compilers have traditionally allocated storage for
12184           uninitialized global variables in a common block.  This allows the
12185           linker to resolve all tentative definitions of the same variable in
12186           different compilation units to the same object, or to a non-
12187           tentative definition.  This is the behavior specified by -fcommon,
12188           and is the default for GCC on most targets.  On the other hand,
12189           this behavior is not required by ISO C, and on some targets may
12190           carry a speed or code size penalty on variable references.
12191
12192           The -fno-common option specifies that the compiler should instead
12193           place uninitialized global variables in the BSS section of the
12194           object file.  This inhibits the merging of tentative definitions by
12195           the linker so you get a multiple-definition error if the same
12196           variable is defined in more than one compilation unit.  Compiling
12197           with -fno-common is useful on targets for which it provides better
12198           performance, or if you wish to verify that the program will work on
12199           other systems that always treat uninitialized variable definitions
12200           this way.
12201
12202       -fno-ident
12203           Ignore the "#ident" directive.
12204
12205       -finhibit-size-directive
12206           Don't output a ".size" assembler directive, or anything else that
12207           would cause trouble if the function is split in the middle, and the
12208           two halves are placed at locations far apart in memory.  This
12209           option is used when compiling crtstuff.c; you should not need to
12210           use it for anything else.
12211
12212       -fverbose-asm
12213           Put extra commentary information in the generated assembly code to
12214           make it more readable.  This option is generally only of use to
12215           those who actually need to read the generated assembly code
12216           (perhaps while debugging the compiler itself).
12217
12218           -fno-verbose-asm, the default, causes the extra information to be
12219           omitted and is useful when comparing two assembler files.
12220
12221           The added comments include:
12222
12223           *   information on the compiler version and command-line options,
12224
12225           *   the source code lines associated with the assembly
12226               instructions, in the form FILENAME:LINENUMBER:CONTENT OF LINE,
12227
12228           *   hints on which high-level expressions correspond to the various
12229               assembly instruction operands.
12230
12231           For example, given this C source file:
12232
12233                   int test (int n)
12234                   {
12235                     int i;
12236                     int total = 0;
12237
12238                     for (i = 0; i < n; i++)
12239                       total += i * i;
12240
12241                     return total;
12242                   }
12243
12244           compiling to (x86_64) assembly via -S and emitting the result
12245           direct to stdout via -o -
12246
12247                   gcc -S test.c -fverbose-asm -Os -o -
12248
12249           gives output similar to this:
12250
12251                           .file   "test.c"
12252                   # GNU C11 (GCC) version 7.0.0 20160809 (experimental) (x86_64-pc-linux-gnu)
12253                     [...snip...]
12254                   # options passed:
12255                     [...snip...]
12256
12257                           .text
12258                           .globl  test
12259                           .type   test, @function
12260                   test:
12261                   .LFB0:
12262                           .cfi_startproc
12263                   # test.c:4:   int total = 0;
12264                           xorl    %eax, %eax      # <retval>
12265                   # test.c:6:   for (i = 0; i < n; i++)
12266                           xorl    %edx, %edx      # i
12267                   .L2:
12268                   # test.c:6:   for (i = 0; i < n; i++)
12269                           cmpl    %edi, %edx      # n, i
12270                           jge     .L5     #,
12271                   # test.c:7:     total += i * i;
12272                           movl    %edx, %ecx      # i, tmp92
12273                           imull   %edx, %ecx      # i, tmp92
12274                   # test.c:6:   for (i = 0; i < n; i++)
12275                           incl    %edx    # i
12276                   # test.c:7:     total += i * i;
12277                           addl    %ecx, %eax      # tmp92, <retval>
12278                           jmp     .L2     #
12279                   .L5:
12280                   # test.c:10: }
12281                           ret
12282                           .cfi_endproc
12283                   .LFE0:
12284                           .size   test, .-test
12285                           .ident  "GCC: (GNU) 7.0.0 20160809 (experimental)"
12286                           .section        .note.GNU-stack,"",@progbits
12287
12288           The comments are intended for humans rather than machines and hence
12289           the precise format of the comments is subject to change.
12290
12291       -frecord-gcc-switches
12292           This switch causes the command line used to invoke the compiler to
12293           be recorded into the object file that is being created.  This
12294           switch is only implemented on some targets and the exact format of
12295           the recording is target and binary file format dependent, but it
12296           usually takes the form of a section containing ASCII text.  This
12297           switch is related to the -fverbose-asm switch, but that switch only
12298           records information in the assembler output file as comments, so it
12299           never reaches the object file.  See also -grecord-gcc-switches for
12300           another way of storing compiler options into the object file.
12301
12302       -fpic
12303           Generate position-independent code (PIC) suitable for use in a
12304           shared library, if supported for the target machine.  Such code
12305           accesses all constant addresses through a global offset table
12306           (GOT).  The dynamic loader resolves the GOT entries when the
12307           program starts (the dynamic loader is not part of GCC; it is part
12308           of the operating system).  If the GOT size for the linked
12309           executable exceeds a machine-specific maximum size, you get an
12310           error message from the linker indicating that -fpic does not work;
12311           in that case, recompile with -fPIC instead.  (These maximums are 8k
12312           on the SPARC, 28k on AArch64 and 32k on the m68k and RS/6000.  The
12313           x86 has no such limit.)
12314
12315           Position-independent code requires special support, and therefore
12316           works only on certain machines.  For the x86, GCC supports PIC for
12317           System V but not for the Sun 386i.  Code generated for the IBM
12318           RS/6000 is always position-independent.
12319
12320           When this flag is set, the macros "__pic__" and "__PIC__" are
12321           defined to 1.
12322
12323       -fPIC
12324           If supported for the target machine, emit position-independent
12325           code, suitable for dynamic linking and avoiding any limit on the
12326           size of the global offset table.  This option makes a difference on
12327           AArch64, m68k, PowerPC and SPARC.
12328
12329           Position-independent code requires special support, and therefore
12330           works only on certain machines.
12331
12332           When this flag is set, the macros "__pic__" and "__PIC__" are
12333           defined to 2.
12334
12335       -fpie
12336       -fPIE
12337           These options are similar to -fpic and -fPIC, but the generated
12338           position-independent code can be only linked into executables.
12339           Usually these options are used to compile code that will be linked
12340           using the -pie GCC option.
12341
12342           -fpie and -fPIE both define the macros "__pie__" and "__PIE__".
12343           The macros have the value 1 for -fpie and 2 for -fPIE.
12344
12345       -fno-plt
12346           Do not use the PLT for external function calls in position-
12347           independent code.  Instead, load the callee address at call sites
12348           from the GOT and branch to it.  This leads to more efficient code
12349           by eliminating PLT stubs and exposing GOT loads to optimizations.
12350           On architectures such as 32-bit x86 where PLT stubs expect the GOT
12351           pointer in a specific register, this gives more register allocation
12352           freedom to the compiler.  Lazy binding requires use of the PLT;
12353           with -fno-plt all external symbols are resolved at load time.
12354
12355           Alternatively, the function attribute "noplt" can be used to avoid
12356           calls through the PLT for specific external functions.
12357
12358           In position-dependent code, a few targets also convert calls to
12359           functions that are marked to not use the PLT to use the GOT
12360           instead.
12361
12362       -fno-jump-tables
12363           Do not use jump tables for switch statements even where it would be
12364           more efficient than other code generation strategies.  This option
12365           is of use in conjunction with -fpic or -fPIC for building code that
12366           forms part of a dynamic linker and cannot reference the address of
12367           a jump table.  On some targets, jump tables do not require a GOT
12368           and this option is not needed.
12369
12370       -ffixed-reg
12371           Treat the register named reg as a fixed register; generated code
12372           should never refer to it (except perhaps as a stack pointer, frame
12373           pointer or in some other fixed role).
12374
12375           reg must be the name of a register.  The register names accepted
12376           are machine-specific and are defined in the "REGISTER_NAMES" macro
12377           in the machine description macro file.
12378
12379           This flag does not have a negative form, because it specifies a
12380           three-way choice.
12381
12382       -fcall-used-reg
12383           Treat the register named reg as an allocable register that is
12384           clobbered by function calls.  It may be allocated for temporaries
12385           or variables that do not live across a call.  Functions compiled
12386           this way do not save and restore the register reg.
12387
12388           It is an error to use this flag with the frame pointer or stack
12389           pointer.  Use of this flag for other registers that have fixed
12390           pervasive roles in the machine's execution model produces
12391           disastrous results.
12392
12393           This flag does not have a negative form, because it specifies a
12394           three-way choice.
12395
12396       -fcall-saved-reg
12397           Treat the register named reg as an allocable register saved by
12398           functions.  It may be allocated even for temporaries or variables
12399           that live across a call.  Functions compiled this way save and
12400           restore the register reg if they use it.
12401
12402           It is an error to use this flag with the frame pointer or stack
12403           pointer.  Use of this flag for other registers that have fixed
12404           pervasive roles in the machine's execution model produces
12405           disastrous results.
12406
12407           A different sort of disaster results from the use of this flag for
12408           a register in which function values may be returned.
12409
12410           This flag does not have a negative form, because it specifies a
12411           three-way choice.
12412
12413       -fpack-struct[=n]
12414           Without a value specified, pack all structure members together
12415           without holes.  When a value is specified (which must be a small
12416           power of two), pack structure members according to this value,
12417           representing the maximum alignment (that is, objects with default
12418           alignment requirements larger than this are output potentially
12419           unaligned at the next fitting location.
12420
12421           Warning: the -fpack-struct switch causes GCC to generate code that
12422           is not binary compatible with code generated without that switch.
12423           Additionally, it makes the code suboptimal.  Use it to conform to a
12424           non-default application binary interface.
12425
12426       -fleading-underscore
12427           This option and its counterpart, -fno-leading-underscore, forcibly
12428           change the way C symbols are represented in the object file.  One
12429           use is to help link with legacy assembly code.
12430
12431           Warning: the -fleading-underscore switch causes GCC to generate
12432           code that is not binary compatible with code generated without that
12433           switch.  Use it to conform to a non-default application binary
12434           interface.  Not all targets provide complete support for this
12435           switch.
12436
12437       -ftls-model=model
12438           Alter the thread-local storage model to be used.  The model
12439           argument should be one of global-dynamic, local-dynamic, initial-
12440           exec or local-exec.  Note that the choice is subject to
12441           optimization: the compiler may use a more efficient model for
12442           symbols not visible outside of the translation unit, or if -fpic is
12443           not given on the command line.
12444
12445           The default without -fpic is initial-exec; with -fpic the default
12446           is global-dynamic.
12447
12448       -ftrampolines
12449           For targets that normally need trampolines for nested functions,
12450           always generate them instead of using descriptors.  Otherwise, for
12451           targets that do not need them, like for example HP-PA or IA-64, do
12452           nothing.
12453
12454           A trampoline is a small piece of code that is created at run time
12455           on the stack when the address of a nested function is taken, and is
12456           used to call the nested function indirectly.  Therefore, it
12457           requires the stack to be made executable in order for the program
12458           to work properly.
12459
12460           -fno-trampolines is enabled by default on a language by language
12461           basis to let the compiler avoid generating them, if it computes
12462           that this is safe, and replace them with descriptors.  Descriptors
12463           are made up of data only, but the generated code must be prepared
12464           to deal with them.  As of this writing, -fno-trampolines is enabled
12465           by default only for Ada.
12466
12467           Moreover, code compiled with -ftrampolines and code compiled with
12468           -fno-trampolines are not binary compatible if nested functions are
12469           present.  This option must therefore be used on a program-wide
12470           basis and be manipulated with extreme care.
12471
12472       -fvisibility=[default|internal|hidden|protected]
12473           Set the default ELF image symbol visibility to the specified
12474           option---all symbols are marked with this unless overridden within
12475           the code.  Using this feature can very substantially improve
12476           linking and load times of shared object libraries, produce more
12477           optimized code, provide near-perfect API export and prevent symbol
12478           clashes.  It is strongly recommended that you use this in any
12479           shared objects you distribute.
12480
12481           Despite the nomenclature, default always means public; i.e.,
12482           available to be linked against from outside the shared object.
12483           protected and internal are pretty useless in real-world usage so
12484           the only other commonly used option is hidden.  The default if
12485           -fvisibility isn't specified is default, i.e., make every symbol
12486           public.
12487
12488           A good explanation of the benefits offered by ensuring ELF symbols
12489           have the correct visibility is given by "How To Write Shared
12490           Libraries" by Ulrich Drepper (which can be found at
12491           <https://www.akkadia.org/drepper/>)---however a superior solution
12492           made possible by this option to marking things hidden when the
12493           default is public is to make the default hidden and mark things
12494           public.  This is the norm with DLLs on Windows and with
12495           -fvisibility=hidden and "__attribute__ ((visibility("default")))"
12496           instead of "__declspec(dllexport)" you get almost identical
12497           semantics with identical syntax.  This is a great boon to those
12498           working with cross-platform projects.
12499
12500           For those adding visibility support to existing code, you may find
12501           "#pragma GCC visibility" of use.  This works by you enclosing the
12502           declarations you wish to set visibility for with (for example)
12503           "#pragma GCC visibility push(hidden)" and "#pragma GCC visibility
12504           pop".  Bear in mind that symbol visibility should be viewed as part
12505           of the API interface contract and thus all new code should always
12506           specify visibility when it is not the default; i.e., declarations
12507           only for use within the local DSO should always be marked
12508           explicitly as hidden as so to avoid PLT indirection
12509           overheads---making this abundantly clear also aids readability and
12510           self-documentation of the code.  Note that due to ISO C++
12511           specification requirements, "operator new" and "operator delete"
12512           must always be of default visibility.
12513
12514           Be aware that headers from outside your project, in particular
12515           system headers and headers from any other library you use, may not
12516           be expecting to be compiled with visibility other than the default.
12517           You may need to explicitly say "#pragma GCC visibility
12518           push(default)" before including any such headers.
12519
12520           "extern" declarations are not affected by -fvisibility, so a lot of
12521           code can be recompiled with -fvisibility=hidden with no
12522           modifications.  However, this means that calls to "extern"
12523           functions with no explicit visibility use the PLT, so it is more
12524           effective to use "__attribute ((visibility))" and/or "#pragma GCC
12525           visibility" to tell the compiler which "extern" declarations should
12526           be treated as hidden.
12527
12528           Note that -fvisibility does affect C++ vague linkage entities. This
12529           means that, for instance, an exception class that is be thrown
12530           between DSOs must be explicitly marked with default visibility so
12531           that the type_info nodes are unified between the DSOs.
12532
12533           An overview of these techniques, their benefits and how to use them
12534           is at <http://gcc.gnu.org/wiki/Visibility>.
12535
12536       -fstrict-volatile-bitfields
12537           This option should be used if accesses to volatile bit-fields (or
12538           other structure fields, although the compiler usually honors those
12539           types anyway) should use a single access of the width of the
12540           field's type, aligned to a natural alignment if possible.  For
12541           example, targets with memory-mapped peripheral registers might
12542           require all such accesses to be 16 bits wide; with this flag you
12543           can declare all peripheral bit-fields as "unsigned short" (assuming
12544           short is 16 bits on these targets) to force GCC to use 16-bit
12545           accesses instead of, perhaps, a more efficient 32-bit access.
12546
12547           If this option is disabled, the compiler uses the most efficient
12548           instruction.  In the previous example, that might be a 32-bit load
12549           instruction, even though that accesses bytes that do not contain
12550           any portion of the bit-field, or memory-mapped registers unrelated
12551           to the one being updated.
12552
12553           In some cases, such as when the "packed" attribute is applied to a
12554           structure field, it may not be possible to access the field with a
12555           single read or write that is correctly aligned for the target
12556           machine.  In this case GCC falls back to generating multiple
12557           accesses rather than code that will fault or truncate the result at
12558           run time.
12559
12560           Note:  Due to restrictions of the C/C++11 memory model, write
12561           accesses are not allowed to touch non bit-field members.  It is
12562           therefore recommended to define all bits of the field's type as
12563           bit-field members.
12564
12565           The default value of this option is determined by the application
12566           binary interface for the target processor.
12567
12568       -fsync-libcalls
12569           This option controls whether any out-of-line instance of the
12570           "__sync" family of functions may be used to implement the C++11
12571           "__atomic" family of functions.
12572
12573           The default value of this option is enabled, thus the only useful
12574           form of the option is -fno-sync-libcalls.  This option is used in
12575           the implementation of the libatomic runtime library.
12576
12577   GCC Developer Options
12578       This section describes command-line options that are primarily of
12579       interest to GCC developers, including options to support compiler
12580       testing and investigation of compiler bugs and compile-time performance
12581       problems.  This includes options that produce debug dumps at various
12582       points in the compilation; that print statistics such as memory use and
12583       execution time; and that print information about GCC's configuration,
12584       such as where it searches for libraries.  You should rarely need to use
12585       any of these options for ordinary compilation and linking tasks.
12586
12587       Many developer options that cause GCC to dump output to a file take an
12588       optional =filename suffix. You can specify stdout or - to dump to
12589       standard output, and stderr for standard error.
12590
12591       If =filename is omitted, a default dump file name is constructed by
12592       concatenating the base dump file name, a pass number, phase letter, and
12593       pass name.  The base dump file name is the name of output file produced
12594       by the compiler if explicitly specified and not an executable;
12595       otherwise it is the source file name.  The pass number is determined by
12596       the order passes are registered with the compiler's pass manager.  This
12597       is generally the same as the order of execution, but passes registered
12598       by plugins, target-specific passes, or passes that are otherwise
12599       registered late are numbered higher than the pass named final, even if
12600       they are executed earlier.  The phase letter is one of i (inter-
12601       procedural analysis), l (language-specific), r (RTL), or t (tree).  The
12602       files are created in the directory of the output file.
12603
12604       -dletters
12605       -fdump-rtl-pass
12606       -fdump-rtl-pass=filename
12607           Says to make debugging dumps during compilation at times specified
12608           by letters.  This is used for debugging the RTL-based passes of the
12609           compiler.
12610
12611           Some -dletters switches have different meaning when -E is used for
12612           preprocessing.
12613
12614           Debug dumps can be enabled with a -fdump-rtl switch or some -d
12615           option letters.  Here are the possible letters for use in pass and
12616           letters, and their meanings:
12617
12618           -fdump-rtl-alignments
12619               Dump after branch alignments have been computed.
12620
12621           -fdump-rtl-asmcons
12622               Dump after fixing rtl statements that have unsatisfied in/out
12623               constraints.
12624
12625           -fdump-rtl-auto_inc_dec
12626               Dump after auto-inc-dec discovery.  This pass is only run on
12627               architectures that have auto inc or auto dec instructions.
12628
12629           -fdump-rtl-barriers
12630               Dump after cleaning up the barrier instructions.
12631
12632           -fdump-rtl-bbpart
12633               Dump after partitioning hot and cold basic blocks.
12634
12635           -fdump-rtl-bbro
12636               Dump after block reordering.
12637
12638           -fdump-rtl-btl1
12639           -fdump-rtl-btl2
12640               -fdump-rtl-btl1 and -fdump-rtl-btl2 enable dumping after the
12641               two branch target load optimization passes.
12642
12643           -fdump-rtl-bypass
12644               Dump after jump bypassing and control flow optimizations.
12645
12646           -fdump-rtl-combine
12647               Dump after the RTL instruction combination pass.
12648
12649           -fdump-rtl-compgotos
12650               Dump after duplicating the computed gotos.
12651
12652           -fdump-rtl-ce1
12653           -fdump-rtl-ce2
12654           -fdump-rtl-ce3
12655               -fdump-rtl-ce1, -fdump-rtl-ce2, and -fdump-rtl-ce3 enable
12656               dumping after the three if conversion passes.
12657
12658           -fdump-rtl-cprop_hardreg
12659               Dump after hard register copy propagation.
12660
12661           -fdump-rtl-csa
12662               Dump after combining stack adjustments.
12663
12664           -fdump-rtl-cse1
12665           -fdump-rtl-cse2
12666               -fdump-rtl-cse1 and -fdump-rtl-cse2 enable dumping after the
12667               two common subexpression elimination passes.
12668
12669           -fdump-rtl-dce
12670               Dump after the standalone dead code elimination passes.
12671
12672           -fdump-rtl-dbr
12673               Dump after delayed branch scheduling.
12674
12675           -fdump-rtl-dce1
12676           -fdump-rtl-dce2
12677               -fdump-rtl-dce1 and -fdump-rtl-dce2 enable dumping after the
12678               two dead store elimination passes.
12679
12680           -fdump-rtl-eh
12681               Dump after finalization of EH handling code.
12682
12683           -fdump-rtl-eh_ranges
12684               Dump after conversion of EH handling range regions.
12685
12686           -fdump-rtl-expand
12687               Dump after RTL generation.
12688
12689           -fdump-rtl-fwprop1
12690           -fdump-rtl-fwprop2
12691               -fdump-rtl-fwprop1 and -fdump-rtl-fwprop2 enable dumping after
12692               the two forward propagation passes.
12693
12694           -fdump-rtl-gcse1
12695           -fdump-rtl-gcse2
12696               -fdump-rtl-gcse1 and -fdump-rtl-gcse2 enable dumping after
12697               global common subexpression elimination.
12698
12699           -fdump-rtl-init-regs
12700               Dump after the initialization of the registers.
12701
12702           -fdump-rtl-initvals
12703               Dump after the computation of the initial value sets.
12704
12705           -fdump-rtl-into_cfglayout
12706               Dump after converting to cfglayout mode.
12707
12708           -fdump-rtl-ira
12709               Dump after iterated register allocation.
12710
12711           -fdump-rtl-jump
12712               Dump after the second jump optimization.
12713
12714           -fdump-rtl-loop2
12715               -fdump-rtl-loop2 enables dumping after the rtl loop
12716               optimization passes.
12717
12718           -fdump-rtl-mach
12719               Dump after performing the machine dependent reorganization
12720               pass, if that pass exists.
12721
12722           -fdump-rtl-mode_sw
12723               Dump after removing redundant mode switches.
12724
12725           -fdump-rtl-rnreg
12726               Dump after register renumbering.
12727
12728           -fdump-rtl-outof_cfglayout
12729               Dump after converting from cfglayout mode.
12730
12731           -fdump-rtl-peephole2
12732               Dump after the peephole pass.
12733
12734           -fdump-rtl-postreload
12735               Dump after post-reload optimizations.
12736
12737           -fdump-rtl-pro_and_epilogue
12738               Dump after generating the function prologues and epilogues.
12739
12740           -fdump-rtl-sched1
12741           -fdump-rtl-sched2
12742               -fdump-rtl-sched1 and -fdump-rtl-sched2 enable dumping after
12743               the basic block scheduling passes.
12744
12745           -fdump-rtl-ree
12746               Dump after sign/zero extension elimination.
12747
12748           -fdump-rtl-seqabstr
12749               Dump after common sequence discovery.
12750
12751           -fdump-rtl-shorten
12752               Dump after shortening branches.
12753
12754           -fdump-rtl-sibling
12755               Dump after sibling call optimizations.
12756
12757           -fdump-rtl-split1
12758           -fdump-rtl-split2
12759           -fdump-rtl-split3
12760           -fdump-rtl-split4
12761           -fdump-rtl-split5
12762               These options enable dumping after five rounds of instruction
12763               splitting.
12764
12765           -fdump-rtl-sms
12766               Dump after modulo scheduling.  This pass is only run on some
12767               architectures.
12768
12769           -fdump-rtl-stack
12770               Dump after conversion from GCC's "flat register file" registers
12771               to the x87's stack-like registers.  This pass is only run on
12772               x86 variants.
12773
12774           -fdump-rtl-subreg1
12775           -fdump-rtl-subreg2
12776               -fdump-rtl-subreg1 and -fdump-rtl-subreg2 enable dumping after
12777               the two subreg expansion passes.
12778
12779           -fdump-rtl-unshare
12780               Dump after all rtl has been unshared.
12781
12782           -fdump-rtl-vartrack
12783               Dump after variable tracking.
12784
12785           -fdump-rtl-vregs
12786               Dump after converting virtual registers to hard registers.
12787
12788           -fdump-rtl-web
12789               Dump after live range splitting.
12790
12791           -fdump-rtl-regclass
12792           -fdump-rtl-subregs_of_mode_init
12793           -fdump-rtl-subregs_of_mode_finish
12794           -fdump-rtl-dfinit
12795           -fdump-rtl-dfinish
12796               These dumps are defined but always produce empty files.
12797
12798           -da
12799           -fdump-rtl-all
12800               Produce all the dumps listed above.
12801
12802           -dA Annotate the assembler output with miscellaneous debugging
12803               information.
12804
12805           -dD Dump all macro definitions, at the end of preprocessing, in
12806               addition to normal output.
12807
12808           -dH Produce a core dump whenever an error occurs.
12809
12810           -dp Annotate the assembler output with a comment indicating which
12811               pattern and alternative is used.  The length and cost of each
12812               instruction are also printed.
12813
12814           -dP Dump the RTL in the assembler output as a comment before each
12815               instruction.  Also turns on -dp annotation.
12816
12817           -dx Just generate RTL for a function instead of compiling it.
12818               Usually used with -fdump-rtl-expand.
12819
12820       -fdump-debug
12821           Dump debugging information generated during the debug generation
12822           phase.
12823
12824       -fdump-earlydebug
12825           Dump debugging information generated during the early debug
12826           generation phase.
12827
12828       -fdump-noaddr
12829           When doing debugging dumps, suppress address output.  This makes it
12830           more feasible to use diff on debugging dumps for compiler
12831           invocations with different compiler binaries and/or different text
12832           / bss / data / heap / stack / dso start locations.
12833
12834       -freport-bug
12835           Collect and dump debug information into a temporary file if an
12836           internal compiler error (ICE) occurs.
12837
12838       -fdump-unnumbered
12839           When doing debugging dumps, suppress instruction numbers and
12840           address output.  This makes it more feasible to use diff on
12841           debugging dumps for compiler invocations with different options, in
12842           particular with and without -g.
12843
12844       -fdump-unnumbered-links
12845           When doing debugging dumps (see -d option above), suppress
12846           instruction numbers for the links to the previous and next
12847           instructions in a sequence.
12848
12849       -fdump-ipa-switch
12850       -fdump-ipa-switch-options
12851           Control the dumping at various stages of inter-procedural analysis
12852           language tree to a file.  The file name is generated by appending a
12853           switch specific suffix to the source file name, and the file is
12854           created in the same directory as the output file.  The following
12855           dumps are possible:
12856
12857           all Enables all inter-procedural analysis dumps.
12858
12859           cgraph
12860               Dumps information about call-graph optimization, unused
12861               function removal, and inlining decisions.
12862
12863           inline
12864               Dump after function inlining.
12865
12866           Additionally, the options -optimized, -missed, -note, and -all can
12867           be provided, with the same meaning as for -fopt-info, defaulting to
12868           -optimized.
12869
12870           For example, -fdump-ipa-inline-optimized-missed will emit
12871           information on callsites that were inlined, along with callsites
12872           that were not inlined.
12873
12874           By default, the dump will contain messages about successful
12875           optimizations (equivalent to -optimized) together with low-level
12876           details about the analysis.
12877
12878       -fdump-lang-all
12879       -fdump-lang-switch
12880       -fdump-lang-switch-options
12881       -fdump-lang-switch-options=filename
12882           Control the dumping of language-specific information.  The options
12883           and filename portions behave as described in the -fdump-tree
12884           option.  The following switch values are accepted:
12885
12886           all Enable all language-specific dumps.
12887
12888           class
12889               Dump class hierarchy information.  Virtual table information is
12890               emitted unless 'slim' is specified.  This option is applicable
12891               to C++ only.
12892
12893           raw Dump the raw internal tree data.  This option is applicable to
12894               C++ only.
12895
12896       -fdump-passes
12897           Print on stderr the list of optimization passes that are turned on
12898           and off by the current command-line options.
12899
12900       -fdump-statistics-option
12901           Enable and control dumping of pass statistics in a separate file.
12902           The file name is generated by appending a suffix ending in
12903           .statistics to the source file name, and the file is created in the
12904           same directory as the output file.  If the -option form is used,
12905           -stats causes counters to be summed over the whole compilation unit
12906           while -details dumps every event as the passes generate them.  The
12907           default with no option is to sum counters for each function
12908           compiled.
12909
12910       -fdump-tree-all
12911       -fdump-tree-switch
12912       -fdump-tree-switch-options
12913       -fdump-tree-switch-options=filename
12914           Control the dumping at various stages of processing the
12915           intermediate language tree to a file.  If the -options form is
12916           used, options is a list of - separated options which control the
12917           details of the dump.  Not all options are applicable to all dumps;
12918           those that are not meaningful are ignored.  The following options
12919           are available
12920
12921           address
12922               Print the address of each node.  Usually this is not meaningful
12923               as it changes according to the environment and source file.
12924               Its primary use is for tying up a dump file with a debug
12925               environment.
12926
12927           asmname
12928               If "DECL_ASSEMBLER_NAME" has been set for a given decl, use
12929               that in the dump instead of "DECL_NAME".  Its primary use is
12930               ease of use working backward from mangled names in the assembly
12931               file.
12932
12933           slim
12934               When dumping front-end intermediate representations, inhibit
12935               dumping of members of a scope or body of a function merely
12936               because that scope has been reached.  Only dump such items when
12937               they are directly reachable by some other path.
12938
12939               When dumping pretty-printed trees, this option inhibits dumping
12940               the bodies of control structures.
12941
12942               When dumping RTL, print the RTL in slim (condensed) form
12943               instead of the default LISP-like representation.
12944
12945           raw Print a raw representation of the tree.  By default, trees are
12946               pretty-printed into a C-like representation.
12947
12948           details
12949               Enable more detailed dumps (not honored by every dump option).
12950               Also include information from the optimization passes.
12951
12952           stats
12953               Enable dumping various statistics about the pass (not honored
12954               by every dump option).
12955
12956           blocks
12957               Enable showing basic block boundaries (disabled in raw dumps).
12958
12959           graph
12960               For each of the other indicated dump files (-fdump-rtl-pass),
12961               dump a representation of the control flow graph suitable for
12962               viewing with GraphViz to file.passid.pass.dot.  Each function
12963               in the file is pretty-printed as a subgraph, so that GraphViz
12964               can render them all in a single plot.
12965
12966               This option currently only works for RTL dumps, and the RTL is
12967               always dumped in slim form.
12968
12969           vops
12970               Enable showing virtual operands for every statement.
12971
12972           lineno
12973               Enable showing line numbers for statements.
12974
12975           uid Enable showing the unique ID ("DECL_UID") for each variable.
12976
12977           verbose
12978               Enable showing the tree dump for each statement.
12979
12980           eh  Enable showing the EH region number holding each statement.
12981
12982           scev
12983               Enable showing scalar evolution analysis details.
12984
12985           optimized
12986               Enable showing optimization information (only available in
12987               certain passes).
12988
12989           missed
12990               Enable showing missed optimization information (only available
12991               in certain passes).
12992
12993           note
12994               Enable other detailed optimization information (only available
12995               in certain passes).
12996
12997           all Turn on all options, except raw, slim, verbose and lineno.
12998
12999           optall
13000               Turn on all optimization options, i.e., optimized, missed, and
13001               note.
13002
13003           To determine what tree dumps are available or find the dump for a
13004           pass of interest follow the steps below.
13005
13006           1.  Invoke GCC with -fdump-passes and in the stderr output look for
13007               a code that corresponds to the pass you are interested in.  For
13008               example, the codes "tree-evrp", "tree-vrp1", and "tree-vrp2"
13009               correspond to the three Value Range Propagation passes.  The
13010               number at the end distinguishes distinct invocations of the
13011               same pass.
13012
13013           2.  To enable the creation of the dump file, append the pass code
13014               to the -fdump- option prefix and invoke GCC with it.  For
13015               example, to enable the dump from the Early Value Range
13016               Propagation pass, invoke GCC with the -fdump-tree-evrp option.
13017               Optionally, you may specify the name of the dump file.  If you
13018               don't specify one, GCC creates as described below.
13019
13020           3.  Find the pass dump in a file whose name is composed of three
13021               components separated by a period: the name of the source file
13022               GCC was invoked to compile, a numeric suffix indicating the
13023               pass number followed by the letter t for tree passes (and the
13024               letter r for RTL passes), and finally the pass code.  For
13025               example, the Early VRP pass dump might be in a file named
13026               myfile.c.038t.evrp in the current working directory.  Note that
13027               the numeric codes are not stable and may change from one
13028               version of GCC to another.
13029
13030       -fopt-info
13031       -fopt-info-options
13032       -fopt-info-options=filename
13033           Controls optimization dumps from various optimization passes. If
13034           the -options form is used, options is a list of - separated option
13035           keywords to select the dump details and optimizations.
13036
13037           The options can be divided into three groups:
13038
13039           1.  options describing what kinds of messages should be emitted,
13040
13041           2.  options describing the verbosity of the dump, and
13042
13043           3.  options describing which optimizations should be included.
13044
13045           The options from each group can be freely mixed as they are non-
13046           overlapping. However, in case of any conflicts, the later options
13047           override the earlier options on the command line.
13048
13049           The following options control which kinds of messages should be
13050           emitted:
13051
13052           optimized
13053               Print information when an optimization is successfully applied.
13054               It is up to a pass to decide which information is relevant. For
13055               example, the vectorizer passes print the source location of
13056               loops which are successfully vectorized.
13057
13058           missed
13059               Print information about missed optimizations. Individual passes
13060               control which information to include in the output.
13061
13062           note
13063               Print verbose information about optimizations, such as certain
13064               transformations, more detailed messages about decisions etc.
13065
13066           all Print detailed optimization information. This includes
13067               optimized, missed, and note.
13068
13069           The following option controls the dump verbosity:
13070
13071           internals
13072               By default, only "high-level" messages are emitted. This option
13073               enables additional, more detailed, messages, which are likely
13074               to only be of interest to GCC developers.
13075
13076           One or more of the following option keywords can be used to
13077           describe a group of optimizations:
13078
13079           ipa Enable dumps from all interprocedural optimizations.
13080
13081           loop
13082               Enable dumps from all loop optimizations.
13083
13084           inline
13085               Enable dumps from all inlining optimizations.
13086
13087           omp Enable dumps from all OMP (Offloading and Multi Processing)
13088               optimizations.
13089
13090           vec Enable dumps from all vectorization optimizations.
13091
13092           optall
13093               Enable dumps from all optimizations. This is a superset of the
13094               optimization groups listed above.
13095
13096           If options is omitted, it defaults to optimized-optall, which means
13097           to dump messages about successful optimizations from all the
13098           passes, omitting messages that are treated as "internals".
13099
13100           If the filename is provided, then the dumps from all the applicable
13101           optimizations are concatenated into the filename.  Otherwise the
13102           dump is output onto stderr. Though multiple -fopt-info options are
13103           accepted, only one of them can include a filename. If other
13104           filenames are provided then all but the first such option are
13105           ignored.
13106
13107           Note that the output filename is overwritten in case of multiple
13108           translation units. If a combined output from multiple translation
13109           units is desired, stderr should be used instead.
13110
13111           In the following example, the optimization info is output to
13112           stderr:
13113
13114                   gcc -O3 -fopt-info
13115
13116           This example:
13117
13118                   gcc -O3 -fopt-info-missed=missed.all
13119
13120           outputs missed optimization report from all the passes into
13121           missed.all, and this one:
13122
13123                   gcc -O2 -ftree-vectorize -fopt-info-vec-missed
13124
13125           prints information about missed optimization opportunities from
13126           vectorization passes on stderr.  Note that -fopt-info-vec-missed is
13127           equivalent to -fopt-info-missed-vec.  The order of the optimization
13128           group names and message types listed after -fopt-info does not
13129           matter.
13130
13131           As another example,
13132
13133                   gcc -O3 -fopt-info-inline-optimized-missed=inline.txt
13134
13135           outputs information about missed optimizations as well as optimized
13136           locations from all the inlining passes into inline.txt.
13137
13138           Finally, consider:
13139
13140                   gcc -fopt-info-vec-missed=vec.miss -fopt-info-loop-optimized=loop.opt
13141
13142           Here the two output filenames vec.miss and loop.opt are in conflict
13143           since only one output file is allowed. In this case, only the first
13144           option takes effect and the subsequent options are ignored. Thus
13145           only vec.miss is produced which contains dumps from the vectorizer
13146           about missed opportunities.
13147
13148       -fsave-optimization-record
13149           Write a SRCFILE.opt-record.json.gz file detailing what
13150           optimizations were performed, for those optimizations that support
13151           -fopt-info.
13152
13153           This option is experimental and the format of the data within the
13154           compressed JSON file is subject to change.
13155
13156           It is roughly equivalent to a machine-readable version of
13157           -fopt-info-all, as a collection of messages with source file, line
13158           number and column number, with the following additional data for
13159           each message:
13160
13161           *   the execution count of the code being optimized, along with
13162               metadata about whether this was from actual profile data, or
13163               just an estimate, allowing consumers to prioritize messages by
13164               code hotness,
13165
13166           *   the function name of the code being optimized, where
13167               applicable,
13168
13169           *   the "inlining chain" for the code being optimized, so that when
13170               a function is inlined into several different places (which
13171               might themselves be inlined), the reader can distinguish
13172               between the copies,
13173
13174           *   objects identifying those parts of the message that refer to
13175               expressions, statements or symbol-table nodes, which of these
13176               categories they are, and, when available, their source code
13177               location,
13178
13179           *   the GCC pass that emitted the message, and
13180
13181           *   the location in GCC's own code from which the message was
13182               emitted
13183
13184           Additionally, some messages are logically nested within other
13185           messages, reflecting implementation details of the optimization
13186           passes.
13187
13188       -fsched-verbose=n
13189           On targets that use instruction scheduling, this option controls
13190           the amount of debugging output the scheduler prints to the dump
13191           files.
13192
13193           For n greater than zero, -fsched-verbose outputs the same
13194           information as -fdump-rtl-sched1 and -fdump-rtl-sched2.  For n
13195           greater than one, it also output basic block probabilities,
13196           detailed ready list information and unit/insn info.  For n greater
13197           than two, it includes RTL at abort point, control-flow and regions
13198           info.  And for n over four, -fsched-verbose also includes
13199           dependence info.
13200
13201       -fenable-kind-pass
13202       -fdisable-kind-pass=range-list
13203           This is a set of options that are used to explicitly disable/enable
13204           optimization passes.  These options are intended for use for
13205           debugging GCC.  Compiler users should use regular options for
13206           enabling/disabling passes instead.
13207
13208           -fdisable-ipa-pass
13209               Disable IPA pass pass. pass is the pass name.  If the same pass
13210               is statically invoked in the compiler multiple times, the pass
13211               name should be appended with a sequential number starting from
13212               1.
13213
13214           -fdisable-rtl-pass
13215           -fdisable-rtl-pass=range-list
13216               Disable RTL pass pass.  pass is the pass name.  If the same
13217               pass is statically invoked in the compiler multiple times, the
13218               pass name should be appended with a sequential number starting
13219               from 1.  range-list is a comma-separated list of function
13220               ranges or assembler names.  Each range is a number pair
13221               separated by a colon.  The range is inclusive in both ends.  If
13222               the range is trivial, the number pair can be simplified as a
13223               single number.  If the function's call graph node's uid falls
13224               within one of the specified ranges, the pass is disabled for
13225               that function.  The uid is shown in the function header of a
13226               dump file, and the pass names can be dumped by using option
13227               -fdump-passes.
13228
13229           -fdisable-tree-pass
13230           -fdisable-tree-pass=range-list
13231               Disable tree pass pass.  See -fdisable-rtl for the description
13232               of option arguments.
13233
13234           -fenable-ipa-pass
13235               Enable IPA pass pass.  pass is the pass name.  If the same pass
13236               is statically invoked in the compiler multiple times, the pass
13237               name should be appended with a sequential number starting from
13238               1.
13239
13240           -fenable-rtl-pass
13241           -fenable-rtl-pass=range-list
13242               Enable RTL pass pass.  See -fdisable-rtl for option argument
13243               description and examples.
13244
13245           -fenable-tree-pass
13246           -fenable-tree-pass=range-list
13247               Enable tree pass pass.  See -fdisable-rtl for the description
13248               of option arguments.
13249
13250           Here are some examples showing uses of these options.
13251
13252                   # disable ccp1 for all functions
13253                      -fdisable-tree-ccp1
13254                   # disable complete unroll for function whose cgraph node uid is 1
13255                      -fenable-tree-cunroll=1
13256                   # disable gcse2 for functions at the following ranges [1,1],
13257                   # [300,400], and [400,1000]
13258                   # disable gcse2 for functions foo and foo2
13259                      -fdisable-rtl-gcse2=foo,foo2
13260                   # disable early inlining
13261                      -fdisable-tree-einline
13262                   # disable ipa inlining
13263                      -fdisable-ipa-inline
13264                   # enable tree full unroll
13265                      -fenable-tree-unroll
13266
13267       -fchecking
13268       -fchecking=n
13269           Enable internal consistency checking.  The default depends on the
13270           compiler configuration.  -fchecking=2 enables further internal
13271           consistency checking that might affect code generation.
13272
13273       -frandom-seed=string
13274           This option provides a seed that GCC uses in place of random
13275           numbers in generating certain symbol names that have to be
13276           different in every compiled file.  It is also used to place unique
13277           stamps in coverage data files and the object files that produce
13278           them.  You can use the -frandom-seed option to produce reproducibly
13279           identical object files.
13280
13281           The string can either be a number (decimal, octal or hex) or an
13282           arbitrary string (in which case it's converted to a number by
13283           computing CRC32).
13284
13285           The string should be different for every file you compile.
13286
13287       -save-temps
13288       -save-temps=cwd
13289           Store the usual "temporary" intermediate files permanently; place
13290           them in the current directory and name them based on the source
13291           file.  Thus, compiling foo.c with -c -save-temps produces files
13292           foo.i and foo.s, as well as foo.o.  This creates a preprocessed
13293           foo.i output file even though the compiler now normally uses an
13294           integrated preprocessor.
13295
13296           When used in combination with the -x command-line option,
13297           -save-temps is sensible enough to avoid over writing an input
13298           source file with the same extension as an intermediate file.  The
13299           corresponding intermediate file may be obtained by renaming the
13300           source file before using -save-temps.
13301
13302           If you invoke GCC in parallel, compiling several different source
13303           files that share a common base name in different subdirectories or
13304           the same source file compiled for multiple output destinations, it
13305           is likely that the different parallel compilers will interfere with
13306           each other, and overwrite the temporary files.  For instance:
13307
13308                   gcc -save-temps -o outdir1/foo.o indir1/foo.c&
13309                   gcc -save-temps -o outdir2/foo.o indir2/foo.c&
13310
13311           may result in foo.i and foo.o being written to simultaneously by
13312           both compilers.
13313
13314       -save-temps=obj
13315           Store the usual "temporary" intermediate files permanently.  If the
13316           -o option is used, the temporary files are based on the object
13317           file.  If the -o option is not used, the -save-temps=obj switch
13318           behaves like -save-temps.
13319
13320           For example:
13321
13322                   gcc -save-temps=obj -c foo.c
13323                   gcc -save-temps=obj -c bar.c -o dir/xbar.o
13324                   gcc -save-temps=obj foobar.c -o dir2/yfoobar
13325
13326           creates foo.i, foo.s, dir/xbar.i, dir/xbar.s, dir2/yfoobar.i,
13327           dir2/yfoobar.s, and dir2/yfoobar.o.
13328
13329       -time[=file]
13330           Report the CPU time taken by each subprocess in the compilation
13331           sequence.  For C source files, this is the compiler proper and
13332           assembler (plus the linker if linking is done).
13333
13334           Without the specification of an output file, the output looks like
13335           this:
13336
13337                   # cc1 0.12 0.01
13338                   # as 0.00 0.01
13339
13340           The first number on each line is the "user time", that is time
13341           spent executing the program itself.  The second number is "system
13342           time", time spent executing operating system routines on behalf of
13343           the program.  Both numbers are in seconds.
13344
13345           With the specification of an output file, the output is appended to
13346           the named file, and it looks like this:
13347
13348                   0.12 0.01 cc1 <options>
13349                   0.00 0.01 as <options>
13350
13351           The "user time" and the "system time" are moved before the program
13352           name, and the options passed to the program are displayed, so that
13353           one can later tell what file was being compiled, and with which
13354           options.
13355
13356       -fdump-final-insns[=file]
13357           Dump the final internal representation (RTL) to file.  If the
13358           optional argument is omitted (or if file is "."), the name of the
13359           dump file is determined by appending ".gkd" to the compilation
13360           output file name.
13361
13362       -fcompare-debug[=opts]
13363           If no error occurs during compilation, run the compiler a second
13364           time, adding opts and -fcompare-debug-second to the arguments
13365           passed to the second compilation.  Dump the final internal
13366           representation in both compilations, and print an error if they
13367           differ.
13368
13369           If the equal sign is omitted, the default -gtoggle is used.
13370
13371           The environment variable GCC_COMPARE_DEBUG, if defined, non-empty
13372           and nonzero, implicitly enables -fcompare-debug.  If
13373           GCC_COMPARE_DEBUG is defined to a string starting with a dash, then
13374           it is used for opts, otherwise the default -gtoggle is used.
13375
13376           -fcompare-debug=, with the equal sign but without opts, is
13377           equivalent to -fno-compare-debug, which disables the dumping of the
13378           final representation and the second compilation, preventing even
13379           GCC_COMPARE_DEBUG from taking effect.
13380
13381           To verify full coverage during -fcompare-debug testing, set
13382           GCC_COMPARE_DEBUG to say -fcompare-debug-not-overridden, which GCC
13383           rejects as an invalid option in any actual compilation (rather than
13384           preprocessing, assembly or linking).  To get just a warning,
13385           setting GCC_COMPARE_DEBUG to -w%n-fcompare-debug not overridden
13386           will do.
13387
13388       -fcompare-debug-second
13389           This option is implicitly passed to the compiler for the second
13390           compilation requested by -fcompare-debug, along with options to
13391           silence warnings, and omitting other options that would cause the
13392           compiler to produce output to files or to standard output as a side
13393           effect.  Dump files and preserved temporary files are renamed so as
13394           to contain the ".gk" additional extension during the second
13395           compilation, to avoid overwriting those generated by the first.
13396
13397           When this option is passed to the compiler driver, it causes the
13398           first compilation to be skipped, which makes it useful for little
13399           other than debugging the compiler proper.
13400
13401       -gtoggle
13402           Turn off generation of debug info, if leaving out this option
13403           generates it, or turn it on at level 2 otherwise.  The position of
13404           this argument in the command line does not matter; it takes effect
13405           after all other options are processed, and it does so only once, no
13406           matter how many times it is given.  This is mainly intended to be
13407           used with -fcompare-debug.
13408
13409       -fvar-tracking-assignments-toggle
13410           Toggle -fvar-tracking-assignments, in the same way that -gtoggle
13411           toggles -g.
13412
13413       -Q  Makes the compiler print out each function name as it is compiled,
13414           and print some statistics about each pass when it finishes.
13415
13416       -ftime-report
13417           Makes the compiler print some statistics about the time consumed by
13418           each pass when it finishes.
13419
13420       -ftime-report-details
13421           Record the time consumed by infrastructure parts separately for
13422           each pass.
13423
13424       -fira-verbose=n
13425           Control the verbosity of the dump file for the integrated register
13426           allocator.  The default value is 5.  If the value n is greater or
13427           equal to 10, the dump output is sent to stderr using the same
13428           format as n minus 10.
13429
13430       -flto-report
13431           Prints a report with internal details on the workings of the link-
13432           time optimizer.  The contents of this report vary from version to
13433           version.  It is meant to be useful to GCC developers when
13434           processing object files in LTO mode (via -flto).
13435
13436           Disabled by default.
13437
13438       -flto-report-wpa
13439           Like -flto-report, but only print for the WPA phase of Link Time
13440           Optimization.
13441
13442       -fmem-report
13443           Makes the compiler print some statistics about permanent memory
13444           allocation when it finishes.
13445
13446       -fmem-report-wpa
13447           Makes the compiler print some statistics about permanent memory
13448           allocation for the WPA phase only.
13449
13450       -fpre-ipa-mem-report
13451       -fpost-ipa-mem-report
13452           Makes the compiler print some statistics about permanent memory
13453           allocation before or after interprocedural optimization.
13454
13455       -fprofile-report
13456           Makes the compiler print some statistics about consistency of the
13457           (estimated) profile and effect of individual passes.
13458
13459       -fstack-usage
13460           Makes the compiler output stack usage information for the program,
13461           on a per-function basis.  The filename for the dump is made by
13462           appending .su to the auxname.  auxname is generated from the name
13463           of the output file, if explicitly specified and it is not an
13464           executable, otherwise it is the basename of the source file.  An
13465           entry is made up of three fields:
13466
13467           *   The name of the function.
13468
13469           *   A number of bytes.
13470
13471           *   One or more qualifiers: "static", "dynamic", "bounded".
13472
13473           The qualifier "static" means that the function manipulates the
13474           stack statically: a fixed number of bytes are allocated for the
13475           frame on function entry and released on function exit; no stack
13476           adjustments are otherwise made in the function.  The second field
13477           is this fixed number of bytes.
13478
13479           The qualifier "dynamic" means that the function manipulates the
13480           stack dynamically: in addition to the static allocation described
13481           above, stack adjustments are made in the body of the function, for
13482           example to push/pop arguments around function calls.  If the
13483           qualifier "bounded" is also present, the amount of these
13484           adjustments is bounded at compile time and the second field is an
13485           upper bound of the total amount of stack used by the function.  If
13486           it is not present, the amount of these adjustments is not bounded
13487           at compile time and the second field only represents the bounded
13488           part.
13489
13490       -fstats
13491           Emit statistics about front-end processing at the end of the
13492           compilation.  This option is supported only by the C++ front end,
13493           and the information is generally only useful to the G++ development
13494           team.
13495
13496       -fdbg-cnt-list
13497           Print the name and the counter upper bound for all debug counters.
13498
13499       -fdbg-cnt=counter-value-list
13500           Set the internal debug counter lower and upper bound.  counter-
13501           value-list is a comma-separated list of
13502           name:lower_bound:upper_bound tuples which sets the lower and the
13503           upper bound of each debug counter name.  The lower_bound is
13504           optional and is zero initialized if not set.  All debug counters
13505           have the initial upper bound of "UINT_MAX"; thus "dbg_cnt" returns
13506           true always unless the upper bound is set by this option.  For
13507           example, with -fdbg-cnt=dce:2:4,tail_call:10, "dbg_cnt(dce)"
13508           returns true only for third and fourth invocation.  For
13509           "dbg_cnt(tail_call)" true is returned for first 10 invocations.
13510
13511       -print-file-name=library
13512           Print the full absolute name of the library file library that would
13513           be used when linking---and don't do anything else.  With this
13514           option, GCC does not compile or link anything; it just prints the
13515           file name.
13516
13517       -print-multi-directory
13518           Print the directory name corresponding to the multilib selected by
13519           any other switches present in the command line.  This directory is
13520           supposed to exist in GCC_EXEC_PREFIX.
13521
13522       -print-multi-lib
13523           Print the mapping from multilib directory names to compiler
13524           switches that enable them.  The directory name is separated from
13525           the switches by ;, and each switch starts with an @ instead of the
13526           -, without spaces between multiple switches.  This is supposed to
13527           ease shell processing.
13528
13529       -print-multi-os-directory
13530           Print the path to OS libraries for the selected multilib, relative
13531           to some lib subdirectory.  If OS libraries are present in the lib
13532           subdirectory and no multilibs are used, this is usually just ., if
13533           OS libraries are present in libsuffix sibling directories this
13534           prints e.g. ../lib64, ../lib or ../lib32, or if OS libraries are
13535           present in lib/subdir subdirectories it prints e.g. amd64, sparcv9
13536           or ev6.
13537
13538       -print-multiarch
13539           Print the path to OS libraries for the selected multiarch, relative
13540           to some lib subdirectory.
13541
13542       -print-prog-name=program
13543           Like -print-file-name, but searches for a program such as cpp.
13544
13545       -print-libgcc-file-name
13546           Same as -print-file-name=libgcc.a.
13547
13548           This is useful when you use -nostdlib or -nodefaultlibs but you do
13549           want to link with libgcc.a.  You can do:
13550
13551                   gcc -nostdlib <files>... `gcc -print-libgcc-file-name`
13552
13553       -print-search-dirs
13554           Print the name of the configured installation directory and a list
13555           of program and library directories gcc searches---and don't do
13556           anything else.
13557
13558           This is useful when gcc prints the error message installation
13559           problem, cannot exec cpp0: No such file or directory.  To resolve
13560           this you either need to put cpp0 and the other compiler components
13561           where gcc expects to find them, or you can set the environment
13562           variable GCC_EXEC_PREFIX to the directory where you installed them.
13563           Don't forget the trailing /.
13564
13565       -print-sysroot
13566           Print the target sysroot directory that is used during compilation.
13567           This is the target sysroot specified either at configure time or
13568           using the --sysroot option, possibly with an extra suffix that
13569           depends on compilation options.  If no target sysroot is specified,
13570           the option prints nothing.
13571
13572       -print-sysroot-headers-suffix
13573           Print the suffix added to the target sysroot when searching for
13574           headers, or give an error if the compiler is not configured with
13575           such a suffix---and don't do anything else.
13576
13577       -dumpmachine
13578           Print the compiler's target machine (for example,
13579           i686-pc-linux-gnu)---and don't do anything else.
13580
13581       -dumpversion
13582           Print the compiler version (for example, 3.0, 6.3.0 or 7)---and
13583           don't do anything else.  This is the compiler version used in
13584           filesystem paths and specs. Depending on how the compiler has been
13585           configured it can be just a single number (major version), two
13586           numbers separated by a dot (major and minor version) or three
13587           numbers separated by dots (major, minor and patchlevel version).
13588
13589       -dumpfullversion
13590           Print the full compiler version---and don't do anything else. The
13591           output is always three numbers separated by dots, major, minor and
13592           patchlevel version.
13593
13594       -dumpspecs
13595           Print the compiler's built-in specs---and don't do anything else.
13596           (This is used when GCC itself is being built.)
13597
13598   Machine-Dependent Options
13599       Each target machine supported by GCC can have its own options---for
13600       example, to allow you to compile for a particular processor variant or
13601       ABI, or to control optimizations specific to that machine.  By
13602       convention, the names of machine-specific options start with -m.
13603
13604       Some configurations of the compiler also support additional target-
13605       specific options, usually for compatibility with other compilers on the
13606       same platform.
13607
13608       AArch64 Options
13609
13610       These options are defined for AArch64 implementations:
13611
13612       -mabi=name
13613           Generate code for the specified data model.  Permissible values are
13614           ilp32 for SysV-like data model where int, long int and pointers are
13615           32 bits, and lp64 for SysV-like data model where int is 32 bits,
13616           but long int and pointers are 64 bits.
13617
13618           The default depends on the specific target configuration.  Note
13619           that the LP64 and ILP32 ABIs are not link-compatible; you must
13620           compile your entire program with the same ABI, and link with a
13621           compatible set of libraries.
13622
13623       -mbig-endian
13624           Generate big-endian code.  This is the default when GCC is
13625           configured for an aarch64_be-*-* target.
13626
13627       -mgeneral-regs-only
13628           Generate code which uses only the general-purpose registers.  This
13629           will prevent the compiler from using floating-point and Advanced
13630           SIMD registers but will not impose any restrictions on the
13631           assembler.
13632
13633       -mlittle-endian
13634           Generate little-endian code.  This is the default when GCC is
13635           configured for an aarch64-*-* but not an aarch64_be-*-* target.
13636
13637       -mcmodel=tiny
13638           Generate code for the tiny code model.  The program and its
13639           statically defined symbols must be within 1MB of each other.
13640           Programs can be statically or dynamically linked.
13641
13642       -mcmodel=small
13643           Generate code for the small code model.  The program and its
13644           statically defined symbols must be within 4GB of each other.
13645           Programs can be statically or dynamically linked.  This is the
13646           default code model.
13647
13648       -mcmodel=large
13649           Generate code for the large code model.  This makes no assumptions
13650           about addresses and sizes of sections.  Programs can be statically
13651           linked only.
13652
13653       -mstrict-align
13654       -mno-strict-align
13655           Avoid or allow generating memory accesses that may not be aligned
13656           on a natural object boundary as described in the architecture
13657           specification.
13658
13659       -momit-leaf-frame-pointer
13660       -mno-omit-leaf-frame-pointer
13661           Omit or keep the frame pointer in leaf functions.  The former
13662           behavior is the default.
13663
13664       -mstack-protector-guard=guard
13665       -mstack-protector-guard-reg=reg
13666       -mstack-protector-guard-offset=offset
13667           Generate stack protection code using canary at guard.  Supported
13668           locations are global for a global canary or sysreg for a canary in
13669           an appropriate system register.
13670
13671           With the latter choice the options -mstack-protector-guard-reg=reg
13672           and -mstack-protector-guard-offset=offset furthermore specify which
13673           system register to use as base register for reading the canary, and
13674           from what offset from that base register. There is no default
13675           register or offset as this is entirely for use within the Linux
13676           kernel.
13677
13678       -mstack-protector-guard=guard
13679       -mstack-protector-guard-reg=reg
13680       -mstack-protector-guard-offset=offset
13681           Generate stack protection code using canary at guard.  Supported
13682           locations are global for a global canary or sysreg for a canary in
13683           an appropriate system register.
13684
13685           With the latter choice the options -mstack-protector-guard-reg=reg
13686           and -mstack-protector-guard-offset=offset furthermore specify which
13687           system register to use as base register for reading the canary, and
13688           from what offset from that base register. There is no default
13689           register or offset as this is entirely for use within the Linux
13690           kernel.
13691
13692       -mtls-dialect=desc
13693           Use TLS descriptors as the thread-local storage mechanism for
13694           dynamic accesses of TLS variables.  This is the default.
13695
13696       -mtls-dialect=traditional
13697           Use traditional TLS as the thread-local storage mechanism for
13698           dynamic accesses of TLS variables.
13699
13700       -mtls-size=size
13701           Specify bit size of immediate TLS offsets.  Valid values are 12,
13702           24, 32, 48.  This option requires binutils 2.26 or newer.
13703
13704       -mfix-cortex-a53-835769
13705       -mno-fix-cortex-a53-835769
13706           Enable or disable the workaround for the ARM Cortex-A53 erratum
13707           number 835769.  This involves inserting a NOP instruction between
13708           memory instructions and 64-bit integer multiply-accumulate
13709           instructions.
13710
13711       -mfix-cortex-a53-843419
13712       -mno-fix-cortex-a53-843419
13713           Enable or disable the workaround for the ARM Cortex-A53 erratum
13714           number 843419.  This erratum workaround is made at link time and
13715           this will only pass the corresponding flag to the linker.
13716
13717       -mlow-precision-recip-sqrt
13718       -mno-low-precision-recip-sqrt
13719           Enable or disable the reciprocal square root approximation.  This
13720           option only has an effect if -ffast-math or
13721           -funsafe-math-optimizations is used as well.  Enabling this reduces
13722           precision of reciprocal square root results to about 16 bits for
13723           single precision and to 32 bits for double precision.
13724
13725       -mlow-precision-sqrt
13726       -mno-low-precision-sqrt
13727           Enable or disable the square root approximation.  This option only
13728           has an effect if -ffast-math or -funsafe-math-optimizations is used
13729           as well.  Enabling this reduces precision of square root results to
13730           about 16 bits for single precision and to 32 bits for double
13731           precision.  If enabled, it implies -mlow-precision-recip-sqrt.
13732
13733       -mlow-precision-div
13734       -mno-low-precision-div
13735           Enable or disable the division approximation.  This option only has
13736           an effect if -ffast-math or -funsafe-math-optimizations is used as
13737           well.  Enabling this reduces precision of division results to about
13738           16 bits for single precision and to 32 bits for double precision.
13739
13740       -mtrack-speculation
13741       -mno-track-speculation
13742           Enable or disable generation of additional code to track
13743           speculative execution through conditional branches.  The tracking
13744           state can then be used by the compiler when expanding calls to
13745           "__builtin_speculation_safe_copy" to permit a more efficient code
13746           sequence to be generated.
13747
13748       -march=name
13749           Specify the name of the target architecture and, optionally, one or
13750           more feature modifiers.  This option has the form
13751           -march=arch{+[no]feature}*.
13752
13753           The permissible values for arch are armv8-a, armv8.1-a, armv8.2-a,
13754           armv8.3-a, armv8.4-a, armv8.5-a or native.
13755
13756           The value armv8.5-a implies armv8.4-a and enables compiler support
13757           for the ARMv8.5-A architecture extensions.
13758
13759           The value armv8.4-a implies armv8.3-a and enables compiler support
13760           for the ARMv8.4-A architecture extensions.
13761
13762           The value armv8.3-a implies armv8.2-a and enables compiler support
13763           for the ARMv8.3-A architecture extensions.
13764
13765           The value armv8.2-a implies armv8.1-a and enables compiler support
13766           for the ARMv8.2-A architecture extensions.
13767
13768           The value armv8.1-a implies armv8-a and enables compiler support
13769           for the ARMv8.1-A architecture extension.  In particular, it
13770           enables the +crc, +lse, and +rdma features.
13771
13772           The value native is available on native AArch64 GNU/Linux and
13773           causes the compiler to pick the architecture of the host system.
13774           This option has no effect if the compiler is unable to recognize
13775           the architecture of the host system,
13776
13777           The permissible values for feature are listed in the sub-section on
13778           aarch64-feature-modifiers,,-march and -mcpu Feature Modifiers.
13779           Where conflicting feature modifiers are specified, the right-most
13780           feature is used.
13781
13782           GCC uses name to determine what kind of instructions it can emit
13783           when generating assembly code.  If -march is specified without
13784           either of -mtune or -mcpu also being specified, the code is tuned
13785           to perform well across a range of target processors implementing
13786           the target architecture.
13787
13788       -mtune=name
13789           Specify the name of the target processor for which GCC should tune
13790           the performance of the code.  Permissible values for this option
13791           are: generic, cortex-a35, cortex-a53, cortex-a55, cortex-a57,
13792           cortex-a72, cortex-a73, cortex-a75, cortex-a76, ares, exynos-m1,
13793           emag, falkor, neoverse-e1,neoverse-n1,qdf24xx, saphira, phecda,
13794           xgene1, vulcan, octeontx, octeontx81,  octeontx83, thunderx,
13795           thunderxt88, thunderxt88p1, thunderxt81, tsv110, thunderxt83,
13796           thunderx2t99, cortex-a57.cortex-a53, cortex-a72.cortex-a53,
13797           cortex-a73.cortex-a35, cortex-a73.cortex-a53,
13798           cortex-a75.cortex-a55, cortex-a76.cortex-a55 native.
13799
13800           The values cortex-a57.cortex-a53, cortex-a72.cortex-a53,
13801           cortex-a73.cortex-a35, cortex-a73.cortex-a53,
13802           cortex-a75.cortex-a55, cortex-a76.cortex-a55 specify that GCC
13803           should tune for a big.LITTLE system.
13804
13805           Additionally on native AArch64 GNU/Linux systems the value native
13806           tunes performance to the host system.  This option has no effect if
13807           the compiler is unable to recognize the processor of the host
13808           system.
13809
13810           Where none of -mtune=, -mcpu= or -march= are specified, the code is
13811           tuned to perform well across a range of target processors.
13812
13813           This option cannot be suffixed by feature modifiers.
13814
13815       -mcpu=name
13816           Specify the name of the target processor, optionally suffixed by
13817           one or more feature modifiers.  This option has the form
13818           -mcpu=cpu{+[no]feature}*, where the permissible values for cpu are
13819           the same as those available for -mtune.  The permissible values for
13820           feature are documented in the sub-section on
13821           aarch64-feature-modifiers,,-march and -mcpu Feature Modifiers.
13822           Where conflicting feature modifiers are specified, the right-most
13823           feature is used.
13824
13825           GCC uses name to determine what kind of instructions it can emit
13826           when generating assembly code (as if by -march) and to determine
13827           the target processor for which to tune for performance (as if by
13828           -mtune).  Where this option is used in conjunction with -march or
13829           -mtune, those options take precedence over the appropriate part of
13830           this option.
13831
13832       -moverride=string
13833           Override tuning decisions made by the back-end in response to a
13834           -mtune= switch.  The syntax, semantics, and accepted values for
13835           string in this option are not guaranteed to be consistent across
13836           releases.
13837
13838           This option is only intended to be useful when developing GCC.
13839
13840       -mverbose-cost-dump
13841           Enable verbose cost model dumping in the debug dump files.  This
13842           option is provided for use in debugging the compiler.
13843
13844       -mpc-relative-literal-loads
13845       -mno-pc-relative-literal-loads
13846           Enable or disable PC-relative literal loads.  With this option
13847           literal pools are accessed using a single instruction and emitted
13848           after each function.  This limits the maximum size of functions to
13849           1MB.  This is enabled by default for -mcmodel=tiny.
13850
13851       -msign-return-address=scope
13852           Select the function scope on which return address signing will be
13853           applied.  Permissible values are none, which disables return
13854           address signing, non-leaf, which enables pointer signing for
13855           functions which are not leaf functions, and all, which enables
13856           pointer signing for all functions.  The default value is none. This
13857           option has been deprecated by -mbranch-protection.
13858
13859       -mbranch-protection=none|standard|pac-ret[+leaf]|bti
13860           Select the branch protection features to use.  none is the default
13861           and turns off all types of branch protection.  standard turns on
13862           all types of branch protection features.  If a feature has
13863           additional tuning options, then standard sets it to its standard
13864           level.  pac-ret[+leaf] turns on return address signing to its
13865           standard level: signing functions that save the return address to
13866           memory (non-leaf functions will practically always do this) using
13867           the a-key.  The optional argument leaf can be used to extend the
13868           signing to include leaf functions.  bti turns on branch target
13869           identification mechanism.
13870
13871       -msve-vector-bits=bits
13872           Specify the number of bits in an SVE vector register.  This option
13873           only has an effect when SVE is enabled.
13874
13875           GCC supports two forms of SVE code generation: "vector-length
13876           agnostic" output that works with any size of vector register and
13877           "vector-length specific" output that allows GCC to make assumptions
13878           about the vector length when it is useful for optimization reasons.
13879           The possible values of bits are: scalable, 128, 256, 512, 1024 and
13880           2048.  Specifying scalable selects vector-length agnostic output.
13881           At present -msve-vector-bits=128 also generates vector-length
13882           agnostic output.  All other values generate vector-length specific
13883           code.  The behavior of these values may change in future releases
13884           and no value except scalable should be relied on for producing code
13885           that is portable across different hardware SVE vector lengths.
13886
13887           The default is -msve-vector-bits=scalable, which produces vector-
13888           length agnostic code.
13889
13890       -march and -mcpu Feature Modifiers
13891
13892       Feature modifiers used with -march and -mcpu can be any of the
13893       following and their inverses nofeature:
13894
13895       crc Enable CRC extension.  This is on by default for -march=armv8.1-a.
13896
13897       crypto
13898           Enable Crypto extension.  This also enables Advanced SIMD and
13899           floating-point instructions.
13900
13901       fp  Enable floating-point instructions.  This is on by default for all
13902           possible values for options -march and -mcpu.
13903
13904       simd
13905           Enable Advanced SIMD instructions.  This also enables floating-
13906           point instructions.  This is on by default for all possible values
13907           for options -march and -mcpu.
13908
13909       sve Enable Scalable Vector Extension instructions.  This also enables
13910           Advanced SIMD and floating-point instructions.
13911
13912       lse Enable Large System Extension instructions.  This is on by default
13913           for -march=armv8.1-a.
13914
13915       rdma
13916           Enable Round Double Multiply Accumulate instructions.  This is on
13917           by default for -march=armv8.1-a.
13918
13919       fp16
13920           Enable FP16 extension.  This also enables floating-point
13921           instructions.
13922
13923       fp16fml
13924           Enable FP16 fmla extension.  This also enables FP16 extensions and
13925           floating-point instructions. This option is enabled by default for
13926           -march=armv8.4-a. Use of this option with architectures prior to
13927           Armv8.2-A is not supported.
13928
13929       rcpc
13930           Enable the RcPc extension.  This does not change code generation
13931           from GCC, but is passed on to the assembler, enabling inline asm
13932           statements to use instructions from the RcPc extension.
13933
13934       dotprod
13935           Enable the Dot Product extension.  This also enables Advanced SIMD
13936           instructions.
13937
13938       aes Enable the Armv8-a aes and pmull crypto extension.  This also
13939           enables Advanced SIMD instructions.
13940
13941       sha2
13942           Enable the Armv8-a sha2 crypto extension.  This also enables
13943           Advanced SIMD instructions.
13944
13945       sha3
13946           Enable the sha512 and sha3 crypto extension.  This also enables
13947           Advanced SIMD instructions. Use of this option with architectures
13948           prior to Armv8.2-A is not supported.
13949
13950       sm4 Enable the sm3 and sm4 crypto extension.  This also enables
13951           Advanced SIMD instructions.  Use of this option with architectures
13952           prior to Armv8.2-A is not supported.
13953
13954       profile
13955           Enable the Statistical Profiling extension.  This option is only to
13956           enable the extension at the assembler level and does not affect
13957           code generation.
13958
13959       rng Enable the Armv8.5-a Random Number instructions.  This option is
13960           only to enable the extension at the assembler level and does not
13961           affect code generation.
13962
13963       memtag
13964           Enable the Armv8.5-a Memory Tagging Extensions.  This option is
13965           only to enable the extension at the assembler level and does not
13966           affect code generation.
13967
13968       sb  Enable the Armv8-a Speculation Barrier instruction.  This option is
13969           only to enable the extension at the assembler level and does not
13970           affect code generation.  This option is enabled by default for
13971           -march=armv8.5-a.
13972
13973       ssbs
13974           Enable the Armv8-a Speculative Store Bypass Safe instruction.  This
13975           option is only to enable the extension at the assembler level and
13976           does not affect code generation.  This option is enabled by default
13977           for -march=armv8.5-a.
13978
13979       predres
13980           Enable the Armv8-a Execution and Data Prediction Restriction
13981           instructions.  This option is only to enable the extension at the
13982           assembler level and does not affect code generation.  This option
13983           is enabled by default for -march=armv8.5-a.
13984
13985       Feature crypto implies aes, sha2, and simd, which implies fp.
13986       Conversely, nofp implies nosimd, which implies nocrypto, noaes and
13987       nosha2.
13988
13989       Adapteva Epiphany Options
13990
13991       These -m options are defined for Adapteva Epiphany:
13992
13993       -mhalf-reg-file
13994           Don't allocate any register in the range "r32"..."r63".  That
13995           allows code to run on hardware variants that lack these registers.
13996
13997       -mprefer-short-insn-regs
13998           Preferentially allocate registers that allow short instruction
13999           generation.  This can result in increased instruction count, so
14000           this may either reduce or increase overall code size.
14001
14002       -mbranch-cost=num
14003           Set the cost of branches to roughly num "simple" instructions.
14004           This cost is only a heuristic and is not guaranteed to produce
14005           consistent results across releases.
14006
14007       -mcmove
14008           Enable the generation of conditional moves.
14009
14010       -mnops=num
14011           Emit num NOPs before every other generated instruction.
14012
14013       -mno-soft-cmpsf
14014           For single-precision floating-point comparisons, emit an "fsub"
14015           instruction and test the flags.  This is faster than a software
14016           comparison, but can get incorrect results in the presence of NaNs,
14017           or when two different small numbers are compared such that their
14018           difference is calculated as zero.  The default is -msoft-cmpsf,
14019           which uses slower, but IEEE-compliant, software comparisons.
14020
14021       -mstack-offset=num
14022           Set the offset between the top of the stack and the stack pointer.
14023           E.g., a value of 8 means that the eight bytes in the range
14024           "sp+0...sp+7" can be used by leaf functions without stack
14025           allocation.  Values other than 8 or 16 are untested and unlikely to
14026           work.  Note also that this option changes the ABI; compiling a
14027           program with a different stack offset than the libraries have been
14028           compiled with generally does not work.  This option can be useful
14029           if you want to evaluate if a different stack offset would give you
14030           better code, but to actually use a different stack offset to build
14031           working programs, it is recommended to configure the toolchain with
14032           the appropriate --with-stack-offset=num option.
14033
14034       -mno-round-nearest
14035           Make the scheduler assume that the rounding mode has been set to
14036           truncating.  The default is -mround-nearest.
14037
14038       -mlong-calls
14039           If not otherwise specified by an attribute, assume all calls might
14040           be beyond the offset range of the "b" / "bl" instructions, and
14041           therefore load the function address into a register before
14042           performing a (otherwise direct) call.  This is the default.
14043
14044       -mshort-calls
14045           If not otherwise specified by an attribute, assume all direct calls
14046           are in the range of the "b" / "bl" instructions, so use these
14047           instructions for direct calls.  The default is -mlong-calls.
14048
14049       -msmall16
14050           Assume addresses can be loaded as 16-bit unsigned values.  This
14051           does not apply to function addresses for which -mlong-calls
14052           semantics are in effect.
14053
14054       -mfp-mode=mode
14055           Set the prevailing mode of the floating-point unit.  This
14056           determines the floating-point mode that is provided and expected at
14057           function call and return time.  Making this mode match the mode you
14058           predominantly need at function start can make your programs smaller
14059           and faster by avoiding unnecessary mode switches.
14060
14061           mode can be set to one the following values:
14062
14063           caller
14064               Any mode at function entry is valid, and retained or restored
14065               when the function returns, and when it calls other functions.
14066               This mode is useful for compiling libraries or other
14067               compilation units you might want to incorporate into different
14068               programs with different prevailing FPU modes, and the
14069               convenience of being able to use a single object file outweighs
14070               the size and speed overhead for any extra mode switching that
14071               might be needed, compared with what would be needed with a more
14072               specific choice of prevailing FPU mode.
14073
14074           truncate
14075               This is the mode used for floating-point calculations with
14076               truncating (i.e. round towards zero) rounding mode.  That
14077               includes conversion from floating point to integer.
14078
14079           round-nearest
14080               This is the mode used for floating-point calculations with
14081               round-to-nearest-or-even rounding mode.
14082
14083           int This is the mode used to perform integer calculations in the
14084               FPU, e.g.  integer multiply, or integer multiply-and-
14085               accumulate.
14086
14087           The default is -mfp-mode=caller
14088
14089       -mno-split-lohi
14090       -mno-postinc
14091       -mno-postmodify
14092           Code generation tweaks that disable, respectively, splitting of
14093           32-bit loads, generation of post-increment addresses, and
14094           generation of post-modify addresses.  The defaults are msplit-lohi,
14095           -mpost-inc, and -mpost-modify.
14096
14097       -mnovect-double
14098           Change the preferred SIMD mode to SImode.  The default is
14099           -mvect-double, which uses DImode as preferred SIMD mode.
14100
14101       -max-vect-align=num
14102           The maximum alignment for SIMD vector mode types.  num may be 4 or
14103           8.  The default is 8.  Note that this is an ABI change, even though
14104           many library function interfaces are unaffected if they don't use
14105           SIMD vector modes in places that affect size and/or alignment of
14106           relevant types.
14107
14108       -msplit-vecmove-early
14109           Split vector moves into single word moves before reload.  In theory
14110           this can give better register allocation, but so far the reverse
14111           seems to be generally the case.
14112
14113       -m1reg-reg
14114           Specify a register to hold the constant -1, which makes loading
14115           small negative constants and certain bitmasks faster.  Allowable
14116           values for reg are r43 and r63, which specify use of that register
14117           as a fixed register, and none, which means that no register is used
14118           for this purpose.  The default is -m1reg-none.
14119
14120       AMD GCN Options
14121
14122       These options are defined specifically for the AMD GCN port.
14123
14124       -march=gpu
14125       -mtune=gpu
14126           Set architecture type or tuning for gpu. Supported values for gpu
14127           are
14128
14129           fiji
14130               Compile for GCN3 Fiji devices (gfx803).
14131
14132           gfx900
14133               Compile for GCN5 Vega 10 devices (gfx900).
14134
14135       -mstack-size=bytes
14136           Specify how many bytes of stack space will be requested for each
14137           GPU thread (wave-front).  Beware that there may be many threads and
14138           limited memory available.  The size of the stack allocation may
14139           also have an impact on run-time performance.  The default is 32KB
14140           when using OpenACC or OpenMP, and 1MB otherwise.
14141
14142       ARC Options
14143
14144       The following options control the architecture variant for which code
14145       is being compiled:
14146
14147       -mbarrel-shifter
14148           Generate instructions supported by barrel shifter.  This is the
14149           default unless -mcpu=ARC601 or -mcpu=ARCEM is in effect.
14150
14151       -mjli-always
14152           Force to call a function using jli_s instruction.  This option is
14153           valid only for ARCv2 architecture.
14154
14155       -mcpu=cpu
14156           Set architecture type, register usage, and instruction scheduling
14157           parameters for cpu.  There are also shortcut alias options
14158           available for backward compatibility and convenience.  Supported
14159           values for cpu are
14160
14161           arc600
14162               Compile for ARC600.  Aliases: -mA6, -mARC600.
14163
14164           arc601
14165               Compile for ARC601.  Alias: -mARC601.
14166
14167           arc700
14168               Compile for ARC700.  Aliases: -mA7, -mARC700.  This is the
14169               default when configured with --with-cpu=arc700.
14170
14171           arcem
14172               Compile for ARC EM.
14173
14174           archs
14175               Compile for ARC HS.
14176
14177           em  Compile for ARC EM CPU with no hardware extensions.
14178
14179           em4 Compile for ARC EM4 CPU.
14180
14181           em4_dmips
14182               Compile for ARC EM4 DMIPS CPU.
14183
14184           em4_fpus
14185               Compile for ARC EM4 DMIPS CPU with the single-precision
14186               floating-point extension.
14187
14188           em4_fpuda
14189               Compile for ARC EM4 DMIPS CPU with single-precision floating-
14190               point and double assist instructions.
14191
14192           hs  Compile for ARC HS CPU with no hardware extensions except the
14193               atomic instructions.
14194
14195           hs34
14196               Compile for ARC HS34 CPU.
14197
14198           hs38
14199               Compile for ARC HS38 CPU.
14200
14201           hs38_linux
14202               Compile for ARC HS38 CPU with all hardware extensions on.
14203
14204           arc600_norm
14205               Compile for ARC 600 CPU with "norm" instructions enabled.
14206
14207           arc600_mul32x16
14208               Compile for ARC 600 CPU with "norm" and 32x16-bit multiply
14209               instructions enabled.
14210
14211           arc600_mul64
14212               Compile for ARC 600 CPU with "norm" and "mul64"-family
14213               instructions enabled.
14214
14215           arc601_norm
14216               Compile for ARC 601 CPU with "norm" instructions enabled.
14217
14218           arc601_mul32x16
14219               Compile for ARC 601 CPU with "norm" and 32x16-bit multiply
14220               instructions enabled.
14221
14222           arc601_mul64
14223               Compile for ARC 601 CPU with "norm" and "mul64"-family
14224               instructions enabled.
14225
14226           nps400
14227               Compile for ARC 700 on NPS400 chip.
14228
14229           em_mini
14230               Compile for ARC EM minimalist configuration featuring reduced
14231               register set.
14232
14233       -mdpfp
14234       -mdpfp-compact
14235           Generate double-precision FPX instructions, tuned for the compact
14236           implementation.
14237
14238       -mdpfp-fast
14239           Generate double-precision FPX instructions, tuned for the fast
14240           implementation.
14241
14242       -mno-dpfp-lrsr
14243           Disable "lr" and "sr" instructions from using FPX extension aux
14244           registers.
14245
14246       -mea
14247           Generate extended arithmetic instructions.  Currently only "divaw",
14248           "adds", "subs", and "sat16" are supported.  This is always enabled
14249           for -mcpu=ARC700.
14250
14251       -mno-mpy
14252           Do not generate "mpy"-family instructions for ARC700.  This option
14253           is deprecated.
14254
14255       -mmul32x16
14256           Generate 32x16-bit multiply and multiply-accumulate instructions.
14257
14258       -mmul64
14259           Generate "mul64" and "mulu64" instructions.  Only valid for
14260           -mcpu=ARC600.
14261
14262       -mnorm
14263           Generate "norm" instructions.  This is the default if -mcpu=ARC700
14264           is in effect.
14265
14266       -mspfp
14267       -mspfp-compact
14268           Generate single-precision FPX instructions, tuned for the compact
14269           implementation.
14270
14271       -mspfp-fast
14272           Generate single-precision FPX instructions, tuned for the fast
14273           implementation.
14274
14275       -msimd
14276           Enable generation of ARC SIMD instructions via target-specific
14277           builtins.  Only valid for -mcpu=ARC700.
14278
14279       -msoft-float
14280           This option ignored; it is provided for compatibility purposes
14281           only.  Software floating-point code is emitted by default, and this
14282           default can overridden by FPX options; -mspfp, -mspfp-compact, or
14283           -mspfp-fast for single precision, and -mdpfp, -mdpfp-compact, or
14284           -mdpfp-fast for double precision.
14285
14286       -mswap
14287           Generate "swap" instructions.
14288
14289       -matomic
14290           This enables use of the locked load/store conditional extension to
14291           implement atomic memory built-in functions.  Not available for ARC
14292           6xx or ARC EM cores.
14293
14294       -mdiv-rem
14295           Enable "div" and "rem" instructions for ARCv2 cores.
14296
14297       -mcode-density
14298           Enable code density instructions for ARC EM.  This option is on by
14299           default for ARC HS.
14300
14301       -mll64
14302           Enable double load/store operations for ARC HS cores.
14303
14304       -mtp-regno=regno
14305           Specify thread pointer register number.
14306
14307       -mmpy-option=multo
14308           Compile ARCv2 code with a multiplier design option.  You can
14309           specify the option using either a string or numeric value for
14310           multo.  wlh1 is the default value.  The recognized values are:
14311
14312           0
14313           none
14314               No multiplier available.
14315
14316           1
14317           w   16x16 multiplier, fully pipelined.  The following instructions
14318               are enabled: "mpyw" and "mpyuw".
14319
14320           2
14321           wlh1
14322               32x32 multiplier, fully pipelined (1 stage).  The following
14323               instructions are additionally enabled: "mpy", "mpyu", "mpym",
14324               "mpymu", and "mpy_s".
14325
14326           3
14327           wlh2
14328               32x32 multiplier, fully pipelined (2 stages).  The following
14329               instructions are additionally enabled: "mpy", "mpyu", "mpym",
14330               "mpymu", and "mpy_s".
14331
14332           4
14333           wlh3
14334               Two 16x16 multipliers, blocking, sequential.  The following
14335               instructions are additionally enabled: "mpy", "mpyu", "mpym",
14336               "mpymu", and "mpy_s".
14337
14338           5
14339           wlh4
14340               One 16x16 multiplier, blocking, sequential.  The following
14341               instructions are additionally enabled: "mpy", "mpyu", "mpym",
14342               "mpymu", and "mpy_s".
14343
14344           6
14345           wlh5
14346               One 32x4 multiplier, blocking, sequential.  The following
14347               instructions are additionally enabled: "mpy", "mpyu", "mpym",
14348               "mpymu", and "mpy_s".
14349
14350           7
14351           plus_dmpy
14352               ARC HS SIMD support.
14353
14354           8
14355           plus_macd
14356               ARC HS SIMD support.
14357
14358           9
14359           plus_qmacw
14360               ARC HS SIMD support.
14361
14362           This option is only available for ARCv2 cores.
14363
14364       -mfpu=fpu
14365           Enables support for specific floating-point hardware extensions for
14366           ARCv2 cores.  Supported values for fpu are:
14367
14368           fpus
14369               Enables support for single-precision floating-point hardware
14370               extensions.
14371
14372           fpud
14373               Enables support for double-precision floating-point hardware
14374               extensions.  The single-precision floating-point extension is
14375               also enabled.  Not available for ARC EM.
14376
14377           fpuda
14378               Enables support for double-precision floating-point hardware
14379               extensions using double-precision assist instructions.  The
14380               single-precision floating-point extension is also enabled.
14381               This option is only available for ARC EM.
14382
14383           fpuda_div
14384               Enables support for double-precision floating-point hardware
14385               extensions using double-precision assist instructions.  The
14386               single-precision floating-point, square-root, and divide
14387               extensions are also enabled.  This option is only available for
14388               ARC EM.
14389
14390           fpuda_fma
14391               Enables support for double-precision floating-point hardware
14392               extensions using double-precision assist instructions.  The
14393               single-precision floating-point and fused multiply and add
14394               hardware extensions are also enabled.  This option is only
14395               available for ARC EM.
14396
14397           fpuda_all
14398               Enables support for double-precision floating-point hardware
14399               extensions using double-precision assist instructions.  All
14400               single-precision floating-point hardware extensions are also
14401               enabled.  This option is only available for ARC EM.
14402
14403           fpus_div
14404               Enables support for single-precision floating-point, square-
14405               root and divide hardware extensions.
14406
14407           fpud_div
14408               Enables support for double-precision floating-point, square-
14409               root and divide hardware extensions.  This option includes
14410               option fpus_div. Not available for ARC EM.
14411
14412           fpus_fma
14413               Enables support for single-precision floating-point and fused
14414               multiply and add hardware extensions.
14415
14416           fpud_fma
14417               Enables support for double-precision floating-point and fused
14418               multiply and add hardware extensions.  This option includes
14419               option fpus_fma.  Not available for ARC EM.
14420
14421           fpus_all
14422               Enables support for all single-precision floating-point
14423               hardware extensions.
14424
14425           fpud_all
14426               Enables support for all single- and double-precision floating-
14427               point hardware extensions.  Not available for ARC EM.
14428
14429       -mirq-ctrl-saved=register-range, blink, lp_count
14430           Specifies general-purposes registers that the processor
14431           automatically saves/restores on interrupt entry and exit.
14432           register-range is specified as two registers separated by a dash.
14433           The register range always starts with "r0", the upper limit is "fp"
14434           register.  blink and lp_count are optional.  This option is only
14435           valid for ARC EM and ARC HS cores.
14436
14437       -mrgf-banked-regs=number
14438           Specifies the number of registers replicated in second register
14439           bank on entry to fast interrupt.  Fast interrupts are interrupts
14440           with the highest priority level P0.  These interrupts save only PC
14441           and STATUS32 registers to avoid memory transactions during
14442           interrupt entry and exit sequences.  Use this option when you are
14443           using fast interrupts in an ARC V2 family processor.  Permitted
14444           values are 4, 8, 16, and 32.
14445
14446       -mlpc-width=width
14447           Specify the width of the "lp_count" register.  Valid values for
14448           width are 8, 16, 20, 24, 28 and 32 bits.  The default width is
14449           fixed to 32 bits.  If the width is less than 32, the compiler does
14450           not attempt to transform loops in your program to use the zero-
14451           delay loop mechanism unless it is known that the "lp_count"
14452           register can hold the required loop-counter value.  Depending on
14453           the width specified, the compiler and run-time library might
14454           continue to use the loop mechanism for various needs.  This option
14455           defines macro "__ARC_LPC_WIDTH__" with the value of width.
14456
14457       -mrf16
14458           This option instructs the compiler to generate code for a 16-entry
14459           register file.  This option defines the "__ARC_RF16__" preprocessor
14460           macro.
14461
14462       -mbranch-index
14463           Enable use of "bi" or "bih" instructions to implement jump tables.
14464
14465       The following options are passed through to the assembler, and also
14466       define preprocessor macro symbols.
14467
14468       -mdsp-packa
14469           Passed down to the assembler to enable the DSP Pack A extensions.
14470           Also sets the preprocessor symbol "__Xdsp_packa".  This option is
14471           deprecated.
14472
14473       -mdvbf
14474           Passed down to the assembler to enable the dual Viterbi butterfly
14475           extension.  Also sets the preprocessor symbol "__Xdvbf".  This
14476           option is deprecated.
14477
14478       -mlock
14479           Passed down to the assembler to enable the locked load/store
14480           conditional extension.  Also sets the preprocessor symbol
14481           "__Xlock".
14482
14483       -mmac-d16
14484           Passed down to the assembler.  Also sets the preprocessor symbol
14485           "__Xxmac_d16".  This option is deprecated.
14486
14487       -mmac-24
14488           Passed down to the assembler.  Also sets the preprocessor symbol
14489           "__Xxmac_24".  This option is deprecated.
14490
14491       -mrtsc
14492           Passed down to the assembler to enable the 64-bit time-stamp
14493           counter extension instruction.  Also sets the preprocessor symbol
14494           "__Xrtsc".  This option is deprecated.
14495
14496       -mswape
14497           Passed down to the assembler to enable the swap byte ordering
14498           extension instruction.  Also sets the preprocessor symbol
14499           "__Xswape".
14500
14501       -mtelephony
14502           Passed down to the assembler to enable dual- and single-operand
14503           instructions for telephony.  Also sets the preprocessor symbol
14504           "__Xtelephony".  This option is deprecated.
14505
14506       -mxy
14507           Passed down to the assembler to enable the XY memory extension.
14508           Also sets the preprocessor symbol "__Xxy".
14509
14510       The following options control how the assembly code is annotated:
14511
14512       -misize
14513           Annotate assembler instructions with estimated addresses.
14514
14515       -mannotate-align
14516           Explain what alignment considerations lead to the decision to make
14517           an instruction short or long.
14518
14519       The following options are passed through to the linker:
14520
14521       -marclinux
14522           Passed through to the linker, to specify use of the "arclinux"
14523           emulation.  This option is enabled by default in tool chains built
14524           for "arc-linux-uclibc" and "arceb-linux-uclibc" targets when
14525           profiling is not requested.
14526
14527       -marclinux_prof
14528           Passed through to the linker, to specify use of the "arclinux_prof"
14529           emulation.  This option is enabled by default in tool chains built
14530           for "arc-linux-uclibc" and "arceb-linux-uclibc" targets when
14531           profiling is requested.
14532
14533       The following options control the semantics of generated code:
14534
14535       -mlong-calls
14536           Generate calls as register indirect calls, thus providing access to
14537           the full 32-bit address range.
14538
14539       -mmedium-calls
14540           Don't use less than 25-bit addressing range for calls, which is the
14541           offset available for an unconditional branch-and-link instruction.
14542           Conditional execution of function calls is suppressed, to allow use
14543           of the 25-bit range, rather than the 21-bit range with conditional
14544           branch-and-link.  This is the default for tool chains built for
14545           "arc-linux-uclibc" and "arceb-linux-uclibc" targets.
14546
14547       -G num
14548           Put definitions of externally-visible data in a small data section
14549           if that data is no bigger than num bytes.  The default value of num
14550           is 4 for any ARC configuration, or 8 when we have double load/store
14551           operations.
14552
14553       -mno-sdata
14554           Do not generate sdata references.  This is the default for tool
14555           chains built for "arc-linux-uclibc" and "arceb-linux-uclibc"
14556           targets.
14557
14558       -mvolatile-cache
14559           Use ordinarily cached memory accesses for volatile references.
14560           This is the default.
14561
14562       -mno-volatile-cache
14563           Enable cache bypass for volatile references.
14564
14565       The following options fine tune code generation:
14566
14567       -malign-call
14568           Do alignment optimizations for call instructions.
14569
14570       -mauto-modify-reg
14571           Enable the use of pre/post modify with register displacement.
14572
14573       -mbbit-peephole
14574           Enable bbit peephole2.
14575
14576       -mno-brcc
14577           This option disables a target-specific pass in arc_reorg to
14578           generate compare-and-branch ("brcc") instructions.  It has no
14579           effect on generation of these instructions driven by the combiner
14580           pass.
14581
14582       -mcase-vector-pcrel
14583           Use PC-relative switch case tables to enable case table shortening.
14584           This is the default for -Os.
14585
14586       -mcompact-casesi
14587           Enable compact "casesi" pattern.  This is the default for -Os, and
14588           only available for ARCv1 cores.  This option is deprecated.
14589
14590       -mno-cond-exec
14591           Disable the ARCompact-specific pass to generate conditional
14592           execution instructions.
14593
14594           Due to delay slot scheduling and interactions between operand
14595           numbers, literal sizes, instruction lengths, and the support for
14596           conditional execution, the target-independent pass to generate
14597           conditional execution is often lacking, so the ARC port has kept a
14598           special pass around that tries to find more conditional execution
14599           generation opportunities after register allocation, branch
14600           shortening, and delay slot scheduling have been done.  This pass
14601           generally, but not always, improves performance and code size, at
14602           the cost of extra compilation time, which is why there is an option
14603           to switch it off.  If you have a problem with call instructions
14604           exceeding their allowable offset range because they are
14605           conditionalized, you should consider using -mmedium-calls instead.
14606
14607       -mearly-cbranchsi
14608           Enable pre-reload use of the "cbranchsi" pattern.
14609
14610       -mexpand-adddi
14611           Expand "adddi3" and "subdi3" at RTL generation time into "add.f",
14612           "adc" etc.  This option is deprecated.
14613
14614       -mindexed-loads
14615           Enable the use of indexed loads.  This can be problematic because
14616           some optimizers then assume that indexed stores exist, which is not
14617           the case.
14618
14619       -mlra
14620           Enable Local Register Allocation.  This is still experimental for
14621           ARC, so by default the compiler uses standard reload (i.e.
14622           -mno-lra).
14623
14624       -mlra-priority-none
14625           Don't indicate any priority for target registers.
14626
14627       -mlra-priority-compact
14628           Indicate target register priority for r0..r3 / r12..r15.
14629
14630       -mlra-priority-noncompact
14631           Reduce target register priority for r0..r3 / r12..r15.
14632
14633       -mmillicode
14634           When optimizing for size (using -Os), prologues and epilogues that
14635           have to save or restore a large number of registers are often
14636           shortened by using call to a special function in libgcc; this is
14637           referred to as a millicode call.  As these calls can pose
14638           performance issues, and/or cause linking issues when linking in a
14639           nonstandard way, this option is provided to turn on or off
14640           millicode call generation.
14641
14642       -mcode-density-frame
14643           This option enable the compiler to emit "enter" and "leave"
14644           instructions.  These instructions are only valid for CPUs with
14645           code-density feature.
14646
14647       -mmixed-code
14648           Tweak register allocation to help 16-bit instruction generation.
14649           This generally has the effect of decreasing the average instruction
14650           size while increasing the instruction count.
14651
14652       -mq-class
14653           Enable q instruction alternatives.  This is the default for -Os.
14654
14655       -mRcq
14656           Enable Rcq constraint handling.  Most short code generation depends
14657           on this.  This is the default.
14658
14659       -mRcw
14660           Enable Rcw constraint handling.  Most ccfsm condexec mostly depends
14661           on this.  This is the default.
14662
14663       -msize-level=level
14664           Fine-tune size optimization with regards to instruction lengths and
14665           alignment.  The recognized values for level are:
14666
14667           0   No size optimization.  This level is deprecated and treated
14668               like 1.
14669
14670           1   Short instructions are used opportunistically.
14671
14672           2   In addition, alignment of loops and of code after barriers are
14673               dropped.
14674
14675           3   In addition, optional data alignment is dropped, and the option
14676               Os is enabled.
14677
14678           This defaults to 3 when -Os is in effect.  Otherwise, the behavior
14679           when this is not set is equivalent to level 1.
14680
14681       -mtune=cpu
14682           Set instruction scheduling parameters for cpu, overriding any
14683           implied by -mcpu=.
14684
14685           Supported values for cpu are
14686
14687           ARC600
14688               Tune for ARC600 CPU.
14689
14690           ARC601
14691               Tune for ARC601 CPU.
14692
14693           ARC700
14694               Tune for ARC700 CPU with standard multiplier block.
14695
14696           ARC700-xmac
14697               Tune for ARC700 CPU with XMAC block.
14698
14699           ARC725D
14700               Tune for ARC725D CPU.
14701
14702           ARC750D
14703               Tune for ARC750D CPU.
14704
14705       -mmultcost=num
14706           Cost to assume for a multiply instruction, with 4 being equal to a
14707           normal instruction.
14708
14709       -munalign-prob-threshold=probability
14710           Set probability threshold for unaligning branches.  When tuning for
14711           ARC700 and optimizing for speed, branches without filled delay slot
14712           are preferably emitted unaligned and long, unless profiling
14713           indicates that the probability for the branch to be taken is below
14714           probability.  The default is (REG_BR_PROB_BASE/2), i.e. 5000.
14715
14716       The following options are maintained for backward compatibility, but
14717       are now deprecated and will be removed in a future release:
14718
14719       -margonaut
14720           Obsolete FPX.
14721
14722       -mbig-endian
14723       -EB Compile code for big-endian targets.  Use of these options is now
14724           deprecated.  Big-endian code is supported by configuring GCC to
14725           build "arceb-elf32" and "arceb-linux-uclibc" targets, for which big
14726           endian is the default.
14727
14728       -mlittle-endian
14729       -EL Compile code for little-endian targets.  Use of these options is
14730           now deprecated.  Little-endian code is supported by configuring GCC
14731           to build "arc-elf32" and "arc-linux-uclibc" targets, for which
14732           little endian is the default.
14733
14734       -mbarrel_shifter
14735           Replaced by -mbarrel-shifter.
14736
14737       -mdpfp_compact
14738           Replaced by -mdpfp-compact.
14739
14740       -mdpfp_fast
14741           Replaced by -mdpfp-fast.
14742
14743       -mdsp_packa
14744           Replaced by -mdsp-packa.
14745
14746       -mEA
14747           Replaced by -mea.
14748
14749       -mmac_24
14750           Replaced by -mmac-24.
14751
14752       -mmac_d16
14753           Replaced by -mmac-d16.
14754
14755       -mspfp_compact
14756           Replaced by -mspfp-compact.
14757
14758       -mspfp_fast
14759           Replaced by -mspfp-fast.
14760
14761       -mtune=cpu
14762           Values arc600, arc601, arc700 and arc700-xmac for cpu are replaced
14763           by ARC600, ARC601, ARC700 and ARC700-xmac respectively.
14764
14765       -multcost=num
14766           Replaced by -mmultcost.
14767
14768       ARM Options
14769
14770       These -m options are defined for the ARM port:
14771
14772       -mabi=name
14773           Generate code for the specified ABI.  Permissible values are: apcs-
14774           gnu, atpcs, aapcs, aapcs-linux and iwmmxt.
14775
14776       -mapcs-frame
14777           Generate a stack frame that is compliant with the ARM Procedure
14778           Call Standard for all functions, even if this is not strictly
14779           necessary for correct execution of the code.  Specifying
14780           -fomit-frame-pointer with this option causes the stack frames not
14781           to be generated for leaf functions.  The default is
14782           -mno-apcs-frame.  This option is deprecated.
14783
14784       -mapcs
14785           This is a synonym for -mapcs-frame and is deprecated.
14786
14787       -mthumb-interwork
14788           Generate code that supports calling between the ARM and Thumb
14789           instruction sets.  Without this option, on pre-v5 architectures,
14790           the two instruction sets cannot be reliably used inside one
14791           program.  The default is -mno-thumb-interwork, since slightly
14792           larger code is generated when -mthumb-interwork is specified.  In
14793           AAPCS configurations this option is meaningless.
14794
14795       -mno-sched-prolog
14796           Prevent the reordering of instructions in the function prologue, or
14797           the merging of those instruction with the instructions in the
14798           function's body.  This means that all functions start with a
14799           recognizable set of instructions (or in fact one of a choice from a
14800           small set of different function prologues), and this information
14801           can be used to locate the start of functions inside an executable
14802           piece of code.  The default is -msched-prolog.
14803
14804       -mfloat-abi=name
14805           Specifies which floating-point ABI to use.  Permissible values are:
14806           soft, softfp and hard.
14807
14808           Specifying soft causes GCC to generate output containing library
14809           calls for floating-point operations.  softfp allows the generation
14810           of code using hardware floating-point instructions, but still uses
14811           the soft-float calling conventions.  hard allows generation of
14812           floating-point instructions and uses FPU-specific calling
14813           conventions.
14814
14815           The default depends on the specific target configuration.  Note
14816           that the hard-float and soft-float ABIs are not link-compatible;
14817           you must compile your entire program with the same ABI, and link
14818           with a compatible set of libraries.
14819
14820       -mgeneral-regs-only
14821           Generate code which uses only the general-purpose registers.  This
14822           will prevent the compiler from using floating-point and Advanced
14823           SIMD registers but will not impose any restrictions on the
14824           assembler.
14825
14826       -mlittle-endian
14827           Generate code for a processor running in little-endian mode.  This
14828           is the default for all standard configurations.
14829
14830       -mbig-endian
14831           Generate code for a processor running in big-endian mode; the
14832           default is to compile code for a little-endian processor.
14833
14834       -mbe8
14835       -mbe32
14836           When linking a big-endian image select between BE8 and BE32
14837           formats.  The option has no effect for little-endian images and is
14838           ignored.  The default is dependent on the selected target
14839           architecture.  For ARMv6 and later architectures the default is
14840           BE8, for older architectures the default is BE32.  BE32 format has
14841           been deprecated by ARM.
14842
14843       -march=name[+extension...]
14844           This specifies the name of the target ARM architecture.  GCC uses
14845           this name to determine what kind of instructions it can emit when
14846           generating assembly code.  This option can be used in conjunction
14847           with or instead of the -mcpu= option.
14848
14849           Permissible names are: armv4t, armv5t, armv5te, armv6, armv6j,
14850           armv6k, armv6kz, armv6t2, armv6z, armv6zk, armv7, armv7-a, armv7ve,
14851           armv8-a, armv8.1-a, armv8.2-a, armv8.3-a, armv8.4-a, armv8.5-a,
14852           armv7-r, armv8-r, armv6-m, armv6s-m, armv7-m, armv7e-m,
14853           armv8-m.base, armv8-m.main, iwmmxt and iwmmxt2.
14854
14855           Additionally, the following architectures, which lack support for
14856           the Thumb execution state, are recognized but support is
14857           deprecated: armv4.
14858
14859           Many of the architectures support extensions.  These can be added
14860           by appending +extension to the architecture name.  Extension
14861           options are processed in order and capabilities accumulate.  An
14862           extension will also enable any necessary base extensions upon which
14863           it depends.  For example, the +crypto extension will always enable
14864           the +simd extension.  The exception to the additive construction is
14865           for extensions that are prefixed with +no...: these extensions
14866           disable the specified option and any other extensions that may
14867           depend on the presence of that extension.
14868
14869           For example, -march=armv7-a+simd+nofp+vfpv4 is equivalent to
14870           writing -march=armv7-a+vfpv4 since the +simd option is entirely
14871           disabled by the +nofp option that follows it.
14872
14873           Most extension names are generically named, but have an effect that
14874           is dependent upon the architecture to which it is applied.  For
14875           example, the +simd option can be applied to both armv7-a and
14876           armv8-a architectures, but will enable the original ARMv7-A
14877           Advanced SIMD (Neon) extensions for armv7-a and the ARMv8-A variant
14878           for armv8-a.
14879
14880           The table below lists the supported extensions for each
14881           architecture.  Architectures not mentioned do not support any
14882           extensions.
14883
14884           armv5te
14885           armv6
14886           armv6j
14887           armv6k
14888           armv6kz
14889           armv6t2
14890           armv6z
14891           armv6zk
14892               +fp The VFPv2 floating-point instructions.  The extension
14893                   +vfpv2 can be used as an alias for this extension.
14894
14895               +nofp
14896                   Disable the floating-point instructions.
14897
14898           armv7
14899               The common subset of the ARMv7-A, ARMv7-R and ARMv7-M
14900               architectures.
14901
14902               +fp The VFPv3 floating-point instructions, with 16 double-
14903                   precision registers.  The extension +vfpv3-d16 can be used
14904                   as an alias for this extension.  Note that floating-point
14905                   is not supported by the base ARMv7-M architecture, but is
14906                   compatible with both the ARMv7-A and ARMv7-R architectures.
14907
14908               +nofp
14909                   Disable the floating-point instructions.
14910
14911           armv7-a
14912               +mp The multiprocessing extension.
14913
14914               +sec
14915                   The security extension.
14916
14917               +fp The VFPv3 floating-point instructions, with 16 double-
14918                   precision registers.  The extension +vfpv3-d16 can be used
14919                   as an alias for this extension.
14920
14921               +simd
14922                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
14923                   instructions.  The extensions +neon and +neon-vfpv3 can be
14924                   used as aliases for this extension.
14925
14926               +vfpv3
14927                   The VFPv3 floating-point instructions, with 32 double-
14928                   precision registers.
14929
14930               +vfpv3-d16-fp16
14931                   The VFPv3 floating-point instructions, with 16 double-
14932                   precision registers and the half-precision floating-point
14933                   conversion operations.
14934
14935               +vfpv3-fp16
14936                   The VFPv3 floating-point instructions, with 32 double-
14937                   precision registers and the half-precision floating-point
14938                   conversion operations.
14939
14940               +vfpv4-d16
14941                   The VFPv4 floating-point instructions, with 16 double-
14942                   precision registers.
14943
14944               +vfpv4
14945                   The VFPv4 floating-point instructions, with 32 double-
14946                   precision registers.
14947
14948               +neon-fp16
14949                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
14950                   instructions, with the half-precision floating-point
14951                   conversion operations.
14952
14953               +neon-vfpv4
14954                   The Advanced SIMD (Neon) v2 and the VFPv4 floating-point
14955                   instructions.
14956
14957               +nosimd
14958                   Disable the Advanced SIMD instructions (does not disable
14959                   floating point).
14960
14961               +nofp
14962                   Disable the floating-point and Advanced SIMD instructions.
14963
14964           armv7ve
14965               The extended version of the ARMv7-A architecture with support
14966               for virtualization.
14967
14968               +fp The VFPv4 floating-point instructions, with 16 double-
14969                   precision registers.  The extension +vfpv4-d16 can be used
14970                   as an alias for this extension.
14971
14972               +simd
14973                   The Advanced SIMD (Neon) v2 and the VFPv4 floating-point
14974                   instructions.  The extension +neon-vfpv4 can be used as an
14975                   alias for this extension.
14976
14977               +vfpv3-d16
14978                   The VFPv3 floating-point instructions, with 16 double-
14979                   precision registers.
14980
14981               +vfpv3
14982                   The VFPv3 floating-point instructions, with 32 double-
14983                   precision registers.
14984
14985               +vfpv3-d16-fp16
14986                   The VFPv3 floating-point instructions, with 16 double-
14987                   precision registers and the half-precision floating-point
14988                   conversion operations.
14989
14990               +vfpv3-fp16
14991                   The VFPv3 floating-point instructions, with 32 double-
14992                   precision registers and the half-precision floating-point
14993                   conversion operations.
14994
14995               +vfpv4-d16
14996                   The VFPv4 floating-point instructions, with 16 double-
14997                   precision registers.
14998
14999               +vfpv4
15000                   The VFPv4 floating-point instructions, with 32 double-
15001                   precision registers.
15002
15003               +neon
15004                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
15005                   instructions.  The extension +neon-vfpv3 can be used as an
15006                   alias for this extension.
15007
15008               +neon-fp16
15009                   The Advanced SIMD (Neon) v1 and the VFPv3 floating-point
15010                   instructions, with the half-precision floating-point
15011                   conversion operations.
15012
15013               +nosimd
15014                   Disable the Advanced SIMD instructions (does not disable
15015                   floating point).
15016
15017               +nofp
15018                   Disable the floating-point and Advanced SIMD instructions.
15019
15020           armv8-a
15021               +crc
15022                   The Cyclic Redundancy Check (CRC) instructions.
15023
15024               +simd
15025                   The ARMv8-A Advanced SIMD and floating-point instructions.
15026
15027               +crypto
15028                   The cryptographic instructions.
15029
15030               +nocrypto
15031                   Disable the cryptographic instructions.
15032
15033               +nofp
15034                   Disable the floating-point, Advanced SIMD and cryptographic
15035                   instructions.
15036
15037               +sb Speculation Barrier Instruction.
15038
15039               +predres
15040                   Execution and Data Prediction Restriction Instructions.
15041
15042           armv8.1-a
15043               +simd
15044                   The ARMv8.1-A Advanced SIMD and floating-point
15045                   instructions.
15046
15047               +crypto
15048                   The cryptographic instructions.  This also enables the
15049                   Advanced SIMD and floating-point instructions.
15050
15051               +nocrypto
15052                   Disable the cryptographic instructions.
15053
15054               +nofp
15055                   Disable the floating-point, Advanced SIMD and cryptographic
15056                   instructions.
15057
15058               +sb Speculation Barrier Instruction.
15059
15060               +predres
15061                   Execution and Data Prediction Restriction Instructions.
15062
15063           armv8.2-a
15064           armv8.3-a
15065               +fp16
15066                   The half-precision floating-point data processing
15067                   instructions.  This also enables the Advanced SIMD and
15068                   floating-point instructions.
15069
15070               +fp16fml
15071                   The half-precision floating-point fmla extension.  This
15072                   also enables the half-precision floating-point extension
15073                   and Advanced SIMD and floating-point instructions.
15074
15075               +simd
15076                   The ARMv8.1-A Advanced SIMD and floating-point
15077                   instructions.
15078
15079               +crypto
15080                   The cryptographic instructions.  This also enables the
15081                   Advanced SIMD and floating-point instructions.
15082
15083               +dotprod
15084                   Enable the Dot Product extension.  This also enables
15085                   Advanced SIMD instructions.
15086
15087               +nocrypto
15088                   Disable the cryptographic extension.
15089
15090               +nofp
15091                   Disable the floating-point, Advanced SIMD and cryptographic
15092                   instructions.
15093
15094               +sb Speculation Barrier Instruction.
15095
15096               +predres
15097                   Execution and Data Prediction Restriction Instructions.
15098
15099           armv8.4-a
15100               +fp16
15101                   The half-precision floating-point data processing
15102                   instructions.  This also enables the Advanced SIMD and
15103                   floating-point instructions as well as the Dot Product
15104                   extension and the half-precision floating-point fmla
15105                   extension.
15106
15107               +simd
15108                   The ARMv8.3-A Advanced SIMD and floating-point instructions
15109                   as well as the Dot Product extension.
15110
15111               +crypto
15112                   The cryptographic instructions.  This also enables the
15113                   Advanced SIMD and floating-point instructions as well as
15114                   the Dot Product extension.
15115
15116               +nocrypto
15117                   Disable the cryptographic extension.
15118
15119               +nofp
15120                   Disable the floating-point, Advanced SIMD and cryptographic
15121                   instructions.
15122
15123               +sb Speculation Barrier Instruction.
15124
15125               +predres
15126                   Execution and Data Prediction Restriction Instructions.
15127
15128           armv8.5-a
15129               +fp16
15130                   The half-precision floating-point data processing
15131                   instructions.  This also enables the Advanced SIMD and
15132                   floating-point instructions as well as the Dot Product
15133                   extension and the half-precision floating-point fmla
15134                   extension.
15135
15136               +simd
15137                   The ARMv8.3-A Advanced SIMD and floating-point instructions
15138                   as well as the Dot Product extension.
15139
15140               +crypto
15141                   The cryptographic instructions.  This also enables the
15142                   Advanced SIMD and floating-point instructions as well as
15143                   the Dot Product extension.
15144
15145               +nocrypto
15146                   Disable the cryptographic extension.
15147
15148               +nofp
15149                   Disable the floating-point, Advanced SIMD and cryptographic
15150                   instructions.
15151
15152           armv7-r
15153               +fp.sp
15154                   The single-precision VFPv3 floating-point instructions.
15155                   The extension +vfpv3xd can be used as an alias for this
15156                   extension.
15157
15158               +fp The VFPv3 floating-point instructions with 16 double-
15159                   precision registers.  The extension +vfpv3-d16 can be used
15160                   as an alias for this extension.
15161
15162               +vfpv3xd-d16-fp16
15163                   The single-precision VFPv3 floating-point instructions with
15164                   16 double-precision registers and the half-precision
15165                   floating-point conversion operations.
15166
15167               +vfpv3-d16-fp16
15168                   The VFPv3 floating-point instructions with 16 double-
15169                   precision registers and the half-precision floating-point
15170                   conversion operations.
15171
15172               +nofp
15173                   Disable the floating-point extension.
15174
15175               +idiv
15176                   The ARM-state integer division instructions.
15177
15178               +noidiv
15179                   Disable the ARM-state integer division extension.
15180
15181           armv7e-m
15182               +fp The single-precision VFPv4 floating-point instructions.
15183
15184               +fpv5
15185                   The single-precision FPv5 floating-point instructions.
15186
15187               +fp.dp
15188                   The single- and double-precision FPv5 floating-point
15189                   instructions.
15190
15191               +nofp
15192                   Disable the floating-point extensions.
15193
15194           armv8-m.main
15195               +dsp
15196                   The DSP instructions.
15197
15198               +nodsp
15199                   Disable the DSP extension.
15200
15201               +fp The single-precision floating-point instructions.
15202
15203               +fp.dp
15204                   The single- and double-precision floating-point
15205                   instructions.
15206
15207               +nofp
15208                   Disable the floating-point extension.
15209
15210           armv8-r
15211               +crc
15212                   The Cyclic Redundancy Check (CRC) instructions.
15213
15214               +fp.sp
15215                   The single-precision FPv5 floating-point instructions.
15216
15217               +simd
15218                   The ARMv8-A Advanced SIMD and floating-point instructions.
15219
15220               +crypto
15221                   The cryptographic instructions.
15222
15223               +nocrypto
15224                   Disable the cryptographic instructions.
15225
15226               +nofp
15227                   Disable the floating-point, Advanced SIMD and cryptographic
15228                   instructions.
15229
15230           -march=native causes the compiler to auto-detect the architecture
15231           of the build computer.  At present, this feature is only supported
15232           on GNU/Linux, and not all architectures are recognized.  If the
15233           auto-detect is unsuccessful the option has no effect.
15234
15235       -mtune=name
15236           This option specifies the name of the target ARM processor for
15237           which GCC should tune the performance of the code.  For some ARM
15238           implementations better performance can be obtained by using this
15239           option.  Permissible names are: arm7tdmi, arm7tdmi-s, arm710t,
15240           arm720t, arm740t, strongarm, strongarm110, strongarm1100,
15241           0strongarm1110, arm8, arm810, arm9, arm9e, arm920, arm920t,
15242           arm922t, arm946e-s, arm966e-s, arm968e-s, arm926ej-s, arm940t,
15243           arm9tdmi, arm10tdmi, arm1020t, arm1026ej-s, arm10e, arm1020e,
15244           arm1022e, arm1136j-s, arm1136jf-s, mpcore, mpcorenovfp,
15245           arm1156t2-s, arm1156t2f-s, arm1176jz-s, arm1176jzf-s,
15246           generic-armv7-a, cortex-a5, cortex-a7, cortex-a8, cortex-a9,
15247           cortex-a12, cortex-a15, cortex-a17, cortex-a32, cortex-a35,
15248           cortex-a53, cortex-a55, cortex-a57, cortex-a72, cortex-a73,
15249           cortex-a75, cortex-a76, ares, cortex-r4, cortex-r4f, cortex-r5,
15250           cortex-r7, cortex-r8, cortex-r52, cortex-m0, cortex-m0plus,
15251           cortex-m1, cortex-m3, cortex-m4, cortex-m7, cortex-m23, cortex-m33,
15252           cortex-m1.small-multiply, cortex-m0.small-multiply,
15253           cortex-m0plus.small-multiply, exynos-m1, marvell-pj4, neoverse-n1,
15254           xscale, iwmmxt, iwmmxt2, ep9312, fa526, fa626, fa606te, fa626te,
15255           fmp626, fa726te, xgene1.
15256
15257           Additionally, this option can specify that GCC should tune the
15258           performance of the code for a big.LITTLE system.  Permissible names
15259           are: cortex-a15.cortex-a7, cortex-a17.cortex-a7,
15260           cortex-a57.cortex-a53, cortex-a72.cortex-a53,
15261           cortex-a72.cortex-a35, cortex-a73.cortex-a53,
15262           cortex-a75.cortex-a55, cortex-a76.cortex-a55.
15263
15264           -mtune=generic-arch specifies that GCC should tune the performance
15265           for a blend of processors within architecture arch.  The aim is to
15266           generate code that run well on the current most popular processors,
15267           balancing between optimizations that benefit some CPUs in the
15268           range, and avoiding performance pitfalls of other CPUs.  The
15269           effects of this option may change in future GCC versions as CPU
15270           models come and go.
15271
15272           -mtune permits the same extension options as -mcpu, but the
15273           extension options do not affect the tuning of the generated code.
15274
15275           -mtune=native causes the compiler to auto-detect the CPU of the
15276           build computer.  At present, this feature is only supported on
15277           GNU/Linux, and not all architectures are recognized.  If the auto-
15278           detect is unsuccessful the option has no effect.
15279
15280       -mcpu=name[+extension...]
15281           This specifies the name of the target ARM processor.  GCC uses this
15282           name to derive the name of the target ARM architecture (as if
15283           specified by -march) and the ARM processor type for which to tune
15284           for performance (as if specified by -mtune).  Where this option is
15285           used in conjunction with -march or -mtune, those options take
15286           precedence over the appropriate part of this option.
15287
15288           Many of the supported CPUs implement optional architectural
15289           extensions.  Where this is so the architectural extensions are
15290           normally enabled by default.  If implementations that lack the
15291           extension exist, then the extension syntax can be used to disable
15292           those extensions that have been omitted.  For floating-point and
15293           Advanced SIMD (Neon) instructions, the settings of the options
15294           -mfloat-abi and -mfpu must also be considered: floating-point and
15295           Advanced SIMD instructions will only be used if -mfloat-abi is not
15296           set to soft; and any setting of -mfpu other than auto will override
15297           the available floating-point and SIMD extension instructions.
15298
15299           For example, cortex-a9 can be found in three major configurations:
15300           integer only, with just a floating-point unit or with floating-
15301           point and Advanced SIMD.  The default is to enable all the
15302           instructions, but the extensions +nosimd and +nofp can be used to
15303           disable just the SIMD or both the SIMD and floating-point
15304           instructions respectively.
15305
15306           Permissible names for this option are the same as those for -mtune.
15307
15308           The following extension options are common to the listed CPUs:
15309
15310           +nodsp
15311               Disable the DSP instructions on cortex-m33.
15312
15313           +nofp
15314               Disables the floating-point instructions on arm9e, arm946e-s,
15315               arm966e-s, arm968e-s, arm10e, arm1020e, arm1022e, arm926ej-s,
15316               arm1026ej-s, cortex-r5, cortex-r7, cortex-r8, cortex-m4,
15317               cortex-m7 and cortex-m33.  Disables the floating-point and SIMD
15318               instructions on generic-armv7-a, cortex-a5, cortex-a7,
15319               cortex-a8, cortex-a9, cortex-a12, cortex-a15, cortex-a17,
15320               cortex-a15.cortex-a7, cortex-a17.cortex-a7, cortex-a32,
15321               cortex-a35, cortex-a53 and cortex-a55.
15322
15323           +nofp.dp
15324               Disables the double-precision component of the floating-point
15325               instructions on cortex-r5, cortex-r7, cortex-r8, cortex-r52 and
15326               cortex-m7.
15327
15328           +nosimd
15329               Disables the SIMD (but not floating-point) instructions on
15330               generic-armv7-a, cortex-a5, cortex-a7 and cortex-a9.
15331
15332           +crypto
15333               Enables the cryptographic instructions on cortex-a32,
15334               cortex-a35, cortex-a53, cortex-a55, cortex-a57, cortex-a72,
15335               cortex-a73, cortex-a75, exynos-m1, xgene1,
15336               cortex-a57.cortex-a53, cortex-a72.cortex-a53,
15337               cortex-a73.cortex-a35, cortex-a73.cortex-a53 and
15338               cortex-a75.cortex-a55.
15339
15340           Additionally the generic-armv7-a pseudo target defaults to VFPv3
15341           with 16 double-precision registers.  It supports the following
15342           extension options: mp, sec, vfpv3-d16, vfpv3, vfpv3-d16-fp16,
15343           vfpv3-fp16, vfpv4-d16, vfpv4, neon, neon-vfpv3, neon-fp16,
15344           neon-vfpv4.  The meanings are the same as for the extensions to
15345           -march=armv7-a.
15346
15347           -mcpu=generic-arch is also permissible, and is equivalent to
15348           -march=arch -mtune=generic-arch.  See -mtune for more information.
15349
15350           -mcpu=native causes the compiler to auto-detect the CPU of the
15351           build computer.  At present, this feature is only supported on
15352           GNU/Linux, and not all architectures are recognized.  If the auto-
15353           detect is unsuccessful the option has no effect.
15354
15355       -mfpu=name
15356           This specifies what floating-point hardware (or hardware emulation)
15357           is available on the target.  Permissible names are: auto, vfpv2,
15358           vfpv3, vfpv3-fp16, vfpv3-d16, vfpv3-d16-fp16, vfpv3xd,
15359           vfpv3xd-fp16, neon-vfpv3, neon-fp16, vfpv4, vfpv4-d16, fpv4-sp-d16,
15360           neon-vfpv4, fpv5-d16, fpv5-sp-d16, fp-armv8, neon-fp-armv8 and
15361           crypto-neon-fp-armv8.  Note that neon is an alias for neon-vfpv3
15362           and vfp is an alias for vfpv2.
15363
15364           The setting auto is the default and is special.  It causes the
15365           compiler to select the floating-point and Advanced SIMD
15366           instructions based on the settings of -mcpu and -march.
15367
15368           If the selected floating-point hardware includes the NEON extension
15369           (e.g. -mfpu=neon), note that floating-point operations are not
15370           generated by GCC's auto-vectorization pass unless
15371           -funsafe-math-optimizations is also specified.  This is because
15372           NEON hardware does not fully implement the IEEE 754 standard for
15373           floating-point arithmetic (in particular denormal values are
15374           treated as zero), so the use of NEON instructions may lead to a
15375           loss of precision.
15376
15377           You can also set the fpu name at function level by using the
15378           "target("fpu=")" function attributes or pragmas.
15379
15380       -mfp16-format=name
15381           Specify the format of the "__fp16" half-precision floating-point
15382           type.  Permissible names are none, ieee, and alternative; the
15383           default is none, in which case the "__fp16" type is not defined.
15384
15385       -mstructure-size-boundary=n
15386           The sizes of all structures and unions are rounded up to a multiple
15387           of the number of bits set by this option.  Permissible values are
15388           8, 32 and 64.  The default value varies for different toolchains.
15389           For the COFF targeted toolchain the default value is 8.  A value of
15390           64 is only allowed if the underlying ABI supports it.
15391
15392           Specifying a larger number can produce faster, more efficient code,
15393           but can also increase the size of the program.  Different values
15394           are potentially incompatible.  Code compiled with one value cannot
15395           necessarily expect to work with code or libraries compiled with
15396           another value, if they exchange information using structures or
15397           unions.
15398
15399           This option is deprecated.
15400
15401       -mabort-on-noreturn
15402           Generate a call to the function "abort" at the end of a "noreturn"
15403           function.  It is executed if the function tries to return.
15404
15405       -mlong-calls
15406       -mno-long-calls
15407           Tells the compiler to perform function calls by first loading the
15408           address of the function into a register and then performing a
15409           subroutine call on this register.  This switch is needed if the
15410           target function lies outside of the 64-megabyte addressing range of
15411           the offset-based version of subroutine call instruction.
15412
15413           Even if this switch is enabled, not all function calls are turned
15414           into long calls.  The heuristic is that static functions, functions
15415           that have the "short_call" attribute, functions that are inside the
15416           scope of a "#pragma no_long_calls" directive, and functions whose
15417           definitions have already been compiled within the current
15418           compilation unit are not turned into long calls.  The exceptions to
15419           this rule are that weak function definitions, functions with the
15420           "long_call" attribute or the "section" attribute, and functions
15421           that are within the scope of a "#pragma long_calls" directive are
15422           always turned into long calls.
15423
15424           This feature is not enabled by default.  Specifying -mno-long-calls
15425           restores the default behavior, as does placing the function calls
15426           within the scope of a "#pragma long_calls_off" directive.  Note
15427           these switches have no effect on how the compiler generates code to
15428           handle function calls via function pointers.
15429
15430       -msingle-pic-base
15431           Treat the register used for PIC addressing as read-only, rather
15432           than loading it in the prologue for each function.  The runtime
15433           system is responsible for initializing this register with an
15434           appropriate value before execution begins.
15435
15436       -mpic-register=reg
15437           Specify the register to be used for PIC addressing.  For standard
15438           PIC base case, the default is any suitable register determined by
15439           compiler.  For single PIC base case, the default is R9 if target is
15440           EABI based or stack-checking is enabled, otherwise the default is
15441           R10.
15442
15443       -mpic-data-is-text-relative
15444           Assume that the displacement between the text and data segments is
15445           fixed at static link time.  This permits using PC-relative
15446           addressing operations to access data known to be in the data
15447           segment.  For non-VxWorks RTP targets, this option is enabled by
15448           default.  When disabled on such targets, it will enable
15449           -msingle-pic-base by default.
15450
15451       -mpoke-function-name
15452           Write the name of each function into the text section, directly
15453           preceding the function prologue.  The generated code is similar to
15454           this:
15455
15456                        t0
15457                            .ascii "arm_poke_function_name", 0
15458                            .align
15459                        t1
15460                            .word 0xff000000 + (t1 - t0)
15461                        arm_poke_function_name
15462                            mov     ip, sp
15463                            stmfd   sp!, {fp, ip, lr, pc}
15464                            sub     fp, ip, #4
15465
15466           When performing a stack backtrace, code can inspect the value of
15467           "pc" stored at "fp + 0".  If the trace function then looks at
15468           location "pc - 12" and the top 8 bits are set, then we know that
15469           there is a function name embedded immediately preceding this
15470           location and has length "((pc[-3]) & 0xff000000)".
15471
15472       -mthumb
15473       -marm
15474           Select between generating code that executes in ARM and Thumb
15475           states.  The default for most configurations is to generate code
15476           that executes in ARM state, but the default can be changed by
15477           configuring GCC with the --with-mode=state configure option.
15478
15479           You can also override the ARM and Thumb mode for each function by
15480           using the "target("thumb")" and "target("arm")" function attributes
15481           or pragmas.
15482
15483       -mflip-thumb
15484           Switch ARM/Thumb modes on alternating functions.  This option is
15485           provided for regression testing of mixed Thumb/ARM code generation,
15486           and is not intended for ordinary use in compiling code.
15487
15488       -mtpcs-frame
15489           Generate a stack frame that is compliant with the Thumb Procedure
15490           Call Standard for all non-leaf functions.  (A leaf function is one
15491           that does not call any other functions.)  The default is
15492           -mno-tpcs-frame.
15493
15494       -mtpcs-leaf-frame
15495           Generate a stack frame that is compliant with the Thumb Procedure
15496           Call Standard for all leaf functions.  (A leaf function is one that
15497           does not call any other functions.)  The default is
15498           -mno-apcs-leaf-frame.
15499
15500       -mcallee-super-interworking
15501           Gives all externally visible functions in the file being compiled
15502           an ARM instruction set header which switches to Thumb mode before
15503           executing the rest of the function.  This allows these functions to
15504           be called from non-interworking code.  This option is not valid in
15505           AAPCS configurations because interworking is enabled by default.
15506
15507       -mcaller-super-interworking
15508           Allows calls via function pointers (including virtual functions) to
15509           execute correctly regardless of whether the target code has been
15510           compiled for interworking or not.  There is a small overhead in the
15511           cost of executing a function pointer if this option is enabled.
15512           This option is not valid in AAPCS configurations because
15513           interworking is enabled by default.
15514
15515       -mtp=name
15516           Specify the access model for the thread local storage pointer.  The
15517           valid models are soft, which generates calls to "__aeabi_read_tp",
15518           cp15, which fetches the thread pointer from "cp15" directly
15519           (supported in the arm6k architecture), and auto, which uses the
15520           best available method for the selected processor.  The default
15521           setting is auto.
15522
15523       -mtls-dialect=dialect
15524           Specify the dialect to use for accessing thread local storage.  Two
15525           dialects are supported---gnu and gnu2.  The gnu dialect selects the
15526           original GNU scheme for supporting local and global dynamic TLS
15527           models.  The gnu2 dialect selects the GNU descriptor scheme, which
15528           provides better performance for shared libraries.  The GNU
15529           descriptor scheme is compatible with the original scheme, but does
15530           require new assembler, linker and library support.  Initial and
15531           local exec TLS models are unaffected by this option and always use
15532           the original scheme.
15533
15534       -mword-relocations
15535           Only generate absolute relocations on word-sized values (i.e.
15536           R_ARM_ABS32).  This is enabled by default on targets (uClinux,
15537           SymbianOS) where the runtime loader imposes this restriction, and
15538           when -fpic or -fPIC is specified. This option conflicts with
15539           -mslow-flash-data.
15540
15541       -mfix-cortex-m3-ldrd
15542           Some Cortex-M3 cores can cause data corruption when "ldrd"
15543           instructions with overlapping destination and base registers are
15544           used.  This option avoids generating these instructions.  This
15545           option is enabled by default when -mcpu=cortex-m3 is specified.
15546
15547       -munaligned-access
15548       -mno-unaligned-access
15549           Enables (or disables) reading and writing of 16- and 32- bit values
15550           from addresses that are not 16- or 32- bit aligned.  By default
15551           unaligned access is disabled for all pre-ARMv6, all ARMv6-M and for
15552           ARMv8-M Baseline architectures, and enabled for all other
15553           architectures.  If unaligned access is not enabled then words in
15554           packed data structures are accessed a byte at a time.
15555
15556           The ARM attribute "Tag_CPU_unaligned_access" is set in the
15557           generated object file to either true or false, depending upon the
15558           setting of this option.  If unaligned access is enabled then the
15559           preprocessor symbol "__ARM_FEATURE_UNALIGNED" is also defined.
15560
15561       -mneon-for-64bits
15562           Enables using Neon to handle scalar 64-bits operations. This is
15563           disabled by default since the cost of moving data from core
15564           registers to Neon is high.
15565
15566       -mslow-flash-data
15567           Assume loading data from flash is slower than fetching instruction.
15568           Therefore literal load is minimized for better performance.  This
15569           option is only supported when compiling for ARMv7 M-profile and off
15570           by default. It conflicts with -mword-relocations.
15571
15572       -masm-syntax-unified
15573           Assume inline assembler is using unified asm syntax.  The default
15574           is currently off which implies divided syntax.  This option has no
15575           impact on Thumb2. However, this may change in future releases of
15576           GCC.  Divided syntax should be considered deprecated.
15577
15578       -mrestrict-it
15579           Restricts generation of IT blocks to conform to the rules of
15580           ARMv8-A.  IT blocks can only contain a single 16-bit instruction
15581           from a select set of instructions. This option is on by default for
15582           ARMv8-A Thumb mode.
15583
15584       -mprint-tune-info
15585           Print CPU tuning information as comment in assembler file.  This is
15586           an option used only for regression testing of the compiler and not
15587           intended for ordinary use in compiling code.  This option is
15588           disabled by default.
15589
15590       -mverbose-cost-dump
15591           Enable verbose cost model dumping in the debug dump files.  This
15592           option is provided for use in debugging the compiler.
15593
15594       -mpure-code
15595           Do not allow constant data to be placed in code sections.
15596           Additionally, when compiling for ELF object format give all text
15597           sections the ELF processor-specific section attribute
15598           "SHF_ARM_PURECODE".  This option is only available when generating
15599           non-pic code for M-profile targets with the MOVT instruction.
15600
15601       -mcmse
15602           Generate secure code as per the "ARMv8-M Security Extensions:
15603           Requirements on Development Tools Engineering Specification", which
15604           can be found on
15605           <http://infocenter.arm.com/help/topic/com.arm.doc.ecm0359818/ECM0359818_armv8m_security_extensions_reqs_on_dev_tools_1_0.pdf>.
15606
15607       AVR Options
15608
15609       These options are defined for AVR implementations:
15610
15611       -mmcu=mcu
15612           Specify Atmel AVR instruction set architectures (ISA) or MCU type.
15613
15614           The default for this option is@tie{}avr2.
15615
15616           GCC supports the following AVR devices and ISAs:
15617
15618           "avr2"
15619               "Classic" devices with up to 8@tie{}KiB of program memory.
15620               mcu@tie{}= "attiny22", "attiny26", "at90c8534", "at90s2313",
15621               "at90s2323", "at90s2333", "at90s2343", "at90s4414",
15622               "at90s4433", "at90s4434", "at90s8515", "at90s8535".
15623
15624           "avr25"
15625               "Classic" devices with up to 8@tie{}KiB of program memory and
15626               with the "MOVW" instruction.  mcu@tie{}= "ata5272", "ata6616c",
15627               "attiny13", "attiny13a", "attiny2313", "attiny2313a",
15628               "attiny24", "attiny24a", "attiny25", "attiny261", "attiny261a",
15629               "attiny43u", "attiny4313", "attiny44", "attiny44a",
15630               "attiny441", "attiny45", "attiny461", "attiny461a", "attiny48",
15631               "attiny828", "attiny84", "attiny84a", "attiny841", "attiny85",
15632               "attiny861", "attiny861a", "attiny87", "attiny88", "at86rf401".
15633
15634           "avr3"
15635               "Classic" devices with 16@tie{}KiB up to 64@tie{}KiB of
15636               program memory.  mcu@tie{}= "at43usb355", "at76c711".
15637
15638           "avr31"
15639               "Classic" devices with 128@tie{}KiB of program memory.
15640               mcu@tie{}= "atmega103", "at43usb320".
15641
15642           "avr35"
15643               "Classic" devices with 16@tie{}KiB up to 64@tie{}KiB of program
15644               memory and with the "MOVW" instruction.  mcu@tie{}= "ata5505",
15645               "ata6617c", "ata664251", "atmega16u2", "atmega32u2",
15646               "atmega8u2", "attiny1634", "attiny167", "at90usb162",
15647               "at90usb82".
15648
15649           "avr4"
15650               "Enhanced" devices with up to 8@tie{}KiB of program memory.
15651               mcu@tie{}= "ata6285", "ata6286", "ata6289", "ata6612c",
15652               "atmega48", "atmega48a", "atmega48p", "atmega48pa",
15653               "atmega48pb", "atmega8", "atmega8a", "atmega8hva",
15654               "atmega8515", "atmega8535", "atmega88", "atmega88a",
15655               "atmega88p", "atmega88pa", "atmega88pb", "at90pwm1",
15656               "at90pwm2", "at90pwm2b", "at90pwm3", "at90pwm3b", "at90pwm81".
15657
15658           "avr5"
15659               "Enhanced" devices with 16@tie{}KiB up to 64@tie{}KiB of
15660               program memory.  mcu@tie{}= "ata5702m322", "ata5782",
15661               "ata5790", "ata5790n", "ata5791", "ata5795", "ata5831",
15662               "ata6613c", "ata6614q", "ata8210", "ata8510", "atmega16",
15663               "atmega16a", "atmega16hva", "atmega16hva2", "atmega16hvb",
15664               "atmega16hvbrevb", "atmega16m1", "atmega16u4", "atmega161",
15665               "atmega162", "atmega163", "atmega164a", "atmega164p",
15666               "atmega164pa", "atmega165", "atmega165a", "atmega165p",
15667               "atmega165pa", "atmega168", "atmega168a", "atmega168p",
15668               "atmega168pa", "atmega168pb", "atmega169", "atmega169a",
15669               "atmega169p", "atmega169pa", "atmega32", "atmega32a",
15670               "atmega32c1", "atmega32hvb", "atmega32hvbrevb", "atmega32m1",
15671               "atmega32u4", "atmega32u6", "atmega323", "atmega324a",
15672               "atmega324p", "atmega324pa", "atmega325", "atmega325a",
15673               "atmega325p", "atmega325pa", "atmega3250", "atmega3250a",
15674               "atmega3250p", "atmega3250pa", "atmega328", "atmega328p",
15675               "atmega328pb", "atmega329", "atmega329a", "atmega329p",
15676               "atmega329pa", "atmega3290", "atmega3290a", "atmega3290p",
15677               "atmega3290pa", "atmega406", "atmega64", "atmega64a",
15678               "atmega64c1", "atmega64hve", "atmega64hve2", "atmega64m1",
15679               "atmega64rfr2", "atmega640", "atmega644", "atmega644a",
15680               "atmega644p", "atmega644pa", "atmega644rfr2", "atmega645",
15681               "atmega645a", "atmega645p", "atmega6450", "atmega6450a",
15682               "atmega6450p", "atmega649", "atmega649a", "atmega649p",
15683               "atmega6490", "atmega6490a", "atmega6490p", "at90can32",
15684               "at90can64", "at90pwm161", "at90pwm216", "at90pwm316",
15685               "at90scr100", "at90usb646", "at90usb647", "at94k", "m3000".
15686
15687           "avr51"
15688               "Enhanced" devices with 128@tie{}KiB of program memory.
15689               mcu@tie{}= "atmega128", "atmega128a", "atmega128rfa1",
15690               "atmega128rfr2", "atmega1280", "atmega1281", "atmega1284",
15691               "atmega1284p", "atmega1284rfr2", "at90can128", "at90usb1286",
15692               "at90usb1287".
15693
15694           "avr6"
15695               "Enhanced" devices with 3-byte PC, i.e. with more than
15696               128@tie{}KiB of program memory.  mcu@tie{}= "atmega256rfr2",
15697               "atmega2560", "atmega2561", "atmega2564rfr2".
15698
15699           "avrxmega2"
15700               "XMEGA" devices with more than 8@tie{}KiB and up to 64@tie{}KiB
15701               of program memory.  mcu@tie{}= "atxmega16a4", "atxmega16a4u",
15702               "atxmega16c4", "atxmega16d4", "atxmega16e5", "atxmega32a4",
15703               "atxmega32a4u", "atxmega32c3", "atxmega32c4", "atxmega32d3",
15704               "atxmega32d4", "atxmega32e5", "atxmega8e5".
15705
15706           "avrxmega3"
15707               "XMEGA" devices with up to 64@tie{}KiB of combined program
15708               memory and RAM, and with program memory visible in the RAM
15709               address space.  mcu@tie{}= "attiny1614", "attiny1616",
15710               "attiny1617", "attiny212", "attiny214", "attiny3214",
15711               "attiny3216", "attiny3217", "attiny412", "attiny414",
15712               "attiny416", "attiny417", "attiny814", "attiny816",
15713               "attiny817".
15714
15715           "avrxmega4"
15716               "XMEGA" devices with more than 64@tie{}KiB and up to
15717               128@tie{}KiB of program memory.  mcu@tie{}= "atxmega64a3",
15718               "atxmega64a3u", "atxmega64a4u", "atxmega64b1", "atxmega64b3",
15719               "atxmega64c3", "atxmega64d3", "atxmega64d4".
15720
15721           "avrxmega5"
15722               "XMEGA" devices with more than 64@tie{}KiB and up to
15723               128@tie{}KiB of program memory and more than 64@tie{}KiB of
15724               RAM.  mcu@tie{}= "atxmega64a1", "atxmega64a1u".
15725
15726           "avrxmega6"
15727               "XMEGA" devices with more than 128@tie{}KiB of program memory.
15728               mcu@tie{}= "atxmega128a3", "atxmega128a3u", "atxmega128b1",
15729               "atxmega128b3", "atxmega128c3", "atxmega128d3", "atxmega128d4",
15730               "atxmega192a3", "atxmega192a3u", "atxmega192c3",
15731               "atxmega192d3", "atxmega256a3", "atxmega256a3b",
15732               "atxmega256a3bu", "atxmega256a3u", "atxmega256c3",
15733               "atxmega256d3", "atxmega384c3", "atxmega384d3".
15734
15735           "avrxmega7"
15736               "XMEGA" devices with more than 128@tie{}KiB of program memory
15737               and more than 64@tie{}KiB of RAM.  mcu@tie{}= "atxmega128a1",
15738               "atxmega128a1u", "atxmega128a4u".
15739
15740           "avrtiny"
15741               "TINY" Tiny core devices with 512@tie{}B up to 4@tie{}KiB of
15742               program memory.  mcu@tie{}= "attiny10", "attiny20", "attiny4",
15743               "attiny40", "attiny5", "attiny9".
15744
15745           "avr1"
15746               This ISA is implemented by the minimal AVR core and supported
15747               for assembler only.  mcu@tie{}= "attiny11", "attiny12",
15748               "attiny15", "attiny28", "at90s1200".
15749
15750       -mabsdata
15751           Assume that all data in static storage can be accessed by LDS / STS
15752           instructions.  This option has only an effect on reduced Tiny
15753           devices like ATtiny40.  See also the "absdata" AVR Variable
15754           Attributes,variable attribute.
15755
15756       -maccumulate-args
15757           Accumulate outgoing function arguments and acquire/release the
15758           needed stack space for outgoing function arguments once in function
15759           prologue/epilogue.  Without this option, outgoing arguments are
15760           pushed before calling a function and popped afterwards.
15761
15762           Popping the arguments after the function call can be expensive on
15763           AVR so that accumulating the stack space might lead to smaller
15764           executables because arguments need not be removed from the stack
15765           after such a function call.
15766
15767           This option can lead to reduced code size for functions that
15768           perform several calls to functions that get their arguments on the
15769           stack like calls to printf-like functions.
15770
15771       -mbranch-cost=cost
15772           Set the branch costs for conditional branch instructions to cost.
15773           Reasonable values for cost are small, non-negative integers. The
15774           default branch cost is 0.
15775
15776       -mcall-prologues
15777           Functions prologues/epilogues are expanded as calls to appropriate
15778           subroutines.  Code size is smaller.
15779
15780       -mgas-isr-prologues
15781           Interrupt service routines (ISRs) may use the "__gcc_isr" pseudo
15782           instruction supported by GNU Binutils.  If this option is on, the
15783           feature can still be disabled for individual ISRs by means of the
15784           AVR Function Attributes,,"no_gccisr" function attribute.  This
15785           feature is activated per default if optimization is on (but not
15786           with -Og, @pxref{Optimize Options}), and if GNU Binutils support
15787           PR21683 ("https://sourceware.org/PR21683").
15788
15789       -mint8
15790           Assume "int" to be 8-bit integer.  This affects the sizes of all
15791           types: a "char" is 1 byte, an "int" is 1 byte, a "long" is 2 bytes,
15792           and "long long" is 4 bytes.  Please note that this option does not
15793           conform to the C standards, but it results in smaller code size.
15794
15795       -mmain-is-OS_task
15796           Do not save registers in "main".  The effect is the same like
15797           attaching attribute AVR Function Attributes,,"OS_task" to "main".
15798           It is activated per default if optimization is on.
15799
15800       -mn-flash=num
15801           Assume that the flash memory has a size of num times 64@tie{}KiB.
15802
15803       -mno-interrupts
15804           Generated code is not compatible with hardware interrupts.  Code
15805           size is smaller.
15806
15807       -mrelax
15808           Try to replace "CALL" resp. "JMP" instruction by the shorter
15809           "RCALL" resp. "RJMP" instruction if applicable.  Setting -mrelax
15810           just adds the --mlink-relax option to the assembler's command line
15811           and the --relax option to the linker's command line.
15812
15813           Jump relaxing is performed by the linker because jump offsets are
15814           not known before code is located. Therefore, the assembler code
15815           generated by the compiler is the same, but the instructions in the
15816           executable may differ from instructions in the assembler code.
15817
15818           Relaxing must be turned on if linker stubs are needed, see the
15819           section on "EIND" and linker stubs below.
15820
15821       -mrmw
15822           Assume that the device supports the Read-Modify-Write instructions
15823           "XCH", "LAC", "LAS" and "LAT".
15824
15825       -mshort-calls
15826           Assume that "RJMP" and "RCALL" can target the whole program memory.
15827
15828           This option is used internally for multilib selection.  It is not
15829           an optimization option, and you don't need to set it by hand.
15830
15831       -msp8
15832           Treat the stack pointer register as an 8-bit register, i.e. assume
15833           the high byte of the stack pointer is zero.  In general, you don't
15834           need to set this option by hand.
15835
15836           This option is used internally by the compiler to select and build
15837           multilibs for architectures "avr2" and "avr25".  These
15838           architectures mix devices with and without "SPH".  For any setting
15839           other than -mmcu=avr2 or -mmcu=avr25 the compiler driver adds or
15840           removes this option from the compiler proper's command line,
15841           because the compiler then knows if the device or architecture has
15842           an 8-bit stack pointer and thus no "SPH" register or not.
15843
15844       -mstrict-X
15845           Use address register "X" in a way proposed by the hardware.  This
15846           means that "X" is only used in indirect, post-increment or pre-
15847           decrement addressing.
15848
15849           Without this option, the "X" register may be used in the same way
15850           as "Y" or "Z" which then is emulated by additional instructions.
15851           For example, loading a value with "X+const" addressing with a small
15852           non-negative "const < 64" to a register Rn is performed as
15853
15854                   adiw r26, const   ; X += const
15855                   ld   <Rn>, X        ; <Rn> = *X
15856                   sbiw r26, const   ; X -= const
15857
15858       -mtiny-stack
15859           Only change the lower 8@tie{}bits of the stack pointer.
15860
15861       -mfract-convert-truncate
15862           Allow to use truncation instead of rounding towards zero for
15863           fractional fixed-point types.
15864
15865       -nodevicelib
15866           Don't link against AVR-LibC's device specific library "lib<mcu>.a".
15867
15868       -Waddr-space-convert
15869           Warn about conversions between address spaces in the case where the
15870           resulting address space is not contained in the incoming address
15871           space.
15872
15873       -Wmisspelled-isr
15874           Warn if the ISR is misspelled, i.e. without __vector prefix.
15875           Enabled by default.
15876
15877       "EIND" and Devices with More Than 128 Ki Bytes of Flash
15878
15879       Pointers in the implementation are 16@tie{}bits wide.  The address of a
15880       function or label is represented as word address so that indirect jumps
15881       and calls can target any code address in the range of 64@tie{}Ki words.
15882
15883       In order to facilitate indirect jump on devices with more than
15884       128@tie{}Ki bytes of program memory space, there is a special function
15885       register called "EIND" that serves as most significant part of the
15886       target address when "EICALL" or "EIJMP" instructions are used.
15887
15888       Indirect jumps and calls on these devices are handled as follows by the
15889       compiler and are subject to some limitations:
15890
15891       *   The compiler never sets "EIND".
15892
15893       *   The compiler uses "EIND" implicitly in "EICALL"/"EIJMP"
15894           instructions or might read "EIND" directly in order to emulate an
15895           indirect call/jump by means of a "RET" instruction.
15896
15897       *   The compiler assumes that "EIND" never changes during the startup
15898           code or during the application. In particular, "EIND" is not
15899           saved/restored in function or interrupt service routine
15900           prologue/epilogue.
15901
15902       *   For indirect calls to functions and computed goto, the linker
15903           generates stubs. Stubs are jump pads sometimes also called
15904           trampolines. Thus, the indirect call/jump jumps to such a stub.
15905           The stub contains a direct jump to the desired address.
15906
15907       *   Linker relaxation must be turned on so that the linker generates
15908           the stubs correctly in all situations. See the compiler option
15909           -mrelax and the linker option --relax.  There are corner cases
15910           where the linker is supposed to generate stubs but aborts without
15911           relaxation and without a helpful error message.
15912
15913       *   The default linker script is arranged for code with "EIND = 0".  If
15914           code is supposed to work for a setup with "EIND != 0", a custom
15915           linker script has to be used in order to place the sections whose
15916           name start with ".trampolines" into the segment where "EIND" points
15917           to.
15918
15919       *   The startup code from libgcc never sets "EIND".  Notice that
15920           startup code is a blend of code from libgcc and AVR-LibC.  For the
15921           impact of AVR-LibC on "EIND", see the AVR-LibC user manual
15922           ("http://nongnu.org/avr-libc/user-manual/").
15923
15924       *   It is legitimate for user-specific startup code to set up "EIND"
15925           early, for example by means of initialization code located in
15926           section ".init3". Such code runs prior to general startup code that
15927           initializes RAM and calls constructors, but after the bit of
15928           startup code from AVR-LibC that sets "EIND" to the segment where
15929           the vector table is located.
15930
15931                   #include <avr/io.h>
15932
15933                   static void
15934                   __attribute__((section(".init3"),naked,used,no_instrument_function))
15935                   init3_set_eind (void)
15936                   {
15937                     __asm volatile ("ldi r24,pm_hh8(__trampolines_start)\n\t"
15938                                     "out %i0,r24" :: "n" (&EIND) : "r24","memory");
15939                   }
15940
15941           The "__trampolines_start" symbol is defined in the linker script.
15942
15943       *   Stubs are generated automatically by the linker if the following
15944           two conditions are met:
15945
15946           -<The address of a label is taken by means of the "gs" modifier>
15947               (short for generate stubs) like so:
15948
15949                       LDI r24, lo8(gs(<func>))
15950                       LDI r25, hi8(gs(<func>))
15951
15952           -<The final location of that label is in a code segment>
15953               outside the segment where the stubs are located.
15954
15955       *   The compiler emits such "gs" modifiers for code labels in the
15956           following situations:
15957
15958           -<Taking address of a function or code label.>
15959           -<Computed goto.>
15960           -<If prologue-save function is used, see -mcall-prologues>
15961               command-line option.
15962
15963           -<Switch/case dispatch tables. If you do not want such dispatch>
15964               tables you can specify the -fno-jump-tables command-line
15965               option.
15966
15967           -<C and C++ constructors/destructors called during
15968           startup/shutdown.>
15969           -<If the tools hit a "gs()" modifier explained above.>
15970       *   Jumping to non-symbolic addresses like so is not supported:
15971
15972                   int main (void)
15973                   {
15974                       /* Call function at word address 0x2 */
15975                       return ((int(*)(void)) 0x2)();
15976                   }
15977
15978           Instead, a stub has to be set up, i.e. the function has to be
15979           called through a symbol ("func_4" in the example):
15980
15981                   int main (void)
15982                   {
15983                       extern int func_4 (void);
15984
15985                       /* Call function at byte address 0x4 */
15986                       return func_4();
15987                   }
15988
15989           and the application be linked with -Wl,--defsym,func_4=0x4.
15990           Alternatively, "func_4" can be defined in the linker script.
15991
15992       Handling of the "RAMPD", "RAMPX", "RAMPY" and "RAMPZ" Special Function
15993       Registers
15994
15995       Some AVR devices support memories larger than the 64@tie{}KiB range
15996       that can be accessed with 16-bit pointers.  To access memory locations
15997       outside this 64@tie{}KiB range, the content of a "RAMP" register is
15998       used as high part of the address: The "X", "Y", "Z" address register is
15999       concatenated with the "RAMPX", "RAMPY", "RAMPZ" special function
16000       register, respectively, to get a wide address. Similarly, "RAMPD" is
16001       used together with direct addressing.
16002
16003       *   The startup code initializes the "RAMP" special function registers
16004           with zero.
16005
16006       *   If a AVR Named Address Spaces,named address space other than
16007           generic or "__flash" is used, then "RAMPZ" is set as needed before
16008           the operation.
16009
16010       *   If the device supports RAM larger than 64@tie{}KiB and the compiler
16011           needs to change "RAMPZ" to accomplish an operation, "RAMPZ" is
16012           reset to zero after the operation.
16013
16014       *   If the device comes with a specific "RAMP" register, the ISR
16015           prologue/epilogue saves/restores that SFR and initializes it with
16016           zero in case the ISR code might (implicitly) use it.
16017
16018       *   RAM larger than 64@tie{}KiB is not supported by GCC for AVR
16019           targets.  If you use inline assembler to read from locations
16020           outside the 16-bit address range and change one of the "RAMP"
16021           registers, you must reset it to zero after the access.
16022
16023       AVR Built-in Macros
16024
16025       GCC defines several built-in macros so that the user code can test for
16026       the presence or absence of features.  Almost any of the following
16027       built-in macros are deduced from device capabilities and thus triggered
16028       by the -mmcu= command-line option.
16029
16030       For even more AVR-specific built-in macros see AVR Named Address Spaces
16031       and AVR Built-in Functions.
16032
16033       "__AVR_ARCH__"
16034           Build-in macro that resolves to a decimal number that identifies
16035           the architecture and depends on the -mmcu=mcu option.  Possible
16036           values are:
16037
16038           2, 25, 3, 31, 35, 4, 5, 51, 6
16039
16040           for mcu="avr2", "avr25", "avr3", "avr31", "avr35", "avr4", "avr5",
16041           "avr51", "avr6",
16042
16043           respectively and
16044
16045           100, 102, 103, 104, 105, 106, 107
16046
16047           for mcu="avrtiny", "avrxmega2", "avrxmega3", "avrxmega4",
16048           "avrxmega5", "avrxmega6", "avrxmega7", respectively.  If mcu
16049           specifies a device, this built-in macro is set accordingly. For
16050           example, with -mmcu=atmega8 the macro is defined to 4.
16051
16052       "__AVR_Device__"
16053           Setting -mmcu=device defines this built-in macro which reflects the
16054           device's name. For example, -mmcu=atmega8 defines the built-in
16055           macro "__AVR_ATmega8__", -mmcu=attiny261a defines
16056           "__AVR_ATtiny261A__", etc.
16057
16058           The built-in macros' names follow the scheme "__AVR_Device__" where
16059           Device is the device name as from the AVR user manual. The
16060           difference between Device in the built-in macro and device in
16061           -mmcu=device is that the latter is always lowercase.
16062
16063           If device is not a device but only a core architecture like avr51,
16064           this macro is not defined.
16065
16066       "__AVR_DEVICE_NAME__"
16067           Setting -mmcu=device defines this built-in macro to the device's
16068           name. For example, with -mmcu=atmega8 the macro is defined to
16069           "atmega8".
16070
16071           If device is not a device but only a core architecture like avr51,
16072           this macro is not defined.
16073
16074       "__AVR_XMEGA__"
16075           The device / architecture belongs to the XMEGA family of devices.
16076
16077       "__AVR_HAVE_ELPM__"
16078           The device has the "ELPM" instruction.
16079
16080       "__AVR_HAVE_ELPMX__"
16081           The device has the "ELPM Rn,Z" and "ELPM Rn,Z+" instructions.
16082
16083       "__AVR_HAVE_MOVW__"
16084           The device has the "MOVW" instruction to perform 16-bit register-
16085           register moves.
16086
16087       "__AVR_HAVE_LPMX__"
16088           The device has the "LPM Rn,Z" and "LPM Rn,Z+" instructions.
16089
16090       "__AVR_HAVE_MUL__"
16091           The device has a hardware multiplier.
16092
16093       "__AVR_HAVE_JMP_CALL__"
16094           The device has the "JMP" and "CALL" instructions.  This is the case
16095           for devices with more than 8@tie{}KiB of program memory.
16096
16097       "__AVR_HAVE_EIJMP_EICALL__"
16098       "__AVR_3_BYTE_PC__"
16099           The device has the "EIJMP" and "EICALL" instructions.  This is the
16100           case for devices with more than 128@tie{}KiB of program memory.
16101           This also means that the program counter (PC) is 3@tie{}bytes wide.
16102
16103       "__AVR_2_BYTE_PC__"
16104           The program counter (PC) is 2@tie{}bytes wide. This is the case for
16105           devices with up to 128@tie{}KiB of program memory.
16106
16107       "__AVR_HAVE_8BIT_SP__"
16108       "__AVR_HAVE_16BIT_SP__"
16109           The stack pointer (SP) register is treated as 8-bit respectively
16110           16-bit register by the compiler.  The definition of these macros is
16111           affected by -mtiny-stack.
16112
16113       "__AVR_HAVE_SPH__"
16114       "__AVR_SP8__"
16115           The device has the SPH (high part of stack pointer) special
16116           function register or has an 8-bit stack pointer, respectively.  The
16117           definition of these macros is affected by -mmcu= and in the cases
16118           of -mmcu=avr2 and -mmcu=avr25 also by -msp8.
16119
16120       "__AVR_HAVE_RAMPD__"
16121       "__AVR_HAVE_RAMPX__"
16122       "__AVR_HAVE_RAMPY__"
16123       "__AVR_HAVE_RAMPZ__"
16124           The device has the "RAMPD", "RAMPX", "RAMPY", "RAMPZ" special
16125           function register, respectively.
16126
16127       "__NO_INTERRUPTS__"
16128           This macro reflects the -mno-interrupts command-line option.
16129
16130       "__AVR_ERRATA_SKIP__"
16131       "__AVR_ERRATA_SKIP_JMP_CALL__"
16132           Some AVR devices (AT90S8515, ATmega103) must not skip 32-bit
16133           instructions because of a hardware erratum.  Skip instructions are
16134           "SBRS", "SBRC", "SBIS", "SBIC" and "CPSE".  The second macro is
16135           only defined if "__AVR_HAVE_JMP_CALL__" is also set.
16136
16137       "__AVR_ISA_RMW__"
16138           The device has Read-Modify-Write instructions (XCH, LAC, LAS and
16139           LAT).
16140
16141       "__AVR_SFR_OFFSET__=offset"
16142           Instructions that can address I/O special function registers
16143           directly like "IN", "OUT", "SBI", etc. may use a different address
16144           as if addressed by an instruction to access RAM like "LD" or "STS".
16145           This offset depends on the device architecture and has to be
16146           subtracted from the RAM address in order to get the respective
16147           I/O@tie{}address.
16148
16149       "__AVR_SHORT_CALLS__"
16150           The -mshort-calls command line option is set.
16151
16152       "__AVR_PM_BASE_ADDRESS__=addr"
16153           Some devices support reading from flash memory by means of "LD*"
16154           instructions.  The flash memory is seen in the data address space
16155           at an offset of "__AVR_PM_BASE_ADDRESS__".  If this macro is not
16156           defined, this feature is not available.  If defined, the address
16157           space is linear and there is no need to put ".rodata" into RAM.
16158           This is handled by the default linker description file, and is
16159           currently available for "avrtiny" and "avrxmega3".  Even more
16160           convenient, there is no need to use address spaces like "__flash"
16161           or features like attribute "progmem" and "pgm_read_*".
16162
16163       "__WITH_AVRLIBC__"
16164           The compiler is configured to be used together with AVR-Libc.  See
16165           the --with-avrlibc configure option.
16166
16167       Blackfin Options
16168
16169       -mcpu=cpu[-sirevision]
16170           Specifies the name of the target Blackfin processor.  Currently,
16171           cpu can be one of bf512, bf514, bf516, bf518, bf522, bf523, bf524,
16172           bf525, bf526, bf527, bf531, bf532, bf533, bf534, bf536, bf537,
16173           bf538, bf539, bf542, bf544, bf547, bf548, bf549, bf542m, bf544m,
16174           bf547m, bf548m, bf549m, bf561, bf592.
16175
16176           The optional sirevision specifies the silicon revision of the
16177           target Blackfin processor.  Any workarounds available for the
16178           targeted silicon revision are enabled.  If sirevision is none, no
16179           workarounds are enabled.  If sirevision is any, all workarounds for
16180           the targeted processor are enabled.  The "__SILICON_REVISION__"
16181           macro is defined to two hexadecimal digits representing the major
16182           and minor numbers in the silicon revision.  If sirevision is none,
16183           the "__SILICON_REVISION__" is not defined.  If sirevision is any,
16184           the "__SILICON_REVISION__" is defined to be 0xffff.  If this
16185           optional sirevision is not used, GCC assumes the latest known
16186           silicon revision of the targeted Blackfin processor.
16187
16188           GCC defines a preprocessor macro for the specified cpu.  For the
16189           bfin-elf toolchain, this option causes the hardware BSP provided by
16190           libgloss to be linked in if -msim is not given.
16191
16192           Without this option, bf532 is used as the processor by default.
16193
16194           Note that support for bf561 is incomplete.  For bf561, only the
16195           preprocessor macro is defined.
16196
16197       -msim
16198           Specifies that the program will be run on the simulator.  This
16199           causes the simulator BSP provided by libgloss to be linked in.
16200           This option has effect only for bfin-elf toolchain.  Certain other
16201           options, such as -mid-shared-library and -mfdpic, imply -msim.
16202
16203       -momit-leaf-frame-pointer
16204           Don't keep the frame pointer in a register for leaf functions.
16205           This avoids the instructions to save, set up and restore frame
16206           pointers and makes an extra register available in leaf functions.
16207
16208       -mspecld-anomaly
16209           When enabled, the compiler ensures that the generated code does not
16210           contain speculative loads after jump instructions. If this option
16211           is used, "__WORKAROUND_SPECULATIVE_LOADS" is defined.
16212
16213       -mno-specld-anomaly
16214           Don't generate extra code to prevent speculative loads from
16215           occurring.
16216
16217       -mcsync-anomaly
16218           When enabled, the compiler ensures that the generated code does not
16219           contain CSYNC or SSYNC instructions too soon after conditional
16220           branches.  If this option is used, "__WORKAROUND_SPECULATIVE_SYNCS"
16221           is defined.
16222
16223       -mno-csync-anomaly
16224           Don't generate extra code to prevent CSYNC or SSYNC instructions
16225           from occurring too soon after a conditional branch.
16226
16227       -mlow64k
16228           When enabled, the compiler is free to take advantage of the
16229           knowledge that the entire program fits into the low 64k of memory.
16230
16231       -mno-low64k
16232           Assume that the program is arbitrarily large.  This is the default.
16233
16234       -mstack-check-l1
16235           Do stack checking using information placed into L1 scratchpad
16236           memory by the uClinux kernel.
16237
16238       -mid-shared-library
16239           Generate code that supports shared libraries via the library ID
16240           method.  This allows for execute in place and shared libraries in
16241           an environment without virtual memory management.  This option
16242           implies -fPIC.  With a bfin-elf target, this option implies -msim.
16243
16244       -mno-id-shared-library
16245           Generate code that doesn't assume ID-based shared libraries are
16246           being used.  This is the default.
16247
16248       -mleaf-id-shared-library
16249           Generate code that supports shared libraries via the library ID
16250           method, but assumes that this library or executable won't link
16251           against any other ID shared libraries.  That allows the compiler to
16252           use faster code for jumps and calls.
16253
16254       -mno-leaf-id-shared-library
16255           Do not assume that the code being compiled won't link against any
16256           ID shared libraries.  Slower code is generated for jump and call
16257           insns.
16258
16259       -mshared-library-id=n
16260           Specifies the identification number of the ID-based shared library
16261           being compiled.  Specifying a value of 0 generates more compact
16262           code; specifying other values forces the allocation of that number
16263           to the current library but is no more space- or time-efficient than
16264           omitting this option.
16265
16266       -msep-data
16267           Generate code that allows the data segment to be located in a
16268           different area of memory from the text segment.  This allows for
16269           execute in place in an environment without virtual memory
16270           management by eliminating relocations against the text section.
16271
16272       -mno-sep-data
16273           Generate code that assumes that the data segment follows the text
16274           segment.  This is the default.
16275
16276       -mlong-calls
16277       -mno-long-calls
16278           Tells the compiler to perform function calls by first loading the
16279           address of the function into a register and then performing a
16280           subroutine call on this register.  This switch is needed if the
16281           target function lies outside of the 24-bit addressing range of the
16282           offset-based version of subroutine call instruction.
16283
16284           This feature is not enabled by default.  Specifying -mno-long-calls
16285           restores the default behavior.  Note these switches have no effect
16286           on how the compiler generates code to handle function calls via
16287           function pointers.
16288
16289       -mfast-fp
16290           Link with the fast floating-point library. This library relaxes
16291           some of the IEEE floating-point standard's rules for checking
16292           inputs against Not-a-Number (NAN), in the interest of performance.
16293
16294       -minline-plt
16295           Enable inlining of PLT entries in function calls to functions that
16296           are not known to bind locally.  It has no effect without -mfdpic.
16297
16298       -mmulticore
16299           Build a standalone application for multicore Blackfin processors.
16300           This option causes proper start files and link scripts supporting
16301           multicore to be used, and defines the macro "__BFIN_MULTICORE".  It
16302           can only be used with -mcpu=bf561[-sirevision].
16303
16304           This option can be used with -mcorea or -mcoreb, which selects the
16305           one-application-per-core programming model.  Without -mcorea or
16306           -mcoreb, the single-application/dual-core programming model is
16307           used. In this model, the main function of Core B should be named as
16308           "coreb_main".
16309
16310           If this option is not used, the single-core application programming
16311           model is used.
16312
16313       -mcorea
16314           Build a standalone application for Core A of BF561 when using the
16315           one-application-per-core programming model. Proper start files and
16316           link scripts are used to support Core A, and the macro
16317           "__BFIN_COREA" is defined.  This option can only be used in
16318           conjunction with -mmulticore.
16319
16320       -mcoreb
16321           Build a standalone application for Core B of BF561 when using the
16322           one-application-per-core programming model. Proper start files and
16323           link scripts are used to support Core B, and the macro
16324           "__BFIN_COREB" is defined. When this option is used, "coreb_main"
16325           should be used instead of "main".  This option can only be used in
16326           conjunction with -mmulticore.
16327
16328       -msdram
16329           Build a standalone application for SDRAM. Proper start files and
16330           link scripts are used to put the application into SDRAM, and the
16331           macro "__BFIN_SDRAM" is defined.  The loader should initialize
16332           SDRAM before loading the application.
16333
16334       -micplb
16335           Assume that ICPLBs are enabled at run time.  This has an effect on
16336           certain anomaly workarounds.  For Linux targets, the default is to
16337           assume ICPLBs are enabled; for standalone applications the default
16338           is off.
16339
16340       C6X Options
16341
16342       -march=name
16343           This specifies the name of the target architecture.  GCC uses this
16344           name to determine what kind of instructions it can emit when
16345           generating assembly code.  Permissible names are: c62x, c64x,
16346           c64x+, c67x, c67x+, c674x.
16347
16348       -mbig-endian
16349           Generate code for a big-endian target.
16350
16351       -mlittle-endian
16352           Generate code for a little-endian target.  This is the default.
16353
16354       -msim
16355           Choose startup files and linker script suitable for the simulator.
16356
16357       -msdata=default
16358           Put small global and static data in the ".neardata" section, which
16359           is pointed to by register "B14".  Put small uninitialized global
16360           and static data in the ".bss" section, which is adjacent to the
16361           ".neardata" section.  Put small read-only data into the ".rodata"
16362           section.  The corresponding sections used for large pieces of data
16363           are ".fardata", ".far" and ".const".
16364
16365       -msdata=all
16366           Put all data, not just small objects, into the sections reserved
16367           for small data, and use addressing relative to the "B14" register
16368           to access them.
16369
16370       -msdata=none
16371           Make no use of the sections reserved for small data, and use
16372           absolute addresses to access all data.  Put all initialized global
16373           and static data in the ".fardata" section, and all uninitialized
16374           data in the ".far" section.  Put all constant data into the
16375           ".const" section.
16376
16377       CRIS Options
16378
16379       These options are defined specifically for the CRIS ports.
16380
16381       -march=architecture-type
16382       -mcpu=architecture-type
16383           Generate code for the specified architecture.  The choices for
16384           architecture-type are v3, v8 and v10 for respectively ETRAX 4,
16385           ETRAX 100, and ETRAX 100 LX.  Default is v0 except for cris-axis-
16386           linux-gnu, where the default is v10.
16387
16388       -mtune=architecture-type
16389           Tune to architecture-type everything applicable about the generated
16390           code, except for the ABI and the set of available instructions.
16391           The choices for architecture-type are the same as for
16392           -march=architecture-type.
16393
16394       -mmax-stack-frame=n
16395           Warn when the stack frame of a function exceeds n bytes.
16396
16397       -metrax4
16398       -metrax100
16399           The options -metrax4 and -metrax100 are synonyms for -march=v3 and
16400           -march=v8 respectively.
16401
16402       -mmul-bug-workaround
16403       -mno-mul-bug-workaround
16404           Work around a bug in the "muls" and "mulu" instructions for CPU
16405           models where it applies.  This option is active by default.
16406
16407       -mpdebug
16408           Enable CRIS-specific verbose debug-related information in the
16409           assembly code.  This option also has the effect of turning off the
16410           #NO_APP formatted-code indicator to the assembler at the beginning
16411           of the assembly file.
16412
16413       -mcc-init
16414           Do not use condition-code results from previous instruction; always
16415           emit compare and test instructions before use of condition codes.
16416
16417       -mno-side-effects
16418           Do not emit instructions with side effects in addressing modes
16419           other than post-increment.
16420
16421       -mstack-align
16422       -mno-stack-align
16423       -mdata-align
16424       -mno-data-align
16425       -mconst-align
16426       -mno-const-align
16427           These options (no- options) arrange (eliminate arrangements) for
16428           the stack frame, individual data and constants to be aligned for
16429           the maximum single data access size for the chosen CPU model.  The
16430           default is to arrange for 32-bit alignment.  ABI details such as
16431           structure layout are not affected by these options.
16432
16433       -m32-bit
16434       -m16-bit
16435       -m8-bit
16436           Similar to the stack- data- and const-align options above, these
16437           options arrange for stack frame, writable data and constants to all
16438           be 32-bit, 16-bit or 8-bit aligned.  The default is 32-bit
16439           alignment.
16440
16441       -mno-prologue-epilogue
16442       -mprologue-epilogue
16443           With -mno-prologue-epilogue, the normal function prologue and
16444           epilogue which set up the stack frame are omitted and no return
16445           instructions or return sequences are generated in the code.  Use
16446           this option only together with visual inspection of the compiled
16447           code: no warnings or errors are generated when call-saved registers
16448           must be saved, or storage for local variables needs to be
16449           allocated.
16450
16451       -mno-gotplt
16452       -mgotplt
16453           With -fpic and -fPIC, don't generate (do generate) instruction
16454           sequences that load addresses for functions from the PLT part of
16455           the GOT rather than (traditional on other architectures) calls to
16456           the PLT.  The default is -mgotplt.
16457
16458       -melf
16459           Legacy no-op option only recognized with the cris-axis-elf and
16460           cris-axis-linux-gnu targets.
16461
16462       -mlinux
16463           Legacy no-op option only recognized with the cris-axis-linux-gnu
16464           target.
16465
16466       -sim
16467           This option, recognized for the cris-axis-elf, arranges to link
16468           with input-output functions from a simulator library.  Code,
16469           initialized data and zero-initialized data are allocated
16470           consecutively.
16471
16472       -sim2
16473           Like -sim, but pass linker options to locate initialized data at
16474           0x40000000 and zero-initialized data at 0x80000000.
16475
16476       CR16 Options
16477
16478       These options are defined specifically for the CR16 ports.
16479
16480       -mmac
16481           Enable the use of multiply-accumulate instructions. Disabled by
16482           default.
16483
16484       -mcr16cplus
16485       -mcr16c
16486           Generate code for CR16C or CR16C+ architecture. CR16C+ architecture
16487           is default.
16488
16489       -msim
16490           Links the library libsim.a which is in compatible with simulator.
16491           Applicable to ELF compiler only.
16492
16493       -mint32
16494           Choose integer type as 32-bit wide.
16495
16496       -mbit-ops
16497           Generates "sbit"/"cbit" instructions for bit manipulations.
16498
16499       -mdata-model=model
16500           Choose a data model. The choices for model are near, far or medium.
16501           medium is default.  However, far is not valid with -mcr16c, as the
16502           CR16C architecture does not support the far data model.
16503
16504       C-SKY Options
16505
16506       GCC supports these options when compiling for C-SKY V2 processors.
16507
16508       -march=arch
16509           Specify the C-SKY target architecture.  Valid values for arch are:
16510           ck801, ck802, ck803, ck807, and ck810.  The default is ck810.
16511
16512       -mcpu=cpu
16513           Specify the C-SKY target processor.  Valid values for cpu are:
16514           ck801, ck801t, ck802, ck802t, ck802j, ck803, ck803h, ck803t,
16515           ck803ht, ck803f, ck803fh, ck803e, ck803eh, ck803et, ck803eht,
16516           ck803ef, ck803efh, ck803ft, ck803eft, ck803efht, ck803r1, ck803hr1,
16517           ck803tr1, ck803htr1, ck803fr1, ck803fhr1, ck803er1, ck803ehr1,
16518           ck803etr1, ck803ehtr1, ck803efr1, ck803efhr1, ck803ftr1,
16519           ck803eftr1, ck803efhtr1, ck803s, ck803st, ck803se, ck803sf,
16520           ck803sef, ck803seft, ck807e, ck807ef, ck807, ck807f, ck810e,
16521           ck810et, ck810ef, ck810eft, ck810, ck810v, ck810f, ck810t, ck810fv,
16522           ck810tv, ck810ft, and ck810ftv.
16523
16524       -mbig-endian
16525       -EB
16526       -mlittle-endian
16527       -EL Select big- or little-endian code.  The default is little-endian.
16528
16529       -mhard-float
16530       -msoft-float
16531           Select hardware or software floating-point implementations.  The
16532           default is soft float.
16533
16534       -mdouble-float
16535       -mno-double-float
16536           When -mhard-float is in effect, enable generation of double-
16537           precision float instructions.  This is the default except when
16538           compiling for CK803.
16539
16540       -mfdivdu
16541       -mno-fdivdu
16542           When -mhard-float is in effect, enable generation of "frecipd",
16543           "fsqrtd", and "fdivd" instructions.  This is the default except
16544           when compiling for CK803.
16545
16546       -mfpu=fpu
16547           Select the floating-point processor.  This option can only be used
16548           with -mhard-float.  Values for fpu are fpv2_sf (equivalent to
16549           -mno-double-float -mno-fdivdu), fpv2 (-mdouble-float -mno-divdu),
16550           and fpv2_divd (-mdouble-float -mdivdu).
16551
16552       -melrw
16553       -mno-elrw
16554           Enable the extended "lrw" instruction.  This option defaults to on
16555           for CK801 and off otherwise.
16556
16557       -mistack
16558       -mno-istack
16559           Enable interrupt stack instructions; the default is off.
16560
16561           The -mistack option is required to handle the "interrupt" and "isr"
16562           function attributes.
16563
16564       -mmp
16565           Enable multiprocessor instructions; the default is off.
16566
16567       -mcp
16568           Enable coprocessor instructions; the default is off.
16569
16570       -mcache
16571           Enable coprocessor instructions; the default is off.
16572
16573       -msecurity
16574           Enable C-SKY security instructions; the default is off.
16575
16576       -mtrust
16577           Enable C-SKY trust instructions; the default is off.
16578
16579       -mdsp
16580       -medsp
16581       -mvdsp
16582           Enable C-SKY DSP, Enhanced DSP, or Vector DSP instructions,
16583           respectively.  All of these options default to off.
16584
16585       -mdiv
16586       -mno-div
16587           Generate divide instructions.  Default is off.
16588
16589       -msmart
16590       -mno-smart
16591           Generate code for Smart Mode, using only registers numbered 0-7 to
16592           allow use of 16-bit instructions.  This option is ignored for CK801
16593           where this is the required behavior, and it defaults to on for
16594           CK802.  For other targets, the default is off.
16595
16596       -mhigh-registers
16597       -mno-high-registers
16598           Generate code using the high registers numbered 16-31.  This option
16599           is not supported on CK801, CK802, or CK803, and is enabled by
16600           default for other processors.
16601
16602       -manchor
16603       -mno-anchor
16604           Generate code using global anchor symbol addresses.
16605
16606       -mpushpop
16607       -mno-pushpop
16608           Generate code using "push" and "pop" instructions.  This option
16609           defaults to on.
16610
16611       -mmultiple-stld
16612       -mstm
16613       -mno-multiple-stld
16614       -mno-stm
16615           Generate code using "stm" and "ldm" instructions.  This option
16616           isn't supported on CK801 but is enabled by default on other
16617           processors.
16618
16619       -mconstpool
16620       -mno-constpool
16621           Create constant pools in the compiler instead of deferring it to
16622           the assembler.  This option is the default and required for correct
16623           code generation on CK801 and CK802, and is optional on other
16624           processors.
16625
16626       -mstack-size
16627       -mno-stack-size
16628           Emit ".stack_size" directives for each function in the assembly
16629           output.  This option defaults to off.
16630
16631       -mccrt
16632       -mno-ccrt
16633           Generate code for the C-SKY compiler runtime instead of libgcc.
16634           This option defaults to off.
16635
16636       -mbranch-cost=n
16637           Set the branch costs to roughly "n" instructions.  The default is
16638           1.
16639
16640       -msched-prolog
16641       -mno-sched-prolog
16642           Permit scheduling of function prologue and epilogue sequences.
16643           Using this option can result in code that is not compliant with the
16644           C-SKY V2 ABI prologue requirements and that cannot be debugged or
16645           backtraced.  It is disabled by default.
16646
16647       Darwin Options
16648
16649       These options are defined for all architectures running the Darwin
16650       operating system.
16651
16652       FSF GCC on Darwin does not create "fat" object files; it creates an
16653       object file for the single architecture that GCC was built to target.
16654       Apple's GCC on Darwin does create "fat" files if multiple -arch options
16655       are used; it does so by running the compiler or linker multiple times
16656       and joining the results together with lipo.
16657
16658       The subtype of the file created (like ppc7400 or ppc970 or i686) is
16659       determined by the flags that specify the ISA that GCC is targeting,
16660       like -mcpu or -march.  The -force_cpusubtype_ALL option can be used to
16661       override this.
16662
16663       The Darwin tools vary in their behavior when presented with an ISA
16664       mismatch.  The assembler, as, only permits instructions to be used that
16665       are valid for the subtype of the file it is generating, so you cannot
16666       put 64-bit instructions in a ppc750 object file.  The linker for shared
16667       libraries, /usr/bin/libtool, fails and prints an error if asked to
16668       create a shared library with a less restrictive subtype than its input
16669       files (for instance, trying to put a ppc970 object file in a ppc7400
16670       library).  The linker for executables, ld, quietly gives the executable
16671       the most restrictive subtype of any of its input files.
16672
16673       -Fdir
16674           Add the framework directory dir to the head of the list of
16675           directories to be searched for header files.  These directories are
16676           interleaved with those specified by -I options and are scanned in a
16677           left-to-right order.
16678
16679           A framework directory is a directory with frameworks in it.  A
16680           framework is a directory with a Headers and/or PrivateHeaders
16681           directory contained directly in it that ends in .framework.  The
16682           name of a framework is the name of this directory excluding the
16683           .framework.  Headers associated with the framework are found in one
16684           of those two directories, with Headers being searched first.  A
16685           subframework is a framework directory that is in a framework's
16686           Frameworks directory.  Includes of subframework headers can only
16687           appear in a header of a framework that contains the subframework,
16688           or in a sibling subframework header.  Two subframeworks are
16689           siblings if they occur in the same framework.  A subframework
16690           should not have the same name as a framework; a warning is issued
16691           if this is violated.  Currently a subframework cannot have
16692           subframeworks; in the future, the mechanism may be extended to
16693           support this.  The standard frameworks can be found in
16694           /System/Library/Frameworks and /Library/Frameworks.  An example
16695           include looks like "#include <Framework/header.h>", where Framework
16696           denotes the name of the framework and header.h is found in the
16697           PrivateHeaders or Headers directory.
16698
16699       -iframeworkdir
16700           Like -F except the directory is a treated as a system directory.
16701           The main difference between this -iframework and -F is that with
16702           -iframework the compiler does not warn about constructs contained
16703           within header files found via dir.  This option is valid only for
16704           the C family of languages.
16705
16706       -gused
16707           Emit debugging information for symbols that are used.  For stabs
16708           debugging format, this enables -feliminate-unused-debug-symbols.
16709           This is by default ON.
16710
16711       -gfull
16712           Emit debugging information for all symbols and types.
16713
16714       -mmacosx-version-min=version
16715           The earliest version of MacOS X that this executable will run on is
16716           version.  Typical values of version include 10.1, 10.2, and 10.3.9.
16717
16718           If the compiler was built to use the system's headers by default,
16719           then the default for this option is the system version on which the
16720           compiler is running, otherwise the default is to make choices that
16721           are compatible with as many systems and code bases as possible.
16722
16723       -mkernel
16724           Enable kernel development mode.  The -mkernel option sets -static,
16725           -fno-common, -fno-use-cxa-atexit, -fno-exceptions,
16726           -fno-non-call-exceptions, -fapple-kext, -fno-weak and -fno-rtti
16727           where applicable.  This mode also sets -mno-altivec, -msoft-float,
16728           -fno-builtin and -mlong-branch for PowerPC targets.
16729
16730       -mone-byte-bool
16731           Override the defaults for "bool" so that "sizeof(bool)==1".  By
16732           default "sizeof(bool)" is 4 when compiling for Darwin/PowerPC and 1
16733           when compiling for Darwin/x86, so this option has no effect on x86.
16734
16735           Warning: The -mone-byte-bool switch causes GCC to generate code
16736           that is not binary compatible with code generated without that
16737           switch.  Using this switch may require recompiling all other
16738           modules in a program, including system libraries.  Use this switch
16739           to conform to a non-default data model.
16740
16741       -mfix-and-continue
16742       -ffix-and-continue
16743       -findirect-data
16744           Generate code suitable for fast turnaround development, such as to
16745           allow GDB to dynamically load .o files into already-running
16746           programs.  -findirect-data and -ffix-and-continue are provided for
16747           backwards compatibility.
16748
16749       -all_load
16750           Loads all members of static archive libraries.  See man ld(1) for
16751           more information.
16752
16753       -arch_errors_fatal
16754           Cause the errors having to do with files that have the wrong
16755           architecture to be fatal.
16756
16757       -bind_at_load
16758           Causes the output file to be marked such that the dynamic linker
16759           will bind all undefined references when the file is loaded or
16760           launched.
16761
16762       -bundle
16763           Produce a Mach-o bundle format file.  See man ld(1) for more
16764           information.
16765
16766       -bundle_loader executable
16767           This option specifies the executable that will load the build
16768           output file being linked.  See man ld(1) for more information.
16769
16770       -dynamiclib
16771           When passed this option, GCC produces a dynamic library instead of
16772           an executable when linking, using the Darwin libtool command.
16773
16774       -force_cpusubtype_ALL
16775           This causes GCC's output file to have the ALL subtype, instead of
16776           one controlled by the -mcpu or -march option.
16777
16778       -allowable_client  client_name
16779       -client_name
16780       -compatibility_version
16781       -current_version
16782       -dead_strip
16783       -dependency-file
16784       -dylib_file
16785       -dylinker_install_name
16786       -dynamic
16787       -exported_symbols_list
16788       -filelist
16789       -flat_namespace
16790       -force_flat_namespace
16791       -headerpad_max_install_names
16792       -image_base
16793       -init
16794       -install_name
16795       -keep_private_externs
16796       -multi_module
16797       -multiply_defined
16798       -multiply_defined_unused
16799       -noall_load
16800       -no_dead_strip_inits_and_terms
16801       -nofixprebinding
16802       -nomultidefs
16803       -noprebind
16804       -noseglinkedit
16805       -pagezero_size
16806       -prebind
16807       -prebind_all_twolevel_modules
16808       -private_bundle
16809       -read_only_relocs
16810       -sectalign
16811       -sectobjectsymbols
16812       -whyload
16813       -seg1addr
16814       -sectcreate
16815       -sectobjectsymbols
16816       -sectorder
16817       -segaddr
16818       -segs_read_only_addr
16819       -segs_read_write_addr
16820       -seg_addr_table
16821       -seg_addr_table_filename
16822       -seglinkedit
16823       -segprot
16824       -segs_read_only_addr
16825       -segs_read_write_addr
16826       -single_module
16827       -static
16828       -sub_library
16829       -sub_umbrella
16830       -twolevel_namespace
16831       -umbrella
16832       -undefined
16833       -unexported_symbols_list
16834       -weak_reference_mismatches
16835       -whatsloaded
16836           These options are passed to the Darwin linker.  The Darwin linker
16837           man page describes them in detail.
16838
16839       DEC Alpha Options
16840
16841       These -m options are defined for the DEC Alpha implementations:
16842
16843       -mno-soft-float
16844       -msoft-float
16845           Use (do not use) the hardware floating-point instructions for
16846           floating-point operations.  When -msoft-float is specified,
16847           functions in libgcc.a are used to perform floating-point
16848           operations.  Unless they are replaced by routines that emulate the
16849           floating-point operations, or compiled in such a way as to call
16850           such emulations routines, these routines issue floating-point
16851           operations.   If you are compiling for an Alpha without floating-
16852           point operations, you must ensure that the library is built so as
16853           not to call them.
16854
16855           Note that Alpha implementations without floating-point operations
16856           are required to have floating-point registers.
16857
16858       -mfp-reg
16859       -mno-fp-regs
16860           Generate code that uses (does not use) the floating-point register
16861           set.  -mno-fp-regs implies -msoft-float.  If the floating-point
16862           register set is not used, floating-point operands are passed in
16863           integer registers as if they were integers and floating-point
16864           results are passed in $0 instead of $f0.  This is a non-standard
16865           calling sequence, so any function with a floating-point argument or
16866           return value called by code compiled with -mno-fp-regs must also be
16867           compiled with that option.
16868
16869           A typical use of this option is building a kernel that does not
16870           use, and hence need not save and restore, any floating-point
16871           registers.
16872
16873       -mieee
16874           The Alpha architecture implements floating-point hardware optimized
16875           for maximum performance.  It is mostly compliant with the IEEE
16876           floating-point standard.  However, for full compliance, software
16877           assistance is required.  This option generates code fully IEEE-
16878           compliant code except that the inexact-flag is not maintained (see
16879           below).  If this option is turned on, the preprocessor macro
16880           "_IEEE_FP" is defined during compilation.  The resulting code is
16881           less efficient but is able to correctly support denormalized
16882           numbers and exceptional IEEE values such as not-a-number and
16883           plus/minus infinity.  Other Alpha compilers call this option
16884           -ieee_with_no_inexact.
16885
16886       -mieee-with-inexact
16887           This is like -mieee except the generated code also maintains the
16888           IEEE inexact-flag.  Turning on this option causes the generated
16889           code to implement fully-compliant IEEE math.  In addition to
16890           "_IEEE_FP", "_IEEE_FP_EXACT" is defined as a preprocessor macro.
16891           On some Alpha implementations the resulting code may execute
16892           significantly slower than the code generated by default.  Since
16893           there is very little code that depends on the inexact-flag, you
16894           should normally not specify this option.  Other Alpha compilers
16895           call this option -ieee_with_inexact.
16896
16897       -mfp-trap-mode=trap-mode
16898           This option controls what floating-point related traps are enabled.
16899           Other Alpha compilers call this option -fptm trap-mode.  The trap
16900           mode can be set to one of four values:
16901
16902           n   This is the default (normal) setting.  The only traps that are
16903               enabled are the ones that cannot be disabled in software (e.g.,
16904               division by zero trap).
16905
16906           u   In addition to the traps enabled by n, underflow traps are
16907               enabled as well.
16908
16909           su  Like u, but the instructions are marked to be safe for software
16910               completion (see Alpha architecture manual for details).
16911
16912           sui Like su, but inexact traps are enabled as well.
16913
16914       -mfp-rounding-mode=rounding-mode
16915           Selects the IEEE rounding mode.  Other Alpha compilers call this
16916           option -fprm rounding-mode.  The rounding-mode can be one of:
16917
16918           n   Normal IEEE rounding mode.  Floating-point numbers are rounded
16919               towards the nearest machine number or towards the even machine
16920               number in case of a tie.
16921
16922           m   Round towards minus infinity.
16923
16924           c   Chopped rounding mode.  Floating-point numbers are rounded
16925               towards zero.
16926
16927           d   Dynamic rounding mode.  A field in the floating-point control
16928               register (fpcr, see Alpha architecture reference manual)
16929               controls the rounding mode in effect.  The C library
16930               initializes this register for rounding towards plus infinity.
16931               Thus, unless your program modifies the fpcr, d corresponds to
16932               round towards plus infinity.
16933
16934       -mtrap-precision=trap-precision
16935           In the Alpha architecture, floating-point traps are imprecise.
16936           This means without software assistance it is impossible to recover
16937           from a floating trap and program execution normally needs to be
16938           terminated.  GCC can generate code that can assist operating system
16939           trap handlers in determining the exact location that caused a
16940           floating-point trap.  Depending on the requirements of an
16941           application, different levels of precisions can be selected:
16942
16943           p   Program precision.  This option is the default and means a trap
16944               handler can only identify which program caused a floating-point
16945               exception.
16946
16947           f   Function precision.  The trap handler can determine the
16948               function that caused a floating-point exception.
16949
16950           i   Instruction precision.  The trap handler can determine the
16951               exact instruction that caused a floating-point exception.
16952
16953           Other Alpha compilers provide the equivalent options called
16954           -scope_safe and -resumption_safe.
16955
16956       -mieee-conformant
16957           This option marks the generated code as IEEE conformant.  You must
16958           not use this option unless you also specify -mtrap-precision=i and
16959           either -mfp-trap-mode=su or -mfp-trap-mode=sui.  Its only effect is
16960           to emit the line .eflag 48 in the function prologue of the
16961           generated assembly file.
16962
16963       -mbuild-constants
16964           Normally GCC examines a 32- or 64-bit integer constant to see if it
16965           can construct it from smaller constants in two or three
16966           instructions.  If it cannot, it outputs the constant as a literal
16967           and generates code to load it from the data segment at run time.
16968
16969           Use this option to require GCC to construct all integer constants
16970           using code, even if it takes more instructions (the maximum is
16971           six).
16972
16973           You typically use this option to build a shared library dynamic
16974           loader.  Itself a shared library, it must relocate itself in memory
16975           before it can find the variables and constants in its own data
16976           segment.
16977
16978       -mbwx
16979       -mno-bwx
16980       -mcix
16981       -mno-cix
16982       -mfix
16983       -mno-fix
16984       -mmax
16985       -mno-max
16986           Indicate whether GCC should generate code to use the optional BWX,
16987           CIX, FIX and MAX instruction sets.  The default is to use the
16988           instruction sets supported by the CPU type specified via -mcpu=
16989           option or that of the CPU on which GCC was built if none is
16990           specified.
16991
16992       -mfloat-vax
16993       -mfloat-ieee
16994           Generate code that uses (does not use) VAX F and G floating-point
16995           arithmetic instead of IEEE single and double precision.
16996
16997       -mexplicit-relocs
16998       -mno-explicit-relocs
16999           Older Alpha assemblers provided no way to generate symbol
17000           relocations except via assembler macros.  Use of these macros does
17001           not allow optimal instruction scheduling.  GNU binutils as of
17002           version 2.12 supports a new syntax that allows the compiler to
17003           explicitly mark which relocations should apply to which
17004           instructions.  This option is mostly useful for debugging, as GCC
17005           detects the capabilities of the assembler when it is built and sets
17006           the default accordingly.
17007
17008       -msmall-data
17009       -mlarge-data
17010           When -mexplicit-relocs is in effect, static data is accessed via
17011           gp-relative relocations.  When -msmall-data is used, objects 8
17012           bytes long or smaller are placed in a small data area (the ".sdata"
17013           and ".sbss" sections) and are accessed via 16-bit relocations off
17014           of the $gp register.  This limits the size of the small data area
17015           to 64KB, but allows the variables to be directly accessed via a
17016           single instruction.
17017
17018           The default is -mlarge-data.  With this option the data area is
17019           limited to just below 2GB.  Programs that require more than 2GB of
17020           data must use "malloc" or "mmap" to allocate the data in the heap
17021           instead of in the program's data segment.
17022
17023           When generating code for shared libraries, -fpic implies
17024           -msmall-data and -fPIC implies -mlarge-data.
17025
17026       -msmall-text
17027       -mlarge-text
17028           When -msmall-text is used, the compiler assumes that the code of
17029           the entire program (or shared library) fits in 4MB, and is thus
17030           reachable with a branch instruction.  When -msmall-data is used,
17031           the compiler can assume that all local symbols share the same $gp
17032           value, and thus reduce the number of instructions required for a
17033           function call from 4 to 1.
17034
17035           The default is -mlarge-text.
17036
17037       -mcpu=cpu_type
17038           Set the instruction set and instruction scheduling parameters for
17039           machine type cpu_type.  You can specify either the EV style name or
17040           the corresponding chip number.  GCC supports scheduling parameters
17041           for the EV4, EV5 and EV6 family of processors and chooses the
17042           default values for the instruction set from the processor you
17043           specify.  If you do not specify a processor type, GCC defaults to
17044           the processor on which the compiler was built.
17045
17046           Supported values for cpu_type are
17047
17048           ev4
17049           ev45
17050           21064
17051               Schedules as an EV4 and has no instruction set extensions.
17052
17053           ev5
17054           21164
17055               Schedules as an EV5 and has no instruction set extensions.
17056
17057           ev56
17058           21164a
17059               Schedules as an EV5 and supports the BWX extension.
17060
17061           pca56
17062           21164pc
17063           21164PC
17064               Schedules as an EV5 and supports the BWX and MAX extensions.
17065
17066           ev6
17067           21264
17068               Schedules as an EV6 and supports the BWX, FIX, and MAX
17069               extensions.
17070
17071           ev67
17072           21264a
17073               Schedules as an EV6 and supports the BWX, CIX, FIX, and MAX
17074               extensions.
17075
17076           Native toolchains also support the value native, which selects the
17077           best architecture option for the host processor.  -mcpu=native has
17078           no effect if GCC does not recognize the processor.
17079
17080       -mtune=cpu_type
17081           Set only the instruction scheduling parameters for machine type
17082           cpu_type.  The instruction set is not changed.
17083
17084           Native toolchains also support the value native, which selects the
17085           best architecture option for the host processor.  -mtune=native has
17086           no effect if GCC does not recognize the processor.
17087
17088       -mmemory-latency=time
17089           Sets the latency the scheduler should assume for typical memory
17090           references as seen by the application.  This number is highly
17091           dependent on the memory access patterns used by the application and
17092           the size of the external cache on the machine.
17093
17094           Valid options for time are
17095
17096           number
17097               A decimal number representing clock cycles.
17098
17099           L1
17100           L2
17101           L3
17102           main
17103               The compiler contains estimates of the number of clock cycles
17104               for "typical" EV4 & EV5 hardware for the Level 1, 2 & 3 caches
17105               (also called Dcache, Scache, and Bcache), as well as to main
17106               memory.  Note that L3 is only valid for EV5.
17107
17108       FR30 Options
17109
17110       These options are defined specifically for the FR30 port.
17111
17112       -msmall-model
17113           Use the small address space model.  This can produce smaller code,
17114           but it does assume that all symbolic values and addresses fit into
17115           a 20-bit range.
17116
17117       -mno-lsim
17118           Assume that runtime support has been provided and so there is no
17119           need to include the simulator library (libsim.a) on the linker
17120           command line.
17121
17122       FT32 Options
17123
17124       These options are defined specifically for the FT32 port.
17125
17126       -msim
17127           Specifies that the program will be run on the simulator.  This
17128           causes an alternate runtime startup and library to be linked.  You
17129           must not use this option when generating programs that will run on
17130           real hardware; you must provide your own runtime library for
17131           whatever I/O functions are needed.
17132
17133       -mlra
17134           Enable Local Register Allocation.  This is still experimental for
17135           FT32, so by default the compiler uses standard reload.
17136
17137       -mnodiv
17138           Do not use div and mod instructions.
17139
17140       -mft32b
17141           Enable use of the extended instructions of the FT32B processor.
17142
17143       -mcompress
17144           Compress all code using the Ft32B code compression scheme.
17145
17146       -mnopm
17147           Do not generate code that reads program memory.
17148
17149       FRV Options
17150
17151       -mgpr-32
17152           Only use the first 32 general-purpose registers.
17153
17154       -mgpr-64
17155           Use all 64 general-purpose registers.
17156
17157       -mfpr-32
17158           Use only the first 32 floating-point registers.
17159
17160       -mfpr-64
17161           Use all 64 floating-point registers.
17162
17163       -mhard-float
17164           Use hardware instructions for floating-point operations.
17165
17166       -msoft-float
17167           Use library routines for floating-point operations.
17168
17169       -malloc-cc
17170           Dynamically allocate condition code registers.
17171
17172       -mfixed-cc
17173           Do not try to dynamically allocate condition code registers, only
17174           use "icc0" and "fcc0".
17175
17176       -mdword
17177           Change ABI to use double word insns.
17178
17179       -mno-dword
17180           Do not use double word instructions.
17181
17182       -mdouble
17183           Use floating-point double instructions.
17184
17185       -mno-double
17186           Do not use floating-point double instructions.
17187
17188       -mmedia
17189           Use media instructions.
17190
17191       -mno-media
17192           Do not use media instructions.
17193
17194       -mmuladd
17195           Use multiply and add/subtract instructions.
17196
17197       -mno-muladd
17198           Do not use multiply and add/subtract instructions.
17199
17200       -mfdpic
17201           Select the FDPIC ABI, which uses function descriptors to represent
17202           pointers to functions.  Without any PIC/PIE-related options, it
17203           implies -fPIE.  With -fpic or -fpie, it assumes GOT entries and
17204           small data are within a 12-bit range from the GOT base address;
17205           with -fPIC or -fPIE, GOT offsets are computed with 32 bits.  With a
17206           bfin-elf target, this option implies -msim.
17207
17208       -minline-plt
17209           Enable inlining of PLT entries in function calls to functions that
17210           are not known to bind locally.  It has no effect without -mfdpic.
17211           It's enabled by default if optimizing for speed and compiling for
17212           shared libraries (i.e., -fPIC or -fpic), or when an optimization
17213           option such as -O3 or above is present in the command line.
17214
17215       -mTLS
17216           Assume a large TLS segment when generating thread-local code.
17217
17218       -mtls
17219           Do not assume a large TLS segment when generating thread-local
17220           code.
17221
17222       -mgprel-ro
17223           Enable the use of "GPREL" relocations in the FDPIC ABI for data
17224           that is known to be in read-only sections.  It's enabled by
17225           default, except for -fpic or -fpie: even though it may help make
17226           the global offset table smaller, it trades 1 instruction for 4.
17227           With -fPIC or -fPIE, it trades 3 instructions for 4, one of which
17228           may be shared by multiple symbols, and it avoids the need for a GOT
17229           entry for the referenced symbol, so it's more likely to be a win.
17230           If it is not, -mno-gprel-ro can be used to disable it.
17231
17232       -multilib-library-pic
17233           Link with the (library, not FD) pic libraries.  It's implied by
17234           -mlibrary-pic, as well as by -fPIC and -fpic without -mfdpic.  You
17235           should never have to use it explicitly.
17236
17237       -mlinked-fp
17238           Follow the EABI requirement of always creating a frame pointer
17239           whenever a stack frame is allocated.  This option is enabled by
17240           default and can be disabled with -mno-linked-fp.
17241
17242       -mlong-calls
17243           Use indirect addressing to call functions outside the current
17244           compilation unit.  This allows the functions to be placed anywhere
17245           within the 32-bit address space.
17246
17247       -malign-labels
17248           Try to align labels to an 8-byte boundary by inserting NOPs into
17249           the previous packet.  This option only has an effect when VLIW
17250           packing is enabled.  It doesn't create new packets; it merely adds
17251           NOPs to existing ones.
17252
17253       -mlibrary-pic
17254           Generate position-independent EABI code.
17255
17256       -macc-4
17257           Use only the first four media accumulator registers.
17258
17259       -macc-8
17260           Use all eight media accumulator registers.
17261
17262       -mpack
17263           Pack VLIW instructions.
17264
17265       -mno-pack
17266           Do not pack VLIW instructions.
17267
17268       -mno-eflags
17269           Do not mark ABI switches in e_flags.
17270
17271       -mcond-move
17272           Enable the use of conditional-move instructions (default).
17273
17274           This switch is mainly for debugging the compiler and will likely be
17275           removed in a future version.
17276
17277       -mno-cond-move
17278           Disable the use of conditional-move instructions.
17279
17280           This switch is mainly for debugging the compiler and will likely be
17281           removed in a future version.
17282
17283       -mscc
17284           Enable the use of conditional set instructions (default).
17285
17286           This switch is mainly for debugging the compiler and will likely be
17287           removed in a future version.
17288
17289       -mno-scc
17290           Disable the use of conditional set instructions.
17291
17292           This switch is mainly for debugging the compiler and will likely be
17293           removed in a future version.
17294
17295       -mcond-exec
17296           Enable the use of conditional execution (default).
17297
17298           This switch is mainly for debugging the compiler and will likely be
17299           removed in a future version.
17300
17301       -mno-cond-exec
17302           Disable the use of conditional execution.
17303
17304           This switch is mainly for debugging the compiler and will likely be
17305           removed in a future version.
17306
17307       -mvliw-branch
17308           Run a pass to pack branches into VLIW instructions (default).
17309
17310           This switch is mainly for debugging the compiler and will likely be
17311           removed in a future version.
17312
17313       -mno-vliw-branch
17314           Do not run a pass to pack branches into VLIW instructions.
17315
17316           This switch is mainly for debugging the compiler and will likely be
17317           removed in a future version.
17318
17319       -mmulti-cond-exec
17320           Enable optimization of "&&" and "||" in conditional execution
17321           (default).
17322
17323           This switch is mainly for debugging the compiler and will likely be
17324           removed in a future version.
17325
17326       -mno-multi-cond-exec
17327           Disable optimization of "&&" and "||" in conditional execution.
17328
17329           This switch is mainly for debugging the compiler and will likely be
17330           removed in a future version.
17331
17332       -mnested-cond-exec
17333           Enable nested conditional execution optimizations (default).
17334
17335           This switch is mainly for debugging the compiler and will likely be
17336           removed in a future version.
17337
17338       -mno-nested-cond-exec
17339           Disable nested conditional execution optimizations.
17340
17341           This switch is mainly for debugging the compiler and will likely be
17342           removed in a future version.
17343
17344       -moptimize-membar
17345           This switch removes redundant "membar" instructions from the
17346           compiler-generated code.  It is enabled by default.
17347
17348       -mno-optimize-membar
17349           This switch disables the automatic removal of redundant "membar"
17350           instructions from the generated code.
17351
17352       -mtomcat-stats
17353           Cause gas to print out tomcat statistics.
17354
17355       -mcpu=cpu
17356           Select the processor type for which to generate code.  Possible
17357           values are frv, fr550, tomcat, fr500, fr450, fr405, fr400, fr300
17358           and simple.
17359
17360       GNU/Linux Options
17361
17362       These -m options are defined for GNU/Linux targets:
17363
17364       -mglibc
17365           Use the GNU C library.  This is the default except on
17366           *-*-linux-*uclibc*, *-*-linux-*musl* and *-*-linux-*android*
17367           targets.
17368
17369       -muclibc
17370           Use uClibc C library.  This is the default on *-*-linux-*uclibc*
17371           targets.
17372
17373       -mmusl
17374           Use the musl C library.  This is the default on *-*-linux-*musl*
17375           targets.
17376
17377       -mbionic
17378           Use Bionic C library.  This is the default on *-*-linux-*android*
17379           targets.
17380
17381       -mandroid
17382           Compile code compatible with Android platform.  This is the default
17383           on *-*-linux-*android* targets.
17384
17385           When compiling, this option enables -mbionic, -fPIC,
17386           -fno-exceptions and -fno-rtti by default.  When linking, this
17387           option makes the GCC driver pass Android-specific options to the
17388           linker.  Finally, this option causes the preprocessor macro
17389           "__ANDROID__" to be defined.
17390
17391       -tno-android-cc
17392           Disable compilation effects of -mandroid, i.e., do not enable
17393           -mbionic, -fPIC, -fno-exceptions and -fno-rtti by default.
17394
17395       -tno-android-ld
17396           Disable linking effects of -mandroid, i.e., pass standard Linux
17397           linking options to the linker.
17398
17399       H8/300 Options
17400
17401       These -m options are defined for the H8/300 implementations:
17402
17403       -mrelax
17404           Shorten some address references at link time, when possible; uses
17405           the linker option -relax.
17406
17407       -mh Generate code for the H8/300H.
17408
17409       -ms Generate code for the H8S.
17410
17411       -mn Generate code for the H8S and H8/300H in the normal mode.  This
17412           switch must be used either with -mh or -ms.
17413
17414       -ms2600
17415           Generate code for the H8S/2600.  This switch must be used with -ms.
17416
17417       -mexr
17418           Extended registers are stored on stack before execution of function
17419           with monitor attribute. Default option is -mexr.  This option is
17420           valid only for H8S targets.
17421
17422       -mno-exr
17423           Extended registers are not stored on stack before execution of
17424           function with monitor attribute. Default option is -mno-exr.  This
17425           option is valid only for H8S targets.
17426
17427       -mint32
17428           Make "int" data 32 bits by default.
17429
17430       -malign-300
17431           On the H8/300H and H8S, use the same alignment rules as for the
17432           H8/300.  The default for the H8/300H and H8S is to align longs and
17433           floats on 4-byte boundaries.  -malign-300 causes them to be aligned
17434           on 2-byte boundaries.  This option has no effect on the H8/300.
17435
17436       HPPA Options
17437
17438       These -m options are defined for the HPPA family of computers:
17439
17440       -march=architecture-type
17441           Generate code for the specified architecture.  The choices for
17442           architecture-type are 1.0 for PA 1.0, 1.1 for PA 1.1, and 2.0 for
17443           PA 2.0 processors.  Refer to /usr/lib/sched.models on an HP-UX
17444           system to determine the proper architecture option for your
17445           machine.  Code compiled for lower numbered architectures runs on
17446           higher numbered architectures, but not the other way around.
17447
17448       -mpa-risc-1-0
17449       -mpa-risc-1-1
17450       -mpa-risc-2-0
17451           Synonyms for -march=1.0, -march=1.1, and -march=2.0 respectively.
17452
17453       -mcaller-copies
17454           The caller copies function arguments passed by hidden reference.
17455           This option should be used with care as it is not compatible with
17456           the default 32-bit runtime.  However, only aggregates larger than
17457           eight bytes are passed by hidden reference and the option provides
17458           better compatibility with OpenMP.
17459
17460       -mjump-in-delay
17461           This option is ignored and provided for compatibility purposes
17462           only.
17463
17464       -mdisable-fpregs
17465           Prevent floating-point registers from being used in any manner.
17466           This is necessary for compiling kernels that perform lazy context
17467           switching of floating-point registers.  If you use this option and
17468           attempt to perform floating-point operations, the compiler aborts.
17469
17470       -mdisable-indexing
17471           Prevent the compiler from using indexing address modes.  This
17472           avoids some rather obscure problems when compiling MIG generated
17473           code under MACH.
17474
17475       -mno-space-regs
17476           Generate code that assumes the target has no space registers.  This
17477           allows GCC to generate faster indirect calls and use unscaled index
17478           address modes.
17479
17480           Such code is suitable for level 0 PA systems and kernels.
17481
17482       -mfast-indirect-calls
17483           Generate code that assumes calls never cross space boundaries.
17484           This allows GCC to emit code that performs faster indirect calls.
17485
17486           This option does not work in the presence of shared libraries or
17487           nested functions.
17488
17489       -mfixed-range=register-range
17490           Generate code treating the given register range as fixed registers.
17491           A fixed register is one that the register allocator cannot use.
17492           This is useful when compiling kernel code.  A register range is
17493           specified as two registers separated by a dash.  Multiple register
17494           ranges can be specified separated by a comma.
17495
17496       -mlong-load-store
17497           Generate 3-instruction load and store sequences as sometimes
17498           required by the HP-UX 10 linker.  This is equivalent to the +k
17499           option to the HP compilers.
17500
17501       -mportable-runtime
17502           Use the portable calling conventions proposed by HP for ELF
17503           systems.
17504
17505       -mgas
17506           Enable the use of assembler directives only GAS understands.
17507
17508       -mschedule=cpu-type
17509           Schedule code according to the constraints for the machine type
17510           cpu-type.  The choices for cpu-type are 700 7100, 7100LC, 7200,
17511           7300 and 8000.  Refer to /usr/lib/sched.models on an HP-UX system
17512           to determine the proper scheduling option for your machine.  The
17513           default scheduling is 8000.
17514
17515       -mlinker-opt
17516           Enable the optimization pass in the HP-UX linker.  Note this makes
17517           symbolic debugging impossible.  It also triggers a bug in the HP-UX
17518           8 and HP-UX 9 linkers in which they give bogus error messages when
17519           linking some programs.
17520
17521       -msoft-float
17522           Generate output containing library calls for floating point.
17523           Warning: the requisite libraries are not available for all HPPA
17524           targets.  Normally the facilities of the machine's usual C compiler
17525           are used, but this cannot be done directly in cross-compilation.
17526           You must make your own arrangements to provide suitable library
17527           functions for cross-compilation.
17528
17529           -msoft-float changes the calling convention in the output file;
17530           therefore, it is only useful if you compile all of a program with
17531           this option.  In particular, you need to compile libgcc.a, the
17532           library that comes with GCC, with -msoft-float in order for this to
17533           work.
17534
17535       -msio
17536           Generate the predefine, "_SIO", for server IO.  The default is
17537           -mwsio.  This generates the predefines, "__hp9000s700",
17538           "__hp9000s700__" and "_WSIO", for workstation IO.  These options
17539           are available under HP-UX and HI-UX.
17540
17541       -mgnu-ld
17542           Use options specific to GNU ld.  This passes -shared to ld when
17543           building a shared library.  It is the default when GCC is
17544           configured, explicitly or implicitly, with the GNU linker.  This
17545           option does not affect which ld is called; it only changes what
17546           parameters are passed to that ld.  The ld that is called is
17547           determined by the --with-ld configure option, GCC's program search
17548           path, and finally by the user's PATH.  The linker used by GCC can
17549           be printed using which `gcc -print-prog-name=ld`.  This option is
17550           only available on the 64-bit HP-UX GCC, i.e. configured with
17551           hppa*64*-*-hpux*.
17552
17553       -mhp-ld
17554           Use options specific to HP ld.  This passes -b to ld when building
17555           a shared library and passes +Accept TypeMismatch to ld on all
17556           links.  It is the default when GCC is configured, explicitly or
17557           implicitly, with the HP linker.  This option does not affect which
17558           ld is called; it only changes what parameters are passed to that
17559           ld.  The ld that is called is determined by the --with-ld configure
17560           option, GCC's program search path, and finally by the user's PATH.
17561           The linker used by GCC can be printed using which `gcc
17562           -print-prog-name=ld`.  This option is only available on the 64-bit
17563           HP-UX GCC, i.e. configured with hppa*64*-*-hpux*.
17564
17565       -mlong-calls
17566           Generate code that uses long call sequences.  This ensures that a
17567           call is always able to reach linker generated stubs.  The default
17568           is to generate long calls only when the distance from the call site
17569           to the beginning of the function or translation unit, as the case
17570           may be, exceeds a predefined limit set by the branch type being
17571           used.  The limits for normal calls are 7,600,000 and 240,000 bytes,
17572           respectively for the PA 2.0 and PA 1.X architectures.  Sibcalls are
17573           always limited at 240,000 bytes.
17574
17575           Distances are measured from the beginning of functions when using
17576           the -ffunction-sections option, or when using the -mgas and
17577           -mno-portable-runtime options together under HP-UX with the SOM
17578           linker.
17579
17580           It is normally not desirable to use this option as it degrades
17581           performance.  However, it may be useful in large applications,
17582           particularly when partial linking is used to build the application.
17583
17584           The types of long calls used depends on the capabilities of the
17585           assembler and linker, and the type of code being generated.  The
17586           impact on systems that support long absolute calls, and long pic
17587           symbol-difference or pc-relative calls should be relatively small.
17588           However, an indirect call is used on 32-bit ELF systems in pic code
17589           and it is quite long.
17590
17591       -munix=unix-std
17592           Generate compiler predefines and select a startfile for the
17593           specified UNIX standard.  The choices for unix-std are 93, 95 and
17594           98.  93 is supported on all HP-UX versions.  95 is available on HP-
17595           UX 10.10 and later.  98 is available on HP-UX 11.11 and later.  The
17596           default values are 93 for HP-UX 10.00, 95 for HP-UX 10.10 though to
17597           11.00, and 98 for HP-UX 11.11 and later.
17598
17599           -munix=93 provides the same predefines as GCC 3.3 and 3.4.
17600           -munix=95 provides additional predefines for "XOPEN_UNIX" and
17601           "_XOPEN_SOURCE_EXTENDED", and the startfile unix95.o.  -munix=98
17602           provides additional predefines for "_XOPEN_UNIX",
17603           "_XOPEN_SOURCE_EXTENDED", "_INCLUDE__STDC_A1_SOURCE" and
17604           "_INCLUDE_XOPEN_SOURCE_500", and the startfile unix98.o.
17605
17606           It is important to note that this option changes the interfaces for
17607           various library routines.  It also affects the operational behavior
17608           of the C library.  Thus, extreme care is needed in using this
17609           option.
17610
17611           Library code that is intended to operate with more than one UNIX
17612           standard must test, set and restore the variable
17613           "__xpg4_extended_mask" as appropriate.  Most GNU software doesn't
17614           provide this capability.
17615
17616       -nolibdld
17617           Suppress the generation of link options to search libdld.sl when
17618           the -static option is specified on HP-UX 10 and later.
17619
17620       -static
17621           The HP-UX implementation of setlocale in libc has a dependency on
17622           libdld.sl.  There isn't an archive version of libdld.sl.  Thus,
17623           when the -static option is specified, special link options are
17624           needed to resolve this dependency.
17625
17626           On HP-UX 10 and later, the GCC driver adds the necessary options to
17627           link with libdld.sl when the -static option is specified.  This
17628           causes the resulting binary to be dynamic.  On the 64-bit port, the
17629           linkers generate dynamic binaries by default in any case.  The
17630           -nolibdld option can be used to prevent the GCC driver from adding
17631           these link options.
17632
17633       -threads
17634           Add support for multithreading with the dce thread library under
17635           HP-UX.  This option sets flags for both the preprocessor and
17636           linker.
17637
17638       IA-64 Options
17639
17640       These are the -m options defined for the Intel IA-64 architecture.
17641
17642       -mbig-endian
17643           Generate code for a big-endian target.  This is the default for HP-
17644           UX.
17645
17646       -mlittle-endian
17647           Generate code for a little-endian target.  This is the default for
17648           AIX5 and GNU/Linux.
17649
17650       -mgnu-as
17651       -mno-gnu-as
17652           Generate (or don't) code for the GNU assembler.  This is the
17653           default.
17654
17655       -mgnu-ld
17656       -mno-gnu-ld
17657           Generate (or don't) code for the GNU linker.  This is the default.
17658
17659       -mno-pic
17660           Generate code that does not use a global pointer register.  The
17661           result is not position independent code, and violates the IA-64
17662           ABI.
17663
17664       -mvolatile-asm-stop
17665       -mno-volatile-asm-stop
17666           Generate (or don't) a stop bit immediately before and after
17667           volatile asm statements.
17668
17669       -mregister-names
17670       -mno-register-names
17671           Generate (or don't) in, loc, and out register names for the stacked
17672           registers.  This may make assembler output more readable.
17673
17674       -mno-sdata
17675       -msdata
17676           Disable (or enable) optimizations that use the small data section.
17677           This may be useful for working around optimizer bugs.
17678
17679       -mconstant-gp
17680           Generate code that uses a single constant global pointer value.
17681           This is useful when compiling kernel code.
17682
17683       -mauto-pic
17684           Generate code that is self-relocatable.  This implies
17685           -mconstant-gp.  This is useful when compiling firmware code.
17686
17687       -minline-float-divide-min-latency
17688           Generate code for inline divides of floating-point values using the
17689           minimum latency algorithm.
17690
17691       -minline-float-divide-max-throughput
17692           Generate code for inline divides of floating-point values using the
17693           maximum throughput algorithm.
17694
17695       -mno-inline-float-divide
17696           Do not generate inline code for divides of floating-point values.
17697
17698       -minline-int-divide-min-latency
17699           Generate code for inline divides of integer values using the
17700           minimum latency algorithm.
17701
17702       -minline-int-divide-max-throughput
17703           Generate code for inline divides of integer values using the
17704           maximum throughput algorithm.
17705
17706       -mno-inline-int-divide
17707           Do not generate inline code for divides of integer values.
17708
17709       -minline-sqrt-min-latency
17710           Generate code for inline square roots using the minimum latency
17711           algorithm.
17712
17713       -minline-sqrt-max-throughput
17714           Generate code for inline square roots using the maximum throughput
17715           algorithm.
17716
17717       -mno-inline-sqrt
17718           Do not generate inline code for "sqrt".
17719
17720       -mfused-madd
17721       -mno-fused-madd
17722           Do (don't) generate code that uses the fused multiply/add or
17723           multiply/subtract instructions.  The default is to use these
17724           instructions.
17725
17726       -mno-dwarf2-asm
17727       -mdwarf2-asm
17728           Don't (or do) generate assembler code for the DWARF line number
17729           debugging info.  This may be useful when not using the GNU
17730           assembler.
17731
17732       -mearly-stop-bits
17733       -mno-early-stop-bits
17734           Allow stop bits to be placed earlier than immediately preceding the
17735           instruction that triggered the stop bit.  This can improve
17736           instruction scheduling, but does not always do so.
17737
17738       -mfixed-range=register-range
17739           Generate code treating the given register range as fixed registers.
17740           A fixed register is one that the register allocator cannot use.
17741           This is useful when compiling kernel code.  A register range is
17742           specified as two registers separated by a dash.  Multiple register
17743           ranges can be specified separated by a comma.
17744
17745       -mtls-size=tls-size
17746           Specify bit size of immediate TLS offsets.  Valid values are 14,
17747           22, and 64.
17748
17749       -mtune=cpu-type
17750           Tune the instruction scheduling for a particular CPU, Valid values
17751           are itanium, itanium1, merced, itanium2, and mckinley.
17752
17753       -milp32
17754       -mlp64
17755           Generate code for a 32-bit or 64-bit environment.  The 32-bit
17756           environment sets int, long and pointer to 32 bits.  The 64-bit
17757           environment sets int to 32 bits and long and pointer to 64 bits.
17758           These are HP-UX specific flags.
17759
17760       -mno-sched-br-data-spec
17761       -msched-br-data-spec
17762           (Dis/En)able data speculative scheduling before reload.  This
17763           results in generation of "ld.a" instructions and the corresponding
17764           check instructions ("ld.c" / "chk.a").  The default setting is
17765           disabled.
17766
17767       -msched-ar-data-spec
17768       -mno-sched-ar-data-spec
17769           (En/Dis)able data speculative scheduling after reload.  This
17770           results in generation of "ld.a" instructions and the corresponding
17771           check instructions ("ld.c" / "chk.a").  The default setting is
17772           enabled.
17773
17774       -mno-sched-control-spec
17775       -msched-control-spec
17776           (Dis/En)able control speculative scheduling.  This feature is
17777           available only during region scheduling (i.e. before reload).  This
17778           results in generation of the "ld.s" instructions and the
17779           corresponding check instructions "chk.s".  The default setting is
17780           disabled.
17781
17782       -msched-br-in-data-spec
17783       -mno-sched-br-in-data-spec
17784           (En/Dis)able speculative scheduling of the instructions that are
17785           dependent on the data speculative loads before reload.  This is
17786           effective only with -msched-br-data-spec enabled.  The default
17787           setting is enabled.
17788
17789       -msched-ar-in-data-spec
17790       -mno-sched-ar-in-data-spec
17791           (En/Dis)able speculative scheduling of the instructions that are
17792           dependent on the data speculative loads after reload.  This is
17793           effective only with -msched-ar-data-spec enabled.  The default
17794           setting is enabled.
17795
17796       -msched-in-control-spec
17797       -mno-sched-in-control-spec
17798           (En/Dis)able speculative scheduling of the instructions that are
17799           dependent on the control speculative loads.  This is effective only
17800           with -msched-control-spec enabled.  The default setting is enabled.
17801
17802       -mno-sched-prefer-non-data-spec-insns
17803       -msched-prefer-non-data-spec-insns
17804           If enabled, data-speculative instructions are chosen for schedule
17805           only if there are no other choices at the moment.  This makes the
17806           use of the data speculation much more conservative.  The default
17807           setting is disabled.
17808
17809       -mno-sched-prefer-non-control-spec-insns
17810       -msched-prefer-non-control-spec-insns
17811           If enabled, control-speculative instructions are chosen for
17812           schedule only if there are no other choices at the moment.  This
17813           makes the use of the control speculation much more conservative.
17814           The default setting is disabled.
17815
17816       -mno-sched-count-spec-in-critical-path
17817       -msched-count-spec-in-critical-path
17818           If enabled, speculative dependencies are considered during
17819           computation of the instructions priorities.  This makes the use of
17820           the speculation a bit more conservative.  The default setting is
17821           disabled.
17822
17823       -msched-spec-ldc
17824           Use a simple data speculation check.  This option is on by default.
17825
17826       -msched-control-spec-ldc
17827           Use a simple check for control speculation.  This option is on by
17828           default.
17829
17830       -msched-stop-bits-after-every-cycle
17831           Place a stop bit after every cycle when scheduling.  This option is
17832           on by default.
17833
17834       -msched-fp-mem-deps-zero-cost
17835           Assume that floating-point stores and loads are not likely to cause
17836           a conflict when placed into the same instruction group.  This
17837           option is disabled by default.
17838
17839       -msel-sched-dont-check-control-spec
17840           Generate checks for control speculation in selective scheduling.
17841           This flag is disabled by default.
17842
17843       -msched-max-memory-insns=max-insns
17844           Limit on the number of memory insns per instruction group, giving
17845           lower priority to subsequent memory insns attempting to schedule in
17846           the same instruction group. Frequently useful to prevent cache bank
17847           conflicts.  The default value is 1.
17848
17849       -msched-max-memory-insns-hard-limit
17850           Makes the limit specified by msched-max-memory-insns a hard limit,
17851           disallowing more than that number in an instruction group.
17852           Otherwise, the limit is "soft", meaning that non-memory operations
17853           are preferred when the limit is reached, but memory operations may
17854           still be scheduled.
17855
17856       LM32 Options
17857
17858       These -m options are defined for the LatticeMico32 architecture:
17859
17860       -mbarrel-shift-enabled
17861           Enable barrel-shift instructions.
17862
17863       -mdivide-enabled
17864           Enable divide and modulus instructions.
17865
17866       -mmultiply-enabled
17867           Enable multiply instructions.
17868
17869       -msign-extend-enabled
17870           Enable sign extend instructions.
17871
17872       -muser-enabled
17873           Enable user-defined instructions.
17874
17875       M32C Options
17876
17877       -mcpu=name
17878           Select the CPU for which code is generated.  name may be one of r8c
17879           for the R8C/Tiny series, m16c for the M16C (up to /60) series,
17880           m32cm for the M16C/80 series, or m32c for the M32C/80 series.
17881
17882       -msim
17883           Specifies that the program will be run on the simulator.  This
17884           causes an alternate runtime library to be linked in which supports,
17885           for example, file I/O.  You must not use this option when
17886           generating programs that will run on real hardware; you must
17887           provide your own runtime library for whatever I/O functions are
17888           needed.
17889
17890       -memregs=number
17891           Specifies the number of memory-based pseudo-registers GCC uses
17892           during code generation.  These pseudo-registers are used like real
17893           registers, so there is a tradeoff between GCC's ability to fit the
17894           code into available registers, and the performance penalty of using
17895           memory instead of registers.  Note that all modules in a program
17896           must be compiled with the same value for this option.  Because of
17897           that, you must not use this option with GCC's default runtime
17898           libraries.
17899
17900       M32R/D Options
17901
17902       These -m options are defined for Renesas M32R/D architectures:
17903
17904       -m32r2
17905           Generate code for the M32R/2.
17906
17907       -m32rx
17908           Generate code for the M32R/X.
17909
17910       -m32r
17911           Generate code for the M32R.  This is the default.
17912
17913       -mmodel=small
17914           Assume all objects live in the lower 16MB of memory (so that their
17915           addresses can be loaded with the "ld24" instruction), and assume
17916           all subroutines are reachable with the "bl" instruction.  This is
17917           the default.
17918
17919           The addressability of a particular object can be set with the
17920           "model" attribute.
17921
17922       -mmodel=medium
17923           Assume objects may be anywhere in the 32-bit address space (the
17924           compiler generates "seth/add3" instructions to load their
17925           addresses), and assume all subroutines are reachable with the "bl"
17926           instruction.
17927
17928       -mmodel=large
17929           Assume objects may be anywhere in the 32-bit address space (the
17930           compiler generates "seth/add3" instructions to load their
17931           addresses), and assume subroutines may not be reachable with the
17932           "bl" instruction (the compiler generates the much slower
17933           "seth/add3/jl" instruction sequence).
17934
17935       -msdata=none
17936           Disable use of the small data area.  Variables are put into one of
17937           ".data", ".bss", or ".rodata" (unless the "section" attribute has
17938           been specified).  This is the default.
17939
17940           The small data area consists of sections ".sdata" and ".sbss".
17941           Objects may be explicitly put in the small data area with the
17942           "section" attribute using one of these sections.
17943
17944       -msdata=sdata
17945           Put small global and static data in the small data area, but do not
17946           generate special code to reference them.
17947
17948       -msdata=use
17949           Put small global and static data in the small data area, and
17950           generate special instructions to reference them.
17951
17952       -G num
17953           Put global and static objects less than or equal to num bytes into
17954           the small data or BSS sections instead of the normal data or BSS
17955           sections.  The default value of num is 8.  The -msdata option must
17956           be set to one of sdata or use for this option to have any effect.
17957
17958           All modules should be compiled with the same -G num value.
17959           Compiling with different values of num may or may not work; if it
17960           doesn't the linker gives an error message---incorrect code is not
17961           generated.
17962
17963       -mdebug
17964           Makes the M32R-specific code in the compiler display some
17965           statistics that might help in debugging programs.
17966
17967       -malign-loops
17968           Align all loops to a 32-byte boundary.
17969
17970       -mno-align-loops
17971           Do not enforce a 32-byte alignment for loops.  This is the default.
17972
17973       -missue-rate=number
17974           Issue number instructions per cycle.  number can only be 1 or 2.
17975
17976       -mbranch-cost=number
17977           number can only be 1 or 2.  If it is 1 then branches are preferred
17978           over conditional code, if it is 2, then the opposite applies.
17979
17980       -mflush-trap=number
17981           Specifies the trap number to use to flush the cache.  The default
17982           is 12.  Valid numbers are between 0 and 15 inclusive.
17983
17984       -mno-flush-trap
17985           Specifies that the cache cannot be flushed by using a trap.
17986
17987       -mflush-func=name
17988           Specifies the name of the operating system function to call to
17989           flush the cache.  The default is _flush_cache, but a function call
17990           is only used if a trap is not available.
17991
17992       -mno-flush-func
17993           Indicates that there is no OS function for flushing the cache.
17994
17995       M680x0 Options
17996
17997       These are the -m options defined for M680x0 and ColdFire processors.
17998       The default settings depend on which architecture was selected when the
17999       compiler was configured; the defaults for the most common choices are
18000       given below.
18001
18002       -march=arch
18003           Generate code for a specific M680x0 or ColdFire instruction set
18004           architecture.  Permissible values of arch for M680x0 architectures
18005           are: 68000, 68010, 68020, 68030, 68040, 68060 and cpu32.  ColdFire
18006           architectures are selected according to Freescale's ISA
18007           classification and the permissible values are: isaa, isaaplus, isab
18008           and isac.
18009
18010           GCC defines a macro "__mcfarch__" whenever it is generating code
18011           for a ColdFire target.  The arch in this macro is one of the -march
18012           arguments given above.
18013
18014           When used together, -march and -mtune select code that runs on a
18015           family of similar processors but that is optimized for a particular
18016           microarchitecture.
18017
18018       -mcpu=cpu
18019           Generate code for a specific M680x0 or ColdFire processor.  The
18020           M680x0 cpus are: 68000, 68010, 68020, 68030, 68040, 68060, 68302,
18021           68332 and cpu32.  The ColdFire cpus are given by the table below,
18022           which also classifies the CPUs into families:
18023
18024           Family : -mcpu arguments
18025           51 : 51 51ac 51ag 51cn 51em 51je 51jf 51jg 51jm 51mm 51qe 51qm
18026           5206 : 5202 5204 5206
18027           5206e : 5206e
18028           5208 : 5207 5208
18029           5211a : 5210a 5211a
18030           5213 : 5211 5212 5213
18031           5216 : 5214 5216
18032           52235 : 52230 52231 52232 52233 52234 52235
18033           5225 : 5224 5225
18034           52259 : 52252 52254 52255 52256 52258 52259
18035           5235 : 5232 5233 5234 5235 523x
18036           5249 : 5249
18037           5250 : 5250
18038           5271 : 5270 5271
18039           5272 : 5272
18040           5275 : 5274 5275
18041           5282 : 5280 5281 5282 528x
18042           53017 : 53011 53012 53013 53014 53015 53016 53017
18043           5307 : 5307
18044           5329 : 5327 5328 5329 532x
18045           5373 : 5372 5373 537x
18046           5407 : 5407
18047           5475 : 5470 5471 5472 5473 5474 5475 547x 5480 5481 5482 5483 5484
18048           5485
18049
18050           -mcpu=cpu overrides -march=arch if arch is compatible with cpu.
18051           Other combinations of -mcpu and -march are rejected.
18052
18053           GCC defines the macro "__mcf_cpu_cpu" when ColdFire target cpu is
18054           selected.  It also defines "__mcf_family_family", where the value
18055           of family is given by the table above.
18056
18057       -mtune=tune
18058           Tune the code for a particular microarchitecture within the
18059           constraints set by -march and -mcpu.  The M680x0 microarchitectures
18060           are: 68000, 68010, 68020, 68030, 68040, 68060 and cpu32.  The
18061           ColdFire microarchitectures are: cfv1, cfv2, cfv3, cfv4 and cfv4e.
18062
18063           You can also use -mtune=68020-40 for code that needs to run
18064           relatively well on 68020, 68030 and 68040 targets.  -mtune=68020-60
18065           is similar but includes 68060 targets as well.  These two options
18066           select the same tuning decisions as -m68020-40 and -m68020-60
18067           respectively.
18068
18069           GCC defines the macros "__mcarch" and "__mcarch__" when tuning for
18070           680x0 architecture arch.  It also defines "mcarch" unless either
18071           -ansi or a non-GNU -std option is used.  If GCC is tuning for a
18072           range of architectures, as selected by -mtune=68020-40 or
18073           -mtune=68020-60, it defines the macros for every architecture in
18074           the range.
18075
18076           GCC also defines the macro "__muarch__" when tuning for ColdFire
18077           microarchitecture uarch, where uarch is one of the arguments given
18078           above.
18079
18080       -m68000
18081       -mc68000
18082           Generate output for a 68000.  This is the default when the compiler
18083           is configured for 68000-based systems.  It is equivalent to
18084           -march=68000.
18085
18086           Use this option for microcontrollers with a 68000 or EC000 core,
18087           including the 68008, 68302, 68306, 68307, 68322, 68328 and 68356.
18088
18089       -m68010
18090           Generate output for a 68010.  This is the default when the compiler
18091           is configured for 68010-based systems.  It is equivalent to
18092           -march=68010.
18093
18094       -m68020
18095       -mc68020
18096           Generate output for a 68020.  This is the default when the compiler
18097           is configured for 68020-based systems.  It is equivalent to
18098           -march=68020.
18099
18100       -m68030
18101           Generate output for a 68030.  This is the default when the compiler
18102           is configured for 68030-based systems.  It is equivalent to
18103           -march=68030.
18104
18105       -m68040
18106           Generate output for a 68040.  This is the default when the compiler
18107           is configured for 68040-based systems.  It is equivalent to
18108           -march=68040.
18109
18110           This option inhibits the use of 68881/68882 instructions that have
18111           to be emulated by software on the 68040.  Use this option if your
18112           68040 does not have code to emulate those instructions.
18113
18114       -m68060
18115           Generate output for a 68060.  This is the default when the compiler
18116           is configured for 68060-based systems.  It is equivalent to
18117           -march=68060.
18118
18119           This option inhibits the use of 68020 and 68881/68882 instructions
18120           that have to be emulated by software on the 68060.  Use this option
18121           if your 68060 does not have code to emulate those instructions.
18122
18123       -mcpu32
18124           Generate output for a CPU32.  This is the default when the compiler
18125           is configured for CPU32-based systems.  It is equivalent to
18126           -march=cpu32.
18127
18128           Use this option for microcontrollers with a CPU32 or CPU32+ core,
18129           including the 68330, 68331, 68332, 68333, 68334, 68336, 68340,
18130           68341, 68349 and 68360.
18131
18132       -m5200
18133           Generate output for a 520X ColdFire CPU.  This is the default when
18134           the compiler is configured for 520X-based systems.  It is
18135           equivalent to -mcpu=5206, and is now deprecated in favor of that
18136           option.
18137
18138           Use this option for microcontroller with a 5200 core, including the
18139           MCF5202, MCF5203, MCF5204 and MCF5206.
18140
18141       -m5206e
18142           Generate output for a 5206e ColdFire CPU.  The option is now
18143           deprecated in favor of the equivalent -mcpu=5206e.
18144
18145       -m528x
18146           Generate output for a member of the ColdFire 528X family.  The
18147           option is now deprecated in favor of the equivalent -mcpu=528x.
18148
18149       -m5307
18150           Generate output for a ColdFire 5307 CPU.  The option is now
18151           deprecated in favor of the equivalent -mcpu=5307.
18152
18153       -m5407
18154           Generate output for a ColdFire 5407 CPU.  The option is now
18155           deprecated in favor of the equivalent -mcpu=5407.
18156
18157       -mcfv4e
18158           Generate output for a ColdFire V4e family CPU (e.g. 547x/548x).
18159           This includes use of hardware floating-point instructions.  The
18160           option is equivalent to -mcpu=547x, and is now deprecated in favor
18161           of that option.
18162
18163       -m68020-40
18164           Generate output for a 68040, without using any of the new
18165           instructions.  This results in code that can run relatively
18166           efficiently on either a 68020/68881 or a 68030 or a 68040.  The
18167           generated code does use the 68881 instructions that are emulated on
18168           the 68040.
18169
18170           The option is equivalent to -march=68020 -mtune=68020-40.
18171
18172       -m68020-60
18173           Generate output for a 68060, without using any of the new
18174           instructions.  This results in code that can run relatively
18175           efficiently on either a 68020/68881 or a 68030 or a 68040.  The
18176           generated code does use the 68881 instructions that are emulated on
18177           the 68060.
18178
18179           The option is equivalent to -march=68020 -mtune=68020-60.
18180
18181       -mhard-float
18182       -m68881
18183           Generate floating-point instructions.  This is the default for
18184           68020 and above, and for ColdFire devices that have an FPU.  It
18185           defines the macro "__HAVE_68881__" on M680x0 targets and
18186           "__mcffpu__" on ColdFire targets.
18187
18188       -msoft-float
18189           Do not generate floating-point instructions; use library calls
18190           instead.  This is the default for 68000, 68010, and 68832 targets.
18191           It is also the default for ColdFire devices that have no FPU.
18192
18193       -mdiv
18194       -mno-div
18195           Generate (do not generate) ColdFire hardware divide and remainder
18196           instructions.  If -march is used without -mcpu, the default is "on"
18197           for ColdFire architectures and "off" for M680x0 architectures.
18198           Otherwise, the default is taken from the target CPU (either the
18199           default CPU, or the one specified by -mcpu).  For example, the
18200           default is "off" for -mcpu=5206 and "on" for -mcpu=5206e.
18201
18202           GCC defines the macro "__mcfhwdiv__" when this option is enabled.
18203
18204       -mshort
18205           Consider type "int" to be 16 bits wide, like "short int".
18206           Additionally, parameters passed on the stack are also aligned to a
18207           16-bit boundary even on targets whose API mandates promotion to
18208           32-bit.
18209
18210       -mno-short
18211           Do not consider type "int" to be 16 bits wide.  This is the
18212           default.
18213
18214       -mnobitfield
18215       -mno-bitfield
18216           Do not use the bit-field instructions.  The -m68000, -mcpu32 and
18217           -m5200 options imply -mnobitfield.
18218
18219       -mbitfield
18220           Do use the bit-field instructions.  The -m68020 option implies
18221           -mbitfield.  This is the default if you use a configuration
18222           designed for a 68020.
18223
18224       -mrtd
18225           Use a different function-calling convention, in which functions
18226           that take a fixed number of arguments return with the "rtd"
18227           instruction, which pops their arguments while returning.  This
18228           saves one instruction in the caller since there is no need to pop
18229           the arguments there.
18230
18231           This calling convention is incompatible with the one normally used
18232           on Unix, so you cannot use it if you need to call libraries
18233           compiled with the Unix compiler.
18234
18235           Also, you must provide function prototypes for all functions that
18236           take variable numbers of arguments (including "printf"); otherwise
18237           incorrect code is generated for calls to those functions.
18238
18239           In addition, seriously incorrect code results if you call a
18240           function with too many arguments.  (Normally, extra arguments are
18241           harmlessly ignored.)
18242
18243           The "rtd" instruction is supported by the 68010, 68020, 68030,
18244           68040, 68060 and CPU32 processors, but not by the 68000 or 5200.
18245
18246           The default is -mno-rtd.
18247
18248       -malign-int
18249       -mno-align-int
18250           Control whether GCC aligns "int", "long", "long long", "float",
18251           "double", and "long double" variables on a 32-bit boundary
18252           (-malign-int) or a 16-bit boundary (-mno-align-int).  Aligning
18253           variables on 32-bit boundaries produces code that runs somewhat
18254           faster on processors with 32-bit busses at the expense of more
18255           memory.
18256
18257           Warning: if you use the -malign-int switch, GCC aligns structures
18258           containing the above types differently than most published
18259           application binary interface specifications for the m68k.
18260
18261       -mpcrel
18262           Use the pc-relative addressing mode of the 68000 directly, instead
18263           of using a global offset table.  At present, this option implies
18264           -fpic, allowing at most a 16-bit offset for pc-relative addressing.
18265           -fPIC is not presently supported with -mpcrel, though this could be
18266           supported for 68020 and higher processors.
18267
18268       -mno-strict-align
18269       -mstrict-align
18270           Do not (do) assume that unaligned memory references are handled by
18271           the system.
18272
18273       -msep-data
18274           Generate code that allows the data segment to be located in a
18275           different area of memory from the text segment.  This allows for
18276           execute-in-place in an environment without virtual memory
18277           management.  This option implies -fPIC.
18278
18279       -mno-sep-data
18280           Generate code that assumes that the data segment follows the text
18281           segment.  This is the default.
18282
18283       -mid-shared-library
18284           Generate code that supports shared libraries via the library ID
18285           method.  This allows for execute-in-place and shared libraries in
18286           an environment without virtual memory management.  This option
18287           implies -fPIC.
18288
18289       -mno-id-shared-library
18290           Generate code that doesn't assume ID-based shared libraries are
18291           being used.  This is the default.
18292
18293       -mshared-library-id=n
18294           Specifies the identification number of the ID-based shared library
18295           being compiled.  Specifying a value of 0 generates more compact
18296           code; specifying other values forces the allocation of that number
18297           to the current library, but is no more space- or time-efficient
18298           than omitting this option.
18299
18300       -mxgot
18301       -mno-xgot
18302           When generating position-independent code for ColdFire, generate
18303           code that works if the GOT has more than 8192 entries.  This code
18304           is larger and slower than code generated without this option.  On
18305           M680x0 processors, this option is not needed; -fPIC suffices.
18306
18307           GCC normally uses a single instruction to load values from the GOT.
18308           While this is relatively efficient, it only works if the GOT is
18309           smaller than about 64k.  Anything larger causes the linker to
18310           report an error such as:
18311
18312                   relocation truncated to fit: R_68K_GOT16O foobar
18313
18314           If this happens, you should recompile your code with -mxgot.  It
18315           should then work with very large GOTs.  However, code generated
18316           with -mxgot is less efficient, since it takes 4 instructions to
18317           fetch the value of a global symbol.
18318
18319           Note that some linkers, including newer versions of the GNU linker,
18320           can create multiple GOTs and sort GOT entries.  If you have such a
18321           linker, you should only need to use -mxgot when compiling a single
18322           object file that accesses more than 8192 GOT entries.  Very few do.
18323
18324           These options have no effect unless GCC is generating position-
18325           independent code.
18326
18327       -mlong-jump-table-offsets
18328           Use 32-bit offsets in "switch" tables.  The default is to use
18329           16-bit offsets.
18330
18331       MCore Options
18332
18333       These are the -m options defined for the Motorola M*Core processors.
18334
18335       -mhardlit
18336       -mno-hardlit
18337           Inline constants into the code stream if it can be done in two
18338           instructions or less.
18339
18340       -mdiv
18341       -mno-div
18342           Use the divide instruction.  (Enabled by default).
18343
18344       -mrelax-immediate
18345       -mno-relax-immediate
18346           Allow arbitrary-sized immediates in bit operations.
18347
18348       -mwide-bitfields
18349       -mno-wide-bitfields
18350           Always treat bit-fields as "int"-sized.
18351
18352       -m4byte-functions
18353       -mno-4byte-functions
18354           Force all functions to be aligned to a 4-byte boundary.
18355
18356       -mcallgraph-data
18357       -mno-callgraph-data
18358           Emit callgraph information.
18359
18360       -mslow-bytes
18361       -mno-slow-bytes
18362           Prefer word access when reading byte quantities.
18363
18364       -mlittle-endian
18365       -mbig-endian
18366           Generate code for a little-endian target.
18367
18368       -m210
18369       -m340
18370           Generate code for the 210 processor.
18371
18372       -mno-lsim
18373           Assume that runtime support has been provided and so omit the
18374           simulator library (libsim.a) from the linker command line.
18375
18376       -mstack-increment=size
18377           Set the maximum amount for a single stack increment operation.
18378           Large values can increase the speed of programs that contain
18379           functions that need a large amount of stack space, but they can
18380           also trigger a segmentation fault if the stack is extended too
18381           much.  The default value is 0x1000.
18382
18383       MeP Options
18384
18385       -mabsdiff
18386           Enables the "abs" instruction, which is the absolute difference
18387           between two registers.
18388
18389       -mall-opts
18390           Enables all the optional instructions---average, multiply, divide,
18391           bit operations, leading zero, absolute difference, min/max, clip,
18392           and saturation.
18393
18394       -maverage
18395           Enables the "ave" instruction, which computes the average of two
18396           registers.
18397
18398       -mbased=n
18399           Variables of size n bytes or smaller are placed in the ".based"
18400           section by default.  Based variables use the $tp register as a base
18401           register, and there is a 128-byte limit to the ".based" section.
18402
18403       -mbitops
18404           Enables the bit operation instructions---bit test ("btstm"), set
18405           ("bsetm"), clear ("bclrm"), invert ("bnotm"), and test-and-set
18406           ("tas").
18407
18408       -mc=name
18409           Selects which section constant data is placed in.  name may be
18410           tiny, near, or far.
18411
18412       -mclip
18413           Enables the "clip" instruction.  Note that -mclip is not useful
18414           unless you also provide -mminmax.
18415
18416       -mconfig=name
18417           Selects one of the built-in core configurations.  Each MeP chip has
18418           one or more modules in it; each module has a core CPU and a variety
18419           of coprocessors, optional instructions, and peripherals.  The
18420           "MeP-Integrator" tool, not part of GCC, provides these
18421           configurations through this option; using this option is the same
18422           as using all the corresponding command-line options.  The default
18423           configuration is default.
18424
18425       -mcop
18426           Enables the coprocessor instructions.  By default, this is a 32-bit
18427           coprocessor.  Note that the coprocessor is normally enabled via the
18428           -mconfig= option.
18429
18430       -mcop32
18431           Enables the 32-bit coprocessor's instructions.
18432
18433       -mcop64
18434           Enables the 64-bit coprocessor's instructions.
18435
18436       -mivc2
18437           Enables IVC2 scheduling.  IVC2 is a 64-bit VLIW coprocessor.
18438
18439       -mdc
18440           Causes constant variables to be placed in the ".near" section.
18441
18442       -mdiv
18443           Enables the "div" and "divu" instructions.
18444
18445       -meb
18446           Generate big-endian code.
18447
18448       -mel
18449           Generate little-endian code.
18450
18451       -mio-volatile
18452           Tells the compiler that any variable marked with the "io" attribute
18453           is to be considered volatile.
18454
18455       -ml Causes variables to be assigned to the ".far" section by default.
18456
18457       -mleadz
18458           Enables the "leadz" (leading zero) instruction.
18459
18460       -mm Causes variables to be assigned to the ".near" section by default.
18461
18462       -mminmax
18463           Enables the "min" and "max" instructions.
18464
18465       -mmult
18466           Enables the multiplication and multiply-accumulate instructions.
18467
18468       -mno-opts
18469           Disables all the optional instructions enabled by -mall-opts.
18470
18471       -mrepeat
18472           Enables the "repeat" and "erepeat" instructions, used for low-
18473           overhead looping.
18474
18475       -ms Causes all variables to default to the ".tiny" section.  Note that
18476           there is a 65536-byte limit to this section.  Accesses to these
18477           variables use the %gp base register.
18478
18479       -msatur
18480           Enables the saturation instructions.  Note that the compiler does
18481           not currently generate these itself, but this option is included
18482           for compatibility with other tools, like "as".
18483
18484       -msdram
18485           Link the SDRAM-based runtime instead of the default ROM-based
18486           runtime.
18487
18488       -msim
18489           Link the simulator run-time libraries.
18490
18491       -msimnovec
18492           Link the simulator runtime libraries, excluding built-in support
18493           for reset and exception vectors and tables.
18494
18495       -mtf
18496           Causes all functions to default to the ".far" section.  Without
18497           this option, functions default to the ".near" section.
18498
18499       -mtiny=n
18500           Variables that are n bytes or smaller are allocated to the ".tiny"
18501           section.  These variables use the $gp base register.  The default
18502           for this option is 4, but note that there's a 65536-byte limit to
18503           the ".tiny" section.
18504
18505       MicroBlaze Options
18506
18507       -msoft-float
18508           Use software emulation for floating point (default).
18509
18510       -mhard-float
18511           Use hardware floating-point instructions.
18512
18513       -mmemcpy
18514           Do not optimize block moves, use "memcpy".
18515
18516       -mno-clearbss
18517           This option is deprecated.  Use -fno-zero-initialized-in-bss
18518           instead.
18519
18520       -mcpu=cpu-type
18521           Use features of, and schedule code for, the given CPU.  Supported
18522           values are in the format vX.YY.Z, where X is a major version, YY is
18523           the minor version, and Z is compatibility code.  Example values are
18524           v3.00.a, v4.00.b, v5.00.a, v5.00.b, v6.00.a.
18525
18526       -mxl-soft-mul
18527           Use software multiply emulation (default).
18528
18529       -mxl-soft-div
18530           Use software emulation for divides (default).
18531
18532       -mxl-barrel-shift
18533           Use the hardware barrel shifter.
18534
18535       -mxl-pattern-compare
18536           Use pattern compare instructions.
18537
18538       -msmall-divides
18539           Use table lookup optimization for small signed integer divisions.
18540
18541       -mxl-stack-check
18542           This option is deprecated.  Use -fstack-check instead.
18543
18544       -mxl-gp-opt
18545           Use GP-relative ".sdata"/".sbss" sections.
18546
18547       -mxl-multiply-high
18548           Use multiply high instructions for high part of 32x32 multiply.
18549
18550       -mxl-float-convert
18551           Use hardware floating-point conversion instructions.
18552
18553       -mxl-float-sqrt
18554           Use hardware floating-point square root instruction.
18555
18556       -mbig-endian
18557           Generate code for a big-endian target.
18558
18559       -mlittle-endian
18560           Generate code for a little-endian target.
18561
18562       -mxl-reorder
18563           Use reorder instructions (swap and byte reversed load/store).
18564
18565       -mxl-mode-app-model
18566           Select application model app-model.  Valid models are
18567
18568           executable
18569               normal executable (default), uses startup code crt0.o.
18570
18571           -mpic-data-is-text-relative
18572               Assume that the displacement between the text and data segments
18573               is fixed at static link time.  This allows data to be
18574               referenced by offset from start of text address instead of GOT
18575               since PC-relative addressing is not supported.
18576
18577           xmdstub
18578               for use with Xilinx Microprocessor Debugger (XMD) based
18579               software intrusive debug agent called xmdstub. This uses
18580               startup file crt1.o and sets the start address of the program
18581               to 0x800.
18582
18583           bootstrap
18584               for applications that are loaded using a bootloader.  This
18585               model uses startup file crt2.o which does not contain a
18586               processor reset vector handler. This is suitable for
18587               transferring control on a processor reset to the bootloader
18588               rather than the application.
18589
18590           novectors
18591               for applications that do not require any of the MicroBlaze
18592               vectors. This option may be useful for applications running
18593               within a monitoring application. This model uses crt3.o as a
18594               startup file.
18595
18596           Option -xl-mode-app-model is a deprecated alias for -mxl-mode-app-
18597           model.
18598
18599       MIPS Options
18600
18601       -EB Generate big-endian code.
18602
18603       -EL Generate little-endian code.  This is the default for mips*el-*-*
18604           configurations.
18605
18606       -march=arch
18607           Generate code that runs on arch, which can be the name of a generic
18608           MIPS ISA, or the name of a particular processor.  The ISA names
18609           are: mips1, mips2, mips3, mips4, mips32, mips32r2, mips32r3,
18610           mips32r5, mips32r6, mips64, mips64r2, mips64r3, mips64r5 and
18611           mips64r6.  The processor names are: 4kc, 4km, 4kp, 4ksc, 4kec,
18612           4kem, 4kep, 4ksd, 5kc, 5kf, 20kc, 24kc, 24kf2_1, 24kf1_1, 24kec,
18613           24kef2_1, 24kef1_1, 34kc, 34kf2_1, 34kf1_1, 34kn, 74kc, 74kf2_1,
18614           74kf1_1, 74kf3_2, 1004kc, 1004kf2_1, 1004kf1_1, i6400, i6500,
18615           interaptiv, loongson2e, loongson2f, loongson3a, gs464, gs464e,
18616           gs264e, m4k, m14k, m14kc, m14ke, m14kec, m5100, m5101, octeon,
18617           octeon+, octeon2, octeon3, orion, p5600, p6600, r2000, r3000,
18618           r3900, r4000, r4400, r4600, r4650, r4700, r5900, r6000, r8000,
18619           rm7000, rm9000, r10000, r12000, r14000, r16000, sb1, sr71000,
18620           vr4100, vr4111, vr4120, vr4130, vr4300, vr5000, vr5400, vr5500, xlr
18621           and xlp.  The special value from-abi selects the most compatible
18622           architecture for the selected ABI (that is, mips1 for 32-bit ABIs
18623           and mips3 for 64-bit ABIs).
18624
18625           The native Linux/GNU toolchain also supports the value native,
18626           which selects the best architecture option for the host processor.
18627           -march=native has no effect if GCC does not recognize the
18628           processor.
18629
18630           In processor names, a final 000 can be abbreviated as k (for
18631           example, -march=r2k).  Prefixes are optional, and vr may be written
18632           r.
18633
18634           Names of the form nf2_1 refer to processors with FPUs clocked at
18635           half the rate of the core, names of the form nf1_1 refer to
18636           processors with FPUs clocked at the same rate as the core, and
18637           names of the form nf3_2 refer to processors with FPUs clocked a
18638           ratio of 3:2 with respect to the core.  For compatibility reasons,
18639           nf is accepted as a synonym for nf2_1 while nx and bfx are accepted
18640           as synonyms for nf1_1.
18641
18642           GCC defines two macros based on the value of this option.  The
18643           first is "_MIPS_ARCH", which gives the name of target architecture,
18644           as a string.  The second has the form "_MIPS_ARCH_foo", where foo
18645           is the capitalized value of "_MIPS_ARCH".  For example,
18646           -march=r2000 sets "_MIPS_ARCH" to "r2000" and defines the macro
18647           "_MIPS_ARCH_R2000".
18648
18649           Note that the "_MIPS_ARCH" macro uses the processor names given
18650           above.  In other words, it has the full prefix and does not
18651           abbreviate 000 as k.  In the case of from-abi, the macro names the
18652           resolved architecture (either "mips1" or "mips3").  It names the
18653           default architecture when no -march option is given.
18654
18655       -mtune=arch
18656           Optimize for arch.  Among other things, this option controls the
18657           way instructions are scheduled, and the perceived cost of
18658           arithmetic operations.  The list of arch values is the same as for
18659           -march.
18660
18661           When this option is not used, GCC optimizes for the processor
18662           specified by -march.  By using -march and -mtune together, it is
18663           possible to generate code that runs on a family of processors, but
18664           optimize the code for one particular member of that family.
18665
18666           -mtune defines the macros "_MIPS_TUNE" and "_MIPS_TUNE_foo", which
18667           work in the same way as the -march ones described above.
18668
18669       -mips1
18670           Equivalent to -march=mips1.
18671
18672       -mips2
18673           Equivalent to -march=mips2.
18674
18675       -mips3
18676           Equivalent to -march=mips3.
18677
18678       -mips4
18679           Equivalent to -march=mips4.
18680
18681       -mips32
18682           Equivalent to -march=mips32.
18683
18684       -mips32r3
18685           Equivalent to -march=mips32r3.
18686
18687       -mips32r5
18688           Equivalent to -march=mips32r5.
18689
18690       -mips32r6
18691           Equivalent to -march=mips32r6.
18692
18693       -mips64
18694           Equivalent to -march=mips64.
18695
18696       -mips64r2
18697           Equivalent to -march=mips64r2.
18698
18699       -mips64r3
18700           Equivalent to -march=mips64r3.
18701
18702       -mips64r5
18703           Equivalent to -march=mips64r5.
18704
18705       -mips64r6
18706           Equivalent to -march=mips64r6.
18707
18708       -mips16
18709       -mno-mips16
18710           Generate (do not generate) MIPS16 code.  If GCC is targeting a
18711           MIPS32 or MIPS64 architecture, it makes use of the MIPS16e ASE.
18712
18713           MIPS16 code generation can also be controlled on a per-function
18714           basis by means of "mips16" and "nomips16" attributes.
18715
18716       -mflip-mips16
18717           Generate MIPS16 code on alternating functions.  This option is
18718           provided for regression testing of mixed MIPS16/non-MIPS16 code
18719           generation, and is not intended for ordinary use in compiling user
18720           code.
18721
18722       -minterlink-compressed
18723       -mno-interlink-compressed
18724           Require (do not require) that code using the standard
18725           (uncompressed) MIPS ISA be link-compatible with MIPS16 and
18726           microMIPS code, and vice versa.
18727
18728           For example, code using the standard ISA encoding cannot jump
18729           directly to MIPS16 or microMIPS code; it must either use a call or
18730           an indirect jump.  -minterlink-compressed therefore disables direct
18731           jumps unless GCC knows that the target of the jump is not
18732           compressed.
18733
18734       -minterlink-mips16
18735       -mno-interlink-mips16
18736           Aliases of -minterlink-compressed and -mno-interlink-compressed.
18737           These options predate the microMIPS ASE and are retained for
18738           backwards compatibility.
18739
18740       -mabi=32
18741       -mabi=o64
18742       -mabi=n32
18743       -mabi=64
18744       -mabi=eabi
18745           Generate code for the given ABI.
18746
18747           Note that the EABI has a 32-bit and a 64-bit variant.  GCC normally
18748           generates 64-bit code when you select a 64-bit architecture, but
18749           you can use -mgp32 to get 32-bit code instead.
18750
18751           For information about the O64 ABI, see
18752           <http://gcc.gnu.org/projects/mipso64-abi.html>.
18753
18754           GCC supports a variant of the o32 ABI in which floating-point
18755           registers are 64 rather than 32 bits wide.  You can select this
18756           combination with -mabi=32 -mfp64.  This ABI relies on the "mthc1"
18757           and "mfhc1" instructions and is therefore only supported for
18758           MIPS32R2, MIPS32R3 and MIPS32R5 processors.
18759
18760           The register assignments for arguments and return values remain the
18761           same, but each scalar value is passed in a single 64-bit register
18762           rather than a pair of 32-bit registers.  For example, scalar
18763           floating-point values are returned in $f0 only, not a $f0/$f1 pair.
18764           The set of call-saved registers also remains the same in that the
18765           even-numbered double-precision registers are saved.
18766
18767           Two additional variants of the o32 ABI are supported to enable a
18768           transition from 32-bit to 64-bit registers.  These are FPXX
18769           (-mfpxx) and FP64A (-mfp64 -mno-odd-spreg).  The FPXX extension
18770           mandates that all code must execute correctly when run using 32-bit
18771           or 64-bit registers.  The code can be interlinked with either FP32
18772           or FP64, but not both.  The FP64A extension is similar to the FP64
18773           extension but forbids the use of odd-numbered single-precision
18774           registers.  This can be used in conjunction with the "FRE" mode of
18775           FPUs in MIPS32R5 processors and allows both FP32 and FP64A code to
18776           interlink and run in the same process without changing FPU modes.
18777
18778       -mabicalls
18779       -mno-abicalls
18780           Generate (do not generate) code that is suitable for SVR4-style
18781           dynamic objects.  -mabicalls is the default for SVR4-based systems.
18782
18783       -mshared
18784       -mno-shared
18785           Generate (do not generate) code that is fully position-independent,
18786           and that can therefore be linked into shared libraries.  This
18787           option only affects -mabicalls.
18788
18789           All -mabicalls code has traditionally been position-independent,
18790           regardless of options like -fPIC and -fpic.  However, as an
18791           extension, the GNU toolchain allows executables to use absolute
18792           accesses for locally-binding symbols.  It can also use shorter GP
18793           initialization sequences and generate direct calls to locally-
18794           defined functions.  This mode is selected by -mno-shared.
18795
18796           -mno-shared depends on binutils 2.16 or higher and generates
18797           objects that can only be linked by the GNU linker.  However, the
18798           option does not affect the ABI of the final executable; it only
18799           affects the ABI of relocatable objects.  Using -mno-shared
18800           generally makes executables both smaller and quicker.
18801
18802           -mshared is the default.
18803
18804       -mplt
18805       -mno-plt
18806           Assume (do not assume) that the static and dynamic linkers support
18807           PLTs and copy relocations.  This option only affects -mno-shared
18808           -mabicalls.  For the n64 ABI, this option has no effect without
18809           -msym32.
18810
18811           You can make -mplt the default by configuring GCC with
18812           --with-mips-plt.  The default is -mno-plt otherwise.
18813
18814       -mxgot
18815       -mno-xgot
18816           Lift (do not lift) the usual restrictions on the size of the global
18817           offset table.
18818
18819           GCC normally uses a single instruction to load values from the GOT.
18820           While this is relatively efficient, it only works if the GOT is
18821           smaller than about 64k.  Anything larger causes the linker to
18822           report an error such as:
18823
18824                   relocation truncated to fit: R_MIPS_GOT16 foobar
18825
18826           If this happens, you should recompile your code with -mxgot.  This
18827           works with very large GOTs, although the code is also less
18828           efficient, since it takes three instructions to fetch the value of
18829           a global symbol.
18830
18831           Note that some linkers can create multiple GOTs.  If you have such
18832           a linker, you should only need to use -mxgot when a single object
18833           file accesses more than 64k's worth of GOT entries.  Very few do.
18834
18835           These options have no effect unless GCC is generating position
18836           independent code.
18837
18838       -mgp32
18839           Assume that general-purpose registers are 32 bits wide.
18840
18841       -mgp64
18842           Assume that general-purpose registers are 64 bits wide.
18843
18844       -mfp32
18845           Assume that floating-point registers are 32 bits wide.
18846
18847       -mfp64
18848           Assume that floating-point registers are 64 bits wide.
18849
18850       -mfpxx
18851           Do not assume the width of floating-point registers.
18852
18853       -mhard-float
18854           Use floating-point coprocessor instructions.
18855
18856       -msoft-float
18857           Do not use floating-point coprocessor instructions.  Implement
18858           floating-point calculations using library calls instead.
18859
18860       -mno-float
18861           Equivalent to -msoft-float, but additionally asserts that the
18862           program being compiled does not perform any floating-point
18863           operations.  This option is presently supported only by some bare-
18864           metal MIPS configurations, where it may select a special set of
18865           libraries that lack all floating-point support (including, for
18866           example, the floating-point "printf" formats).  If code compiled
18867           with -mno-float accidentally contains floating-point operations, it
18868           is likely to suffer a link-time or run-time failure.
18869
18870       -msingle-float
18871           Assume that the floating-point coprocessor only supports single-
18872           precision operations.
18873
18874       -mdouble-float
18875           Assume that the floating-point coprocessor supports double-
18876           precision operations.  This is the default.
18877
18878       -modd-spreg
18879       -mno-odd-spreg
18880           Enable the use of odd-numbered single-precision floating-point
18881           registers for the o32 ABI.  This is the default for processors that
18882           are known to support these registers.  When using the o32 FPXX ABI,
18883           -mno-odd-spreg is set by default.
18884
18885       -mabs=2008
18886       -mabs=legacy
18887           These options control the treatment of the special not-a-number
18888           (NaN) IEEE 754 floating-point data with the "abs.fmt" and "neg.fmt"
18889           machine instructions.
18890
18891           By default or when -mabs=legacy is used the legacy treatment is
18892           selected.  In this case these instructions are considered
18893           arithmetic and avoided where correct operation is required and the
18894           input operand might be a NaN.  A longer sequence of instructions
18895           that manipulate the sign bit of floating-point datum manually is
18896           used instead unless the -ffinite-math-only option has also been
18897           specified.
18898
18899           The -mabs=2008 option selects the IEEE 754-2008 treatment.  In this
18900           case these instructions are considered non-arithmetic and therefore
18901           operating correctly in all cases, including in particular where the
18902           input operand is a NaN.  These instructions are therefore always
18903           used for the respective operations.
18904
18905       -mnan=2008
18906       -mnan=legacy
18907           These options control the encoding of the special not-a-number
18908           (NaN) IEEE 754 floating-point data.
18909
18910           The -mnan=legacy option selects the legacy encoding.  In this case
18911           quiet NaNs (qNaNs) are denoted by the first bit of their trailing
18912           significand field being 0, whereas signaling NaNs (sNaNs) are
18913           denoted by the first bit of their trailing significand field being
18914           1.
18915
18916           The -mnan=2008 option selects the IEEE 754-2008 encoding.  In this
18917           case qNaNs are denoted by the first bit of their trailing
18918           significand field being 1, whereas sNaNs are denoted by the first
18919           bit of their trailing significand field being 0.
18920
18921           The default is -mnan=legacy unless GCC has been configured with
18922           --with-nan=2008.
18923
18924       -mllsc
18925       -mno-llsc
18926           Use (do not use) ll, sc, and sync instructions to implement atomic
18927           memory built-in functions.  When neither option is specified, GCC
18928           uses the instructions if the target architecture supports them.
18929
18930           -mllsc is useful if the runtime environment can emulate the
18931           instructions and -mno-llsc can be useful when compiling for
18932           nonstandard ISAs.  You can make either option the default by
18933           configuring GCC with --with-llsc and --without-llsc respectively.
18934           --with-llsc is the default for some configurations; see the
18935           installation documentation for details.
18936
18937       -mdsp
18938       -mno-dsp
18939           Use (do not use) revision 1 of the MIPS DSP ASE.
18940             This option defines the preprocessor macro "__mips_dsp".  It also
18941           defines "__mips_dsp_rev" to 1.
18942
18943       -mdspr2
18944       -mno-dspr2
18945           Use (do not use) revision 2 of the MIPS DSP ASE.
18946             This option defines the preprocessor macros "__mips_dsp" and
18947           "__mips_dspr2".  It also defines "__mips_dsp_rev" to 2.
18948
18949       -msmartmips
18950       -mno-smartmips
18951           Use (do not use) the MIPS SmartMIPS ASE.
18952
18953       -mpaired-single
18954       -mno-paired-single
18955           Use (do not use) paired-single floating-point instructions.
18956             This option requires hardware floating-point support to be
18957           enabled.
18958
18959       -mdmx
18960       -mno-mdmx
18961           Use (do not use) MIPS Digital Media Extension instructions.  This
18962           option can only be used when generating 64-bit code and requires
18963           hardware floating-point support to be enabled.
18964
18965       -mips3d
18966       -mno-mips3d
18967           Use (do not use) the MIPS-3D ASE.  The option -mips3d implies
18968           -mpaired-single.
18969
18970       -mmicromips
18971       -mno-micromips
18972           Generate (do not generate) microMIPS code.
18973
18974           MicroMIPS code generation can also be controlled on a per-function
18975           basis by means of "micromips" and "nomicromips" attributes.
18976
18977       -mmt
18978       -mno-mt
18979           Use (do not use) MT Multithreading instructions.
18980
18981       -mmcu
18982       -mno-mcu
18983           Use (do not use) the MIPS MCU ASE instructions.
18984
18985       -meva
18986       -mno-eva
18987           Use (do not use) the MIPS Enhanced Virtual Addressing instructions.
18988
18989       -mvirt
18990       -mno-virt
18991           Use (do not use) the MIPS Virtualization (VZ) instructions.
18992
18993       -mxpa
18994       -mno-xpa
18995           Use (do not use) the MIPS eXtended Physical Address (XPA)
18996           instructions.
18997
18998       -mcrc
18999       -mno-crc
19000           Use (do not use) the MIPS Cyclic Redundancy Check (CRC)
19001           instructions.
19002
19003       -mginv
19004       -mno-ginv
19005           Use (do not use) the MIPS Global INValidate (GINV) instructions.
19006
19007       -mloongson-mmi
19008       -mno-loongson-mmi
19009           Use (do not use) the MIPS Loongson MultiMedia extensions
19010           Instructions (MMI).
19011
19012       -mloongson-ext
19013       -mno-loongson-ext
19014           Use (do not use) the MIPS Loongson EXTensions (EXT) instructions.
19015
19016       -mloongson-ext2
19017       -mno-loongson-ext2
19018           Use (do not use) the MIPS Loongson EXTensions r2 (EXT2)
19019           instructions.
19020
19021       -mlong64
19022           Force "long" types to be 64 bits wide.  See -mlong32 for an
19023           explanation of the default and the way that the pointer size is
19024           determined.
19025
19026       -mlong32
19027           Force "long", "int", and pointer types to be 32 bits wide.
19028
19029           The default size of "int"s, "long"s and pointers depends on the
19030           ABI.  All the supported ABIs use 32-bit "int"s.  The n64 ABI uses
19031           64-bit "long"s, as does the 64-bit EABI; the others use 32-bit
19032           "long"s.  Pointers are the same size as "long"s, or the same size
19033           as integer registers, whichever is smaller.
19034
19035       -msym32
19036       -mno-sym32
19037           Assume (do not assume) that all symbols have 32-bit values,
19038           regardless of the selected ABI.  This option is useful in
19039           combination with -mabi=64 and -mno-abicalls because it allows GCC
19040           to generate shorter and faster references to symbolic addresses.
19041
19042       -G num
19043           Put definitions of externally-visible data in a small data section
19044           if that data is no bigger than num bytes.  GCC can then generate
19045           more efficient accesses to the data; see -mgpopt for details.
19046
19047           The default -G option depends on the configuration.
19048
19049       -mlocal-sdata
19050       -mno-local-sdata
19051           Extend (do not extend) the -G behavior to local data too, such as
19052           to static variables in C.  -mlocal-sdata is the default for all
19053           configurations.
19054
19055           If the linker complains that an application is using too much small
19056           data, you might want to try rebuilding the less performance-
19057           critical parts with -mno-local-sdata.  You might also want to build
19058           large libraries with -mno-local-sdata, so that the libraries leave
19059           more room for the main program.
19060
19061       -mextern-sdata
19062       -mno-extern-sdata
19063           Assume (do not assume) that externally-defined data is in a small
19064           data section if the size of that data is within the -G limit.
19065           -mextern-sdata is the default for all configurations.
19066
19067           If you compile a module Mod with -mextern-sdata -G num -mgpopt, and
19068           Mod references a variable Var that is no bigger than num bytes, you
19069           must make sure that Var is placed in a small data section.  If Var
19070           is defined by another module, you must either compile that module
19071           with a high-enough -G setting or attach a "section" attribute to
19072           Var's definition.  If Var is common, you must link the application
19073           with a high-enough -G setting.
19074
19075           The easiest way of satisfying these restrictions is to compile and
19076           link every module with the same -G option.  However, you may wish
19077           to build a library that supports several different small data
19078           limits.  You can do this by compiling the library with the highest
19079           supported -G setting and additionally using -mno-extern-sdata to
19080           stop the library from making assumptions about externally-defined
19081           data.
19082
19083       -mgpopt
19084       -mno-gpopt
19085           Use (do not use) GP-relative accesses for symbols that are known to
19086           be in a small data section; see -G, -mlocal-sdata and
19087           -mextern-sdata.  -mgpopt is the default for all configurations.
19088
19089           -mno-gpopt is useful for cases where the $gp register might not
19090           hold the value of "_gp".  For example, if the code is part of a
19091           library that might be used in a boot monitor, programs that call
19092           boot monitor routines pass an unknown value in $gp.  (In such
19093           situations, the boot monitor itself is usually compiled with -G0.)
19094
19095           -mno-gpopt implies -mno-local-sdata and -mno-extern-sdata.
19096
19097       -membedded-data
19098       -mno-embedded-data
19099           Allocate variables to the read-only data section first if possible,
19100           then next in the small data section if possible, otherwise in data.
19101           This gives slightly slower code than the default, but reduces the
19102           amount of RAM required when executing, and thus may be preferred
19103           for some embedded systems.
19104
19105       -muninit-const-in-rodata
19106       -mno-uninit-const-in-rodata
19107           Put uninitialized "const" variables in the read-only data section.
19108           This option is only meaningful in conjunction with -membedded-data.
19109
19110       -mcode-readable=setting
19111           Specify whether GCC may generate code that reads from executable
19112           sections.  There are three possible settings:
19113
19114           -mcode-readable=yes
19115               Instructions may freely access executable sections.  This is
19116               the default setting.
19117
19118           -mcode-readable=pcrel
19119               MIPS16 PC-relative load instructions can access executable
19120               sections, but other instructions must not do so.  This option
19121               is useful on 4KSc and 4KSd processors when the code TLBs have
19122               the Read Inhibit bit set.  It is also useful on processors that
19123               can be configured to have a dual instruction/data SRAM
19124               interface and that, like the M4K, automatically redirect PC-
19125               relative loads to the instruction RAM.
19126
19127           -mcode-readable=no
19128               Instructions must not access executable sections.  This option
19129               can be useful on targets that are configured to have a dual
19130               instruction/data SRAM interface but that (unlike the M4K) do
19131               not automatically redirect PC-relative loads to the instruction
19132               RAM.
19133
19134       -msplit-addresses
19135       -mno-split-addresses
19136           Enable (disable) use of the "%hi()" and "%lo()" assembler
19137           relocation operators.  This option has been superseded by
19138           -mexplicit-relocs but is retained for backwards compatibility.
19139
19140       -mexplicit-relocs
19141       -mno-explicit-relocs
19142           Use (do not use) assembler relocation operators when dealing with
19143           symbolic addresses.  The alternative, selected by
19144           -mno-explicit-relocs, is to use assembler macros instead.
19145
19146           -mexplicit-relocs is the default if GCC was configured to use an
19147           assembler that supports relocation operators.
19148
19149       -mcheck-zero-division
19150       -mno-check-zero-division
19151           Trap (do not trap) on integer division by zero.
19152
19153           The default is -mcheck-zero-division.
19154
19155       -mdivide-traps
19156       -mdivide-breaks
19157           MIPS systems check for division by zero by generating either a
19158           conditional trap or a break instruction.  Using traps results in
19159           smaller code, but is only supported on MIPS II and later.  Also,
19160           some versions of the Linux kernel have a bug that prevents trap
19161           from generating the proper signal ("SIGFPE").  Use -mdivide-traps
19162           to allow conditional traps on architectures that support them and
19163           -mdivide-breaks to force the use of breaks.
19164
19165           The default is usually -mdivide-traps, but this can be overridden
19166           at configure time using --with-divide=breaks.  Divide-by-zero
19167           checks can be completely disabled using -mno-check-zero-division.
19168
19169       -mload-store-pairs
19170       -mno-load-store-pairs
19171           Enable (disable) an optimization that pairs consecutive load or
19172           store instructions to enable load/store bonding.  This option is
19173           enabled by default but only takes effect when the selected
19174           architecture is known to support bonding.
19175
19176       -mmemcpy
19177       -mno-memcpy
19178           Force (do not force) the use of "memcpy" for non-trivial block
19179           moves.  The default is -mno-memcpy, which allows GCC to inline most
19180           constant-sized copies.
19181
19182       -mlong-calls
19183       -mno-long-calls
19184           Disable (do not disable) use of the "jal" instruction.  Calling
19185           functions using "jal" is more efficient but requires the caller and
19186           callee to be in the same 256 megabyte segment.
19187
19188           This option has no effect on abicalls code.  The default is
19189           -mno-long-calls.
19190
19191       -mmad
19192       -mno-mad
19193           Enable (disable) use of the "mad", "madu" and "mul" instructions,
19194           as provided by the R4650 ISA.
19195
19196       -mimadd
19197       -mno-imadd
19198           Enable (disable) use of the "madd" and "msub" integer instructions.
19199           The default is -mimadd on architectures that support "madd" and
19200           "msub" except for the 74k architecture where it was found to
19201           generate slower code.
19202
19203       -mfused-madd
19204       -mno-fused-madd
19205           Enable (disable) use of the floating-point multiply-accumulate
19206           instructions, when they are available.  The default is
19207           -mfused-madd.
19208
19209           On the R8000 CPU when multiply-accumulate instructions are used,
19210           the intermediate product is calculated to infinite precision and is
19211           not subject to the FCSR Flush to Zero bit.  This may be undesirable
19212           in some circumstances.  On other processors the result is
19213           numerically identical to the equivalent computation using separate
19214           multiply, add, subtract and negate instructions.
19215
19216       -nocpp
19217           Tell the MIPS assembler to not run its preprocessor over user
19218           assembler files (with a .s suffix) when assembling them.
19219
19220       -mfix-24k
19221       -mno-fix-24k
19222           Work around the 24K E48 (lost data on stores during refill) errata.
19223           The workarounds are implemented by the assembler rather than by
19224           GCC.
19225
19226       -mfix-r4000
19227       -mno-fix-r4000
19228           Work around certain R4000 CPU errata:
19229
19230           -   A double-word or a variable shift may give an incorrect result
19231               if executed immediately after starting an integer division.
19232
19233           -   A double-word or a variable shift may give an incorrect result
19234               if executed while an integer multiplication is in progress.
19235
19236           -   An integer division may give an incorrect result if started in
19237               a delay slot of a taken branch or a jump.
19238
19239       -mfix-r4400
19240       -mno-fix-r4400
19241           Work around certain R4400 CPU errata:
19242
19243           -   A double-word or a variable shift may give an incorrect result
19244               if executed immediately after starting an integer division.
19245
19246       -mfix-r10000
19247       -mno-fix-r10000
19248           Work around certain R10000 errata:
19249
19250           -   "ll"/"sc" sequences may not behave atomically on revisions
19251               prior to 3.0.  They may deadlock on revisions 2.6 and earlier.
19252
19253           This option can only be used if the target architecture supports
19254           branch-likely instructions.  -mfix-r10000 is the default when
19255           -march=r10000 is used; -mno-fix-r10000 is the default otherwise.
19256
19257       -mfix-r5900
19258       -mno-fix-r5900
19259           Do not attempt to schedule the preceding instruction into the delay
19260           slot of a branch instruction placed at the end of a short loop of
19261           six instructions or fewer and always schedule a "nop" instruction
19262           there instead.  The short loop bug under certain conditions causes
19263           loops to execute only once or twice, due to a hardware bug in the
19264           R5900 chip.  The workaround is implemented by the assembler rather
19265           than by GCC.
19266
19267       -mfix-rm7000
19268       -mno-fix-rm7000
19269           Work around the RM7000 "dmult"/"dmultu" errata.  The workarounds
19270           are implemented by the assembler rather than by GCC.
19271
19272       -mfix-vr4120
19273       -mno-fix-vr4120
19274           Work around certain VR4120 errata:
19275
19276           -   "dmultu" does not always produce the correct result.
19277
19278           -   "div" and "ddiv" do not always produce the correct result if
19279               one of the operands is negative.
19280
19281           The workarounds for the division errata rely on special functions
19282           in libgcc.a.  At present, these functions are only provided by the
19283           "mips64vr*-elf" configurations.
19284
19285           Other VR4120 errata require a NOP to be inserted between certain
19286           pairs of instructions.  These errata are handled by the assembler,
19287           not by GCC itself.
19288
19289       -mfix-vr4130
19290           Work around the VR4130 "mflo"/"mfhi" errata.  The workarounds are
19291           implemented by the assembler rather than by GCC, although GCC
19292           avoids using "mflo" and "mfhi" if the VR4130 "macc", "macchi",
19293           "dmacc" and "dmacchi" instructions are available instead.
19294
19295       -mfix-sb1
19296       -mno-fix-sb1
19297           Work around certain SB-1 CPU core errata.  (This flag currently
19298           works around the SB-1 revision 2 "F1" and "F2" floating-point
19299           errata.)
19300
19301       -mr10k-cache-barrier=setting
19302           Specify whether GCC should insert cache barriers to avoid the side
19303           effects of speculation on R10K processors.
19304
19305           In common with many processors, the R10K tries to predict the
19306           outcome of a conditional branch and speculatively executes
19307           instructions from the "taken" branch.  It later aborts these
19308           instructions if the predicted outcome is wrong.  However, on the
19309           R10K, even aborted instructions can have side effects.
19310
19311           This problem only affects kernel stores and, depending on the
19312           system, kernel loads.  As an example, a speculatively-executed
19313           store may load the target memory into cache and mark the cache line
19314           as dirty, even if the store itself is later aborted.  If a DMA
19315           operation writes to the same area of memory before the "dirty" line
19316           is flushed, the cached data overwrites the DMA-ed data.  See the
19317           R10K processor manual for a full description, including other
19318           potential problems.
19319
19320           One workaround is to insert cache barrier instructions before every
19321           memory access that might be speculatively executed and that might
19322           have side effects even if aborted.  -mr10k-cache-barrier=setting
19323           controls GCC's implementation of this workaround.  It assumes that
19324           aborted accesses to any byte in the following regions does not have
19325           side effects:
19326
19327           1.  the memory occupied by the current function's stack frame;
19328
19329           2.  the memory occupied by an incoming stack argument;
19330
19331           3.  the memory occupied by an object with a link-time-constant
19332               address.
19333
19334           It is the kernel's responsibility to ensure that speculative
19335           accesses to these regions are indeed safe.
19336
19337           If the input program contains a function declaration such as:
19338
19339                   void foo (void);
19340
19341           then the implementation of "foo" must allow "j foo" and "jal foo"
19342           to be executed speculatively.  GCC honors this restriction for
19343           functions it compiles itself.  It expects non-GCC functions (such
19344           as hand-written assembly code) to do the same.
19345
19346           The option has three forms:
19347
19348           -mr10k-cache-barrier=load-store
19349               Insert a cache barrier before a load or store that might be
19350               speculatively executed and that might have side effects even if
19351               aborted.
19352
19353           -mr10k-cache-barrier=store
19354               Insert a cache barrier before a store that might be
19355               speculatively executed and that might have side effects even if
19356               aborted.
19357
19358           -mr10k-cache-barrier=none
19359               Disable the insertion of cache barriers.  This is the default
19360               setting.
19361
19362       -mflush-func=func
19363       -mno-flush-func
19364           Specifies the function to call to flush the I and D caches, or to
19365           not call any such function.  If called, the function must take the
19366           same arguments as the common "_flush_func", that is, the address of
19367           the memory range for which the cache is being flushed, the size of
19368           the memory range, and the number 3 (to flush both caches).  The
19369           default depends on the target GCC was configured for, but commonly
19370           is either "_flush_func" or "__cpu_flush".
19371
19372       mbranch-cost=num
19373           Set the cost of branches to roughly num "simple" instructions.
19374           This cost is only a heuristic and is not guaranteed to produce
19375           consistent results across releases.  A zero cost redundantly
19376           selects the default, which is based on the -mtune setting.
19377
19378       -mbranch-likely
19379       -mno-branch-likely
19380           Enable or disable use of Branch Likely instructions, regardless of
19381           the default for the selected architecture.  By default, Branch
19382           Likely instructions may be generated if they are supported by the
19383           selected architecture.  An exception is for the MIPS32 and MIPS64
19384           architectures and processors that implement those architectures;
19385           for those, Branch Likely instructions are not be generated by
19386           default because the MIPS32 and MIPS64 architectures specifically
19387           deprecate their use.
19388
19389       -mcompact-branches=never
19390       -mcompact-branches=optimal
19391       -mcompact-branches=always
19392           These options control which form of branches will be generated.
19393           The default is -mcompact-branches=optimal.
19394
19395           The -mcompact-branches=never option ensures that compact branch
19396           instructions will never be generated.
19397
19398           The -mcompact-branches=always option ensures that a compact branch
19399           instruction will be generated if available.  If a compact branch
19400           instruction is not available, a delay slot form of the branch will
19401           be used instead.
19402
19403           This option is supported from MIPS Release 6 onwards.
19404
19405           The -mcompact-branches=optimal option will cause a delay slot
19406           branch to be used if one is available in the current ISA and the
19407           delay slot is successfully filled.  If the delay slot is not
19408           filled, a compact branch will be chosen if one is available.
19409
19410       -mfp-exceptions
19411       -mno-fp-exceptions
19412           Specifies whether FP exceptions are enabled.  This affects how FP
19413           instructions are scheduled for some processors.  The default is
19414           that FP exceptions are enabled.
19415
19416           For instance, on the SB-1, if FP exceptions are disabled, and we
19417           are emitting 64-bit code, then we can use both FP pipes.
19418           Otherwise, we can only use one FP pipe.
19419
19420       -mvr4130-align
19421       -mno-vr4130-align
19422           The VR4130 pipeline is two-way superscalar, but can only issue two
19423           instructions together if the first one is 8-byte aligned.  When
19424           this option is enabled, GCC aligns pairs of instructions that it
19425           thinks should execute in parallel.
19426
19427           This option only has an effect when optimizing for the VR4130.  It
19428           normally makes code faster, but at the expense of making it bigger.
19429           It is enabled by default at optimization level -O3.
19430
19431       -msynci
19432       -mno-synci
19433           Enable (disable) generation of "synci" instructions on
19434           architectures that support it.  The "synci" instructions (if
19435           enabled) are generated when "__builtin___clear_cache" is compiled.
19436
19437           This option defaults to -mno-synci, but the default can be
19438           overridden by configuring GCC with --with-synci.
19439
19440           When compiling code for single processor systems, it is generally
19441           safe to use "synci".  However, on many multi-core (SMP) systems, it
19442           does not invalidate the instruction caches on all cores and may
19443           lead to undefined behavior.
19444
19445       -mrelax-pic-calls
19446       -mno-relax-pic-calls
19447           Try to turn PIC calls that are normally dispatched via register $25
19448           into direct calls.  This is only possible if the linker can resolve
19449           the destination at link time and if the destination is within range
19450           for a direct call.
19451
19452           -mrelax-pic-calls is the default if GCC was configured to use an
19453           assembler and a linker that support the ".reloc" assembly directive
19454           and -mexplicit-relocs is in effect.  With -mno-explicit-relocs,
19455           this optimization can be performed by the assembler and the linker
19456           alone without help from the compiler.
19457
19458       -mmcount-ra-address
19459       -mno-mcount-ra-address
19460           Emit (do not emit) code that allows "_mcount" to modify the calling
19461           function's return address.  When enabled, this option extends the
19462           usual "_mcount" interface with a new ra-address parameter, which
19463           has type "intptr_t *" and is passed in register $12.  "_mcount" can
19464           then modify the return address by doing both of the following:
19465
19466           *   Returning the new address in register $31.
19467
19468           *   Storing the new address in "*ra-address", if ra-address is
19469               nonnull.
19470
19471           The default is -mno-mcount-ra-address.
19472
19473       -mframe-header-opt
19474       -mno-frame-header-opt
19475           Enable (disable) frame header optimization in the o32 ABI.  When
19476           using the o32 ABI, calling functions will allocate 16 bytes on the
19477           stack for the called function to write out register arguments.
19478           When enabled, this optimization will suppress the allocation of the
19479           frame header if it can be determined that it is unused.
19480
19481           This optimization is off by default at all optimization levels.
19482
19483       -mlxc1-sxc1
19484       -mno-lxc1-sxc1
19485           When applicable, enable (disable) the generation of "lwxc1",
19486           "swxc1", "ldxc1", "sdxc1" instructions.  Enabled by default.
19487
19488       -mmadd4
19489       -mno-madd4
19490           When applicable, enable (disable) the generation of 4-operand
19491           "madd.s", "madd.d" and related instructions.  Enabled by default.
19492
19493       MMIX Options
19494
19495       These options are defined for the MMIX:
19496
19497       -mlibfuncs
19498       -mno-libfuncs
19499           Specify that intrinsic library functions are being compiled,
19500           passing all values in registers, no matter the size.
19501
19502       -mepsilon
19503       -mno-epsilon
19504           Generate floating-point comparison instructions that compare with
19505           respect to the "rE" epsilon register.
19506
19507       -mabi=mmixware
19508       -mabi=gnu
19509           Generate code that passes function parameters and return values
19510           that (in the called function) are seen as registers $0 and up, as
19511           opposed to the GNU ABI which uses global registers $231 and up.
19512
19513       -mzero-extend
19514       -mno-zero-extend
19515           When reading data from memory in sizes shorter than 64 bits, use
19516           (do not use) zero-extending load instructions by default, rather
19517           than sign-extending ones.
19518
19519       -mknuthdiv
19520       -mno-knuthdiv
19521           Make the result of a division yielding a remainder have the same
19522           sign as the divisor.  With the default, -mno-knuthdiv, the sign of
19523           the remainder follows the sign of the dividend.  Both methods are
19524           arithmetically valid, the latter being almost exclusively used.
19525
19526       -mtoplevel-symbols
19527       -mno-toplevel-symbols
19528           Prepend (do not prepend) a : to all global symbols, so the assembly
19529           code can be used with the "PREFIX" assembly directive.
19530
19531       -melf
19532           Generate an executable in the ELF format, rather than the default
19533           mmo format used by the mmix simulator.
19534
19535       -mbranch-predict
19536       -mno-branch-predict
19537           Use (do not use) the probable-branch instructions, when static
19538           branch prediction indicates a probable branch.
19539
19540       -mbase-addresses
19541       -mno-base-addresses
19542           Generate (do not generate) code that uses base addresses.  Using a
19543           base address automatically generates a request (handled by the
19544           assembler and the linker) for a constant to be set up in a global
19545           register.  The register is used for one or more base address
19546           requests within the range 0 to 255 from the value held in the
19547           register.  The generally leads to short and fast code, but the
19548           number of different data items that can be addressed is limited.
19549           This means that a program that uses lots of static data may require
19550           -mno-base-addresses.
19551
19552       -msingle-exit
19553       -mno-single-exit
19554           Force (do not force) generated code to have a single exit point in
19555           each function.
19556
19557       MN10300 Options
19558
19559       These -m options are defined for Matsushita MN10300 architectures:
19560
19561       -mmult-bug
19562           Generate code to avoid bugs in the multiply instructions for the
19563           MN10300 processors.  This is the default.
19564
19565       -mno-mult-bug
19566           Do not generate code to avoid bugs in the multiply instructions for
19567           the MN10300 processors.
19568
19569       -mam33
19570           Generate code using features specific to the AM33 processor.
19571
19572       -mno-am33
19573           Do not generate code using features specific to the AM33 processor.
19574           This is the default.
19575
19576       -mam33-2
19577           Generate code using features specific to the AM33/2.0 processor.
19578
19579       -mam34
19580           Generate code using features specific to the AM34 processor.
19581
19582       -mtune=cpu-type
19583           Use the timing characteristics of the indicated CPU type when
19584           scheduling instructions.  This does not change the targeted
19585           processor type.  The CPU type must be one of mn10300, am33, am33-2
19586           or am34.
19587
19588       -mreturn-pointer-on-d0
19589           When generating a function that returns a pointer, return the
19590           pointer in both "a0" and "d0".  Otherwise, the pointer is returned
19591           only in "a0", and attempts to call such functions without a
19592           prototype result in errors.  Note that this option is on by
19593           default; use -mno-return-pointer-on-d0 to disable it.
19594
19595       -mno-crt0
19596           Do not link in the C run-time initialization object file.
19597
19598       -mrelax
19599           Indicate to the linker that it should perform a relaxation
19600           optimization pass to shorten branches, calls and absolute memory
19601           addresses.  This option only has an effect when used on the command
19602           line for the final link step.
19603
19604           This option makes symbolic debugging impossible.
19605
19606       -mliw
19607           Allow the compiler to generate Long Instruction Word instructions
19608           if the target is the AM33 or later.  This is the default.  This
19609           option defines the preprocessor macro "__LIW__".
19610
19611       -mno-liw
19612           Do not allow the compiler to generate Long Instruction Word
19613           instructions.  This option defines the preprocessor macro
19614           "__NO_LIW__".
19615
19616       -msetlb
19617           Allow the compiler to generate the SETLB and Lcc instructions if
19618           the target is the AM33 or later.  This is the default.  This option
19619           defines the preprocessor macro "__SETLB__".
19620
19621       -mno-setlb
19622           Do not allow the compiler to generate SETLB or Lcc instructions.
19623           This option defines the preprocessor macro "__NO_SETLB__".
19624
19625       Moxie Options
19626
19627       -meb
19628           Generate big-endian code.  This is the default for moxie-*-*
19629           configurations.
19630
19631       -mel
19632           Generate little-endian code.
19633
19634       -mmul.x
19635           Generate mul.x and umul.x instructions.  This is the default for
19636           moxiebox-*-* configurations.
19637
19638       -mno-crt0
19639           Do not link in the C run-time initialization object file.
19640
19641       MSP430 Options
19642
19643       These options are defined for the MSP430:
19644
19645       -masm-hex
19646           Force assembly output to always use hex constants.  Normally such
19647           constants are signed decimals, but this option is available for
19648           testsuite and/or aesthetic purposes.
19649
19650       -mmcu=
19651           Select the MCU to target.  This is used to create a C preprocessor
19652           symbol based upon the MCU name, converted to upper case and pre-
19653           and post-fixed with __.  This in turn is used by the msp430.h
19654           header file to select an MCU-specific supplementary header file.
19655
19656           The option also sets the ISA to use.  If the MCU name is one that
19657           is known to only support the 430 ISA then that is selected,
19658           otherwise the 430X ISA is selected.  A generic MCU name of msp430
19659           can also be used to select the 430 ISA.  Similarly the generic
19660           msp430x MCU name selects the 430X ISA.
19661
19662           In addition an MCU-specific linker script is added to the linker
19663           command line.  The script's name is the name of the MCU with .ld
19664           appended.  Thus specifying -mmcu=xxx on the gcc command line
19665           defines the C preprocessor symbol "__XXX__" and cause the linker to
19666           search for a script called xxx.ld.
19667
19668           This option is also passed on to the assembler.
19669
19670       -mwarn-mcu
19671       -mno-warn-mcu
19672           This option enables or disables warnings about conflicts between
19673           the MCU name specified by the -mmcu option and the ISA set by the
19674           -mcpu option and/or the hardware multiply support set by the
19675           -mhwmult option.  It also toggles warnings about unrecognized MCU
19676           names.  This option is on by default.
19677
19678       -mcpu=
19679           Specifies the ISA to use.  Accepted values are msp430, msp430x and
19680           msp430xv2.  This option is deprecated.  The -mmcu= option should be
19681           used to select the ISA.
19682
19683       -msim
19684           Link to the simulator runtime libraries and linker script.
19685           Overrides any scripts that would be selected by the -mmcu= option.
19686
19687       -mlarge
19688           Use large-model addressing (20-bit pointers, 32-bit "size_t").
19689
19690       -msmall
19691           Use small-model addressing (16-bit pointers, 16-bit "size_t").
19692
19693       -mrelax
19694           This option is passed to the assembler and linker, and allows the
19695           linker to perform certain optimizations that cannot be done until
19696           the final link.
19697
19698       mhwmult=
19699           Describes the type of hardware multiply supported by the target.
19700           Accepted values are none for no hardware multiply, 16bit for the
19701           original 16-bit-only multiply supported by early MCUs.  32bit for
19702           the 16/32-bit multiply supported by later MCUs and f5series for the
19703           16/32-bit multiply supported by F5-series MCUs.  A value of auto
19704           can also be given.  This tells GCC to deduce the hardware multiply
19705           support based upon the MCU name provided by the -mmcu option.  If
19706           no -mmcu option is specified or if the MCU name is not recognized
19707           then no hardware multiply support is assumed.  "auto" is the
19708           default setting.
19709
19710           Hardware multiplies are normally performed by calling a library
19711           routine.  This saves space in the generated code.  When compiling
19712           at -O3 or higher however the hardware multiplier is invoked inline.
19713           This makes for bigger, but faster code.
19714
19715           The hardware multiply routines disable interrupts whilst running
19716           and restore the previous interrupt state when they finish.  This
19717           makes them safe to use inside interrupt handlers as well as in
19718           normal code.
19719
19720       -minrt
19721           Enable the use of a minimum runtime environment - no static
19722           initializers or constructors.  This is intended for memory-
19723           constrained devices.  The compiler includes special symbols in some
19724           objects that tell the linker and runtime which code fragments are
19725           required.
19726
19727       -mcode-region=
19728       -mdata-region=
19729           These options tell the compiler where to place functions and data
19730           that do not have one of the "lower", "upper", "either" or "section"
19731           attributes.  Possible values are "lower", "upper", "either" or
19732           "any".  The first three behave like the corresponding attribute.
19733           The fourth possible value - "any" - is the default.  It leaves
19734           placement entirely up to the linker script and how it assigns the
19735           standard sections (".text", ".data", etc) to the memory regions.
19736
19737       -msilicon-errata=
19738           This option passes on a request to assembler to enable the fixes
19739           for the named silicon errata.
19740
19741       -msilicon-errata-warn=
19742           This option passes on a request to the assembler to enable warning
19743           messages when a silicon errata might need to be applied.
19744
19745       NDS32 Options
19746
19747       These options are defined for NDS32 implementations:
19748
19749       -mbig-endian
19750           Generate code in big-endian mode.
19751
19752       -mlittle-endian
19753           Generate code in little-endian mode.
19754
19755       -mreduced-regs
19756           Use reduced-set registers for register allocation.
19757
19758       -mfull-regs
19759           Use full-set registers for register allocation.
19760
19761       -mcmov
19762           Generate conditional move instructions.
19763
19764       -mno-cmov
19765           Do not generate conditional move instructions.
19766
19767       -mext-perf
19768           Generate performance extension instructions.
19769
19770       -mno-ext-perf
19771           Do not generate performance extension instructions.
19772
19773       -mext-perf2
19774           Generate performance extension 2 instructions.
19775
19776       -mno-ext-perf2
19777           Do not generate performance extension 2 instructions.
19778
19779       -mext-string
19780           Generate string extension instructions.
19781
19782       -mno-ext-string
19783           Do not generate string extension instructions.
19784
19785       -mv3push
19786           Generate v3 push25/pop25 instructions.
19787
19788       -mno-v3push
19789           Do not generate v3 push25/pop25 instructions.
19790
19791       -m16-bit
19792           Generate 16-bit instructions.
19793
19794       -mno-16-bit
19795           Do not generate 16-bit instructions.
19796
19797       -misr-vector-size=num
19798           Specify the size of each interrupt vector, which must be 4 or 16.
19799
19800       -mcache-block-size=num
19801           Specify the size of each cache block, which must be a power of 2
19802           between 4 and 512.
19803
19804       -march=arch
19805           Specify the name of the target architecture.
19806
19807       -mcmodel=code-model
19808           Set the code model to one of
19809
19810           small
19811               All the data and read-only data segments must be within 512KB
19812               addressing space.  The text segment must be within 16MB
19813               addressing space.
19814
19815           medium
19816               The data segment must be within 512KB while the read-only data
19817               segment can be within 4GB addressing space.  The text segment
19818               should be still within 16MB addressing space.
19819
19820           large
19821               All the text and data segments can be within 4GB addressing
19822               space.
19823
19824       -mctor-dtor
19825           Enable constructor/destructor feature.
19826
19827       -mrelax
19828           Guide linker to relax instructions.
19829
19830       Nios II Options
19831
19832       These are the options defined for the Altera Nios II processor.
19833
19834       -G num
19835           Put global and static objects less than or equal to num bytes into
19836           the small data or BSS sections instead of the normal data or BSS
19837           sections.  The default value of num is 8.
19838
19839       -mgpopt=option
19840       -mgpopt
19841       -mno-gpopt
19842           Generate (do not generate) GP-relative accesses.  The following
19843           option names are recognized:
19844
19845           none
19846               Do not generate GP-relative accesses.
19847
19848           local
19849               Generate GP-relative accesses for small data objects that are
19850               not external, weak, or uninitialized common symbols.  Also use
19851               GP-relative addressing for objects that have been explicitly
19852               placed in a small data section via a "section" attribute.
19853
19854           global
19855               As for local, but also generate GP-relative accesses for small
19856               data objects that are external, weak, or common.  If you use
19857               this option, you must ensure that all parts of your program
19858               (including libraries) are compiled with the same -G setting.
19859
19860           data
19861               Generate GP-relative accesses for all data objects in the
19862               program.  If you use this option, the entire data and BSS
19863               segments of your program must fit in 64K of memory and you must
19864               use an appropriate linker script to allocate them within the
19865               addressable range of the global pointer.
19866
19867           all Generate GP-relative addresses for function pointers as well as
19868               data pointers.  If you use this option, the entire text, data,
19869               and BSS segments of your program must fit in 64K of memory and
19870               you must use an appropriate linker script to allocate them
19871               within the addressable range of the global pointer.
19872
19873           -mgpopt is equivalent to -mgpopt=local, and -mno-gpopt is
19874           equivalent to -mgpopt=none.
19875
19876           The default is -mgpopt except when -fpic or -fPIC is specified to
19877           generate position-independent code.  Note that the Nios II ABI does
19878           not permit GP-relative accesses from shared libraries.
19879
19880           You may need to specify -mno-gpopt explicitly when building
19881           programs that include large amounts of small data, including large
19882           GOT data sections.  In this case, the 16-bit offset for GP-relative
19883           addressing may not be large enough to allow access to the entire
19884           small data section.
19885
19886       -mgprel-sec=regexp
19887           This option specifies additional section names that can be accessed
19888           via GP-relative addressing.  It is most useful in conjunction with
19889           "section" attributes on variable declarations and a custom linker
19890           script.  The regexp is a POSIX Extended Regular Expression.
19891
19892           This option does not affect the behavior of the -G option, and the
19893           specified sections are in addition to the standard ".sdata" and
19894           ".sbss" small-data sections that are recognized by -mgpopt.
19895
19896       -mr0rel-sec=regexp
19897           This option specifies names of sections that can be accessed via a
19898           16-bit offset from "r0"; that is, in the low 32K or high 32K of the
19899           32-bit address space.  It is most useful in conjunction with
19900           "section" attributes on variable declarations and a custom linker
19901           script.  The regexp is a POSIX Extended Regular Expression.
19902
19903           In contrast to the use of GP-relative addressing for small data,
19904           zero-based addressing is never generated by default and there are
19905           no conventional section names used in standard linker scripts for
19906           sections in the low or high areas of memory.
19907
19908       -mel
19909       -meb
19910           Generate little-endian (default) or big-endian (experimental) code,
19911           respectively.
19912
19913       -march=arch
19914           This specifies the name of the target Nios II architecture.  GCC
19915           uses this name to determine what kind of instructions it can emit
19916           when generating assembly code.  Permissible names are: r1, r2.
19917
19918           The preprocessor macro "__nios2_arch__" is available to programs,
19919           with value 1 or 2, indicating the targeted ISA level.
19920
19921       -mbypass-cache
19922       -mno-bypass-cache
19923           Force all load and store instructions to always bypass cache by
19924           using I/O variants of the instructions. The default is not to
19925           bypass the cache.
19926
19927       -mno-cache-volatile
19928       -mcache-volatile
19929           Volatile memory access bypass the cache using the I/O variants of
19930           the load and store instructions. The default is not to bypass the
19931           cache.
19932
19933       -mno-fast-sw-div
19934       -mfast-sw-div
19935           Do not use table-based fast divide for small numbers. The default
19936           is to use the fast divide at -O3 and above.
19937
19938       -mno-hw-mul
19939       -mhw-mul
19940       -mno-hw-mulx
19941       -mhw-mulx
19942       -mno-hw-div
19943       -mhw-div
19944           Enable or disable emitting "mul", "mulx" and "div" family of
19945           instructions by the compiler. The default is to emit "mul" and not
19946           emit "div" and "mulx".
19947
19948       -mbmx
19949       -mno-bmx
19950       -mcdx
19951       -mno-cdx
19952           Enable or disable generation of Nios II R2 BMX (bit manipulation)
19953           and CDX (code density) instructions.  Enabling these instructions
19954           also requires -march=r2.  Since these instructions are optional
19955           extensions to the R2 architecture, the default is not to emit them.
19956
19957       -mcustom-insn=N
19958       -mno-custom-insn
19959           Each -mcustom-insn=N option enables use of a custom instruction
19960           with encoding N when generating code that uses insn.  For example,
19961           -mcustom-fadds=253 generates custom instruction 253 for single-
19962           precision floating-point add operations instead of the default
19963           behavior of using a library call.
19964
19965           The following values of insn are supported.  Except as otherwise
19966           noted, floating-point operations are expected to be implemented
19967           with normal IEEE 754 semantics and correspond directly to the C
19968           operators or the equivalent GCC built-in functions.
19969
19970           Single-precision floating point:
19971
19972           fadds, fsubs, fdivs, fmuls
19973               Binary arithmetic operations.
19974
19975           fnegs
19976               Unary negation.
19977
19978           fabss
19979               Unary absolute value.
19980
19981           fcmpeqs, fcmpges, fcmpgts, fcmples, fcmplts, fcmpnes
19982               Comparison operations.
19983
19984           fmins, fmaxs
19985               Floating-point minimum and maximum.  These instructions are
19986               only generated if -ffinite-math-only is specified.
19987
19988           fsqrts
19989               Unary square root operation.
19990
19991           fcoss, fsins, ftans, fatans, fexps, flogs
19992               Floating-point trigonometric and exponential functions.  These
19993               instructions are only generated if -funsafe-math-optimizations
19994               is also specified.
19995
19996           Double-precision floating point:
19997
19998           faddd, fsubd, fdivd, fmuld
19999               Binary arithmetic operations.
20000
20001           fnegd
20002               Unary negation.
20003
20004           fabsd
20005               Unary absolute value.
20006
20007           fcmpeqd, fcmpged, fcmpgtd, fcmpled, fcmpltd, fcmpned
20008               Comparison operations.
20009
20010           fmind, fmaxd
20011               Double-precision minimum and maximum.  These instructions are
20012               only generated if -ffinite-math-only is specified.
20013
20014           fsqrtd
20015               Unary square root operation.
20016
20017           fcosd, fsind, ftand, fatand, fexpd, flogd
20018               Double-precision trigonometric and exponential functions.
20019               These instructions are only generated if
20020               -funsafe-math-optimizations is also specified.
20021
20022           Conversions:
20023
20024           fextsd
20025               Conversion from single precision to double precision.
20026
20027           ftruncds
20028               Conversion from double precision to single precision.
20029
20030           fixsi, fixsu, fixdi, fixdu
20031               Conversion from floating point to signed or unsigned integer
20032               types, with truncation towards zero.
20033
20034           round
20035               Conversion from single-precision floating point to signed
20036               integer, rounding to the nearest integer and ties away from
20037               zero.  This corresponds to the "__builtin_lroundf" function
20038               when -fno-math-errno is used.
20039
20040           floatis, floatus, floatid, floatud
20041               Conversion from signed or unsigned integer types to floating-
20042               point types.
20043
20044           In addition, all of the following transfer instructions for
20045           internal registers X and Y must be provided to use any of the
20046           double-precision floating-point instructions.  Custom instructions
20047           taking two double-precision source operands expect the first
20048           operand in the 64-bit register X.  The other operand (or only
20049           operand of a unary operation) is given to the custom arithmetic
20050           instruction with the least significant half in source register src1
20051           and the most significant half in src2.  A custom instruction that
20052           returns a double-precision result returns the most significant 32
20053           bits in the destination register and the other half in 32-bit
20054           register Y.  GCC automatically generates the necessary code
20055           sequences to write register X and/or read register Y when double-
20056           precision floating-point instructions are used.
20057
20058           fwrx
20059               Write src1 into the least significant half of X and src2 into
20060               the most significant half of X.
20061
20062           fwry
20063               Write src1 into Y.
20064
20065           frdxhi, frdxlo
20066               Read the most or least (respectively) significant half of X and
20067               store it in dest.
20068
20069           frdy
20070               Read the value of Y and store it into dest.
20071
20072           Note that you can gain more local control over generation of Nios
20073           II custom instructions by using the "target("custom-insn=N")" and
20074           "target("no-custom-insn")" function attributes or pragmas.
20075
20076       -mcustom-fpu-cfg=name
20077           This option enables a predefined, named set of custom instruction
20078           encodings (see -mcustom-insn above).  Currently, the following sets
20079           are defined:
20080
20081           -mcustom-fpu-cfg=60-1 is equivalent to: -mcustom-fmuls=252
20082           -mcustom-fadds=253 -mcustom-fsubs=254 -fsingle-precision-constant
20083
20084           -mcustom-fpu-cfg=60-2 is equivalent to: -mcustom-fmuls=252
20085           -mcustom-fadds=253 -mcustom-fsubs=254 -mcustom-fdivs=255
20086           -fsingle-precision-constant
20087
20088           -mcustom-fpu-cfg=72-3 is equivalent to: -mcustom-floatus=243
20089           -mcustom-fixsi=244 -mcustom-floatis=245 -mcustom-fcmpgts=246
20090           -mcustom-fcmples=249 -mcustom-fcmpeqs=250 -mcustom-fcmpnes=251
20091           -mcustom-fmuls=252 -mcustom-fadds=253 -mcustom-fsubs=254
20092           -mcustom-fdivs=255 -fsingle-precision-constant
20093
20094           Custom instruction assignments given by individual -mcustom-insn=
20095           options override those given by -mcustom-fpu-cfg=, regardless of
20096           the order of the options on the command line.
20097
20098           Note that you can gain more local control over selection of a FPU
20099           configuration by using the "target("custom-fpu-cfg=name")" function
20100           attribute or pragma.
20101
20102       These additional -m options are available for the Altera Nios II ELF
20103       (bare-metal) target:
20104
20105       -mhal
20106           Link with HAL BSP.  This suppresses linking with the GCC-provided C
20107           runtime startup and termination code, and is typically used in
20108           conjunction with -msys-crt0= to specify the location of the
20109           alternate startup code provided by the HAL BSP.
20110
20111       -msmallc
20112           Link with a limited version of the C library, -lsmallc, rather than
20113           Newlib.
20114
20115       -msys-crt0=startfile
20116           startfile is the file name of the startfile (crt0) to use when
20117           linking.  This option is only useful in conjunction with -mhal.
20118
20119       -msys-lib=systemlib
20120           systemlib is the library name of the library that provides low-
20121           level system calls required by the C library, e.g. "read" and
20122           "write".  This option is typically used to link with a library
20123           provided by a HAL BSP.
20124
20125       Nvidia PTX Options
20126
20127       These options are defined for Nvidia PTX:
20128
20129       -m32
20130       -m64
20131           Generate code for 32-bit or 64-bit ABI.
20132
20133       -misa=ISA-string
20134           Generate code for given the specified PTX ISA (e.g. sm_35).  ISA
20135           strings must be lower-case.  Valid ISA strings include sm_30 and
20136           sm_35.  The default ISA is sm_30.
20137
20138       -mmainkernel
20139           Link in code for a __main kernel.  This is for stand-alone instead
20140           of offloading execution.
20141
20142       -moptimize
20143           Apply partitioned execution optimizations.  This is the default
20144           when any level of optimization is selected.
20145
20146       -msoft-stack
20147           Generate code that does not use ".local" memory directly for stack
20148           storage. Instead, a per-warp stack pointer is maintained
20149           explicitly. This enables variable-length stack allocation (with
20150           variable-length arrays or "alloca"), and when global memory is used
20151           for underlying storage, makes it possible to access automatic
20152           variables from other threads, or with atomic instructions. This
20153           code generation variant is used for OpenMP offloading, but the
20154           option is exposed on its own for the purpose of testing the
20155           compiler; to generate code suitable for linking into programs using
20156           OpenMP offloading, use option -mgomp.
20157
20158       -muniform-simt
20159           Switch to code generation variant that allows to execute all
20160           threads in each warp, while maintaining memory state and side
20161           effects as if only one thread in each warp was active outside of
20162           OpenMP SIMD regions.  All atomic operations and calls to runtime
20163           (malloc, free, vprintf) are conditionally executed (iff current
20164           lane index equals the master lane index), and the register being
20165           assigned is copied via a shuffle instruction from the master lane.
20166           Outside of SIMD regions lane 0 is the master; inside, each thread
20167           sees itself as the master.  Shared memory array "int __nvptx_uni[]"
20168           stores all-zeros or all-ones bitmasks for each warp, indicating
20169           current mode (0 outside of SIMD regions).  Each thread can bitwise-
20170           and the bitmask at position "tid.y" with current lane index to
20171           compute the master lane index.
20172
20173       -mgomp
20174           Generate code for use in OpenMP offloading: enables -msoft-stack
20175           and -muniform-simt options, and selects corresponding multilib
20176           variant.
20177
20178       OpenRISC Options
20179
20180       These options are defined for OpenRISC:
20181
20182       -mboard=name
20183           Configure a board specific runtime.  This will be passed to the
20184           linker for newlib board library linking.  The default is "or1ksim".
20185
20186       -mnewlib
20187           For compatibility, it's always newlib for elf now.
20188
20189       -mhard-div
20190           Generate code for hardware which supports divide instructions.
20191           This is the default.
20192
20193       -mhard-mul
20194           Generate code for hardware which supports multiply instructions.
20195           This is the default.
20196
20197       -mcmov
20198           Generate code for hardware which supports the conditional move
20199           ("l.cmov") instruction.
20200
20201       -mror
20202           Generate code for hardware which supports rotate right
20203           instructions.
20204
20205       -msext
20206           Generate code for hardware which supports sign-extension
20207           instructions.
20208
20209       -msfimm
20210           Generate code for hardware which supports set flag immediate
20211           ("l.sf*i") instructions.
20212
20213       -mshftimm
20214           Generate code for hardware which supports shift immediate related
20215           instructions (i.e. "l.srai", "l.srli", "l.slli", "1.rori").  Note,
20216           to enable generation of the "l.rori" instruction the -mror flag
20217           must also be specified.
20218
20219       -msoft-div
20220           Generate code for hardware which requires divide instruction
20221           emulation.
20222
20223       -msoft-mul
20224           Generate code for hardware which requires multiply instruction
20225           emulation.
20226
20227       PDP-11 Options
20228
20229       These options are defined for the PDP-11:
20230
20231       -mfpu
20232           Use hardware FPP floating point.  This is the default.  (FIS
20233           floating point on the PDP-11/40 is not supported.)  Implies -m45.
20234
20235       -msoft-float
20236           Do not use hardware floating point.
20237
20238       -mac0
20239           Return floating-point results in ac0 (fr0 in Unix assembler
20240           syntax).
20241
20242       -mno-ac0
20243           Return floating-point results in memory.  This is the default.
20244
20245       -m40
20246           Generate code for a PDP-11/40.  Implies -msoft-float -mno-split.
20247
20248       -m45
20249           Generate code for a PDP-11/45.  This is the default.
20250
20251       -m10
20252           Generate code for a PDP-11/10.  Implies -msoft-float -mno-split.
20253
20254       -mint16
20255       -mno-int32
20256           Use 16-bit "int".  This is the default.
20257
20258       -mint32
20259       -mno-int16
20260           Use 32-bit "int".
20261
20262       -msplit
20263           Target has split instruction and data space.  Implies -m45.
20264
20265       -munix-asm
20266           Use Unix assembler syntax.
20267
20268       -mdec-asm
20269           Use DEC assembler syntax.
20270
20271       -mgnu-asm
20272           Use GNU assembler syntax.  This is the default.
20273
20274       -mlra
20275           Use the new LRA register allocator.  By default, the old "reload"
20276           allocator is used.
20277
20278       picoChip Options
20279
20280       These -m options are defined for picoChip implementations:
20281
20282       -mae=ae_type
20283           Set the instruction set, register set, and instruction scheduling
20284           parameters for array element type ae_type.  Supported values for
20285           ae_type are ANY, MUL, and MAC.
20286
20287           -mae=ANY selects a completely generic AE type.  Code generated with
20288           this option runs on any of the other AE types.  The code is not as
20289           efficient as it would be if compiled for a specific AE type, and
20290           some types of operation (e.g., multiplication) do not work properly
20291           on all types of AE.
20292
20293           -mae=MUL selects a MUL AE type.  This is the most useful AE type
20294           for compiled code, and is the default.
20295
20296           -mae=MAC selects a DSP-style MAC AE.  Code compiled with this
20297           option may suffer from poor performance of byte (char)
20298           manipulation, since the DSP AE does not provide hardware support
20299           for byte load/stores.
20300
20301       -msymbol-as-address
20302           Enable the compiler to directly use a symbol name as an address in
20303           a load/store instruction, without first loading it into a register.
20304           Typically, the use of this option generates larger programs, which
20305           run faster than when the option isn't used.  However, the results
20306           vary from program to program, so it is left as a user option,
20307           rather than being permanently enabled.
20308
20309       -mno-inefficient-warnings
20310           Disables warnings about the generation of inefficient code.  These
20311           warnings can be generated, for example, when compiling code that
20312           performs byte-level memory operations on the MAC AE type.  The MAC
20313           AE has no hardware support for byte-level memory operations, so all
20314           byte load/stores must be synthesized from word load/store
20315           operations.  This is inefficient and a warning is generated to
20316           indicate that you should rewrite the code to avoid byte operations,
20317           or to target an AE type that has the necessary hardware support.
20318           This option disables these warnings.
20319
20320       PowerPC Options
20321
20322       These are listed under
20323
20324       RISC-V Options
20325
20326       These command-line options are defined for RISC-V targets:
20327
20328       -mbranch-cost=n
20329           Set the cost of branches to roughly n instructions.
20330
20331       -mplt
20332       -mno-plt
20333           When generating PIC code, do or don't allow the use of PLTs.
20334           Ignored for non-PIC.  The default is -mplt.
20335
20336       -mabi=ABI-string
20337           Specify integer and floating-point calling convention.  ABI-string
20338           contains two parts: the size of integer types and the registers
20339           used for floating-point types.  For example -march=rv64ifd
20340           -mabi=lp64d means that long and pointers are 64-bit (implicitly
20341           defining int to be 32-bit), and that floating-point values up to 64
20342           bits wide are passed in F registers.  Contrast this with
20343           -march=rv64ifd -mabi=lp64f, which still allows the compiler to
20344           generate code that uses the F and D extensions but only allows
20345           floating-point values up to 32 bits long to be passed in registers;
20346           or -march=rv64ifd -mabi=lp64, in which no floating-point arguments
20347           will be passed in registers.
20348
20349           The default for this argument is system dependent, users who want a
20350           specific calling convention should specify one explicitly.  The
20351           valid calling conventions are: ilp32, ilp32f, ilp32d, lp64, lp64f,
20352           and lp64d.  Some calling conventions are impossible to implement on
20353           some ISAs: for example, -march=rv32if -mabi=ilp32d is invalid
20354           because the ABI requires 64-bit values be passed in F registers,
20355           but F registers are only 32 bits wide.  There is also the ilp32e
20356           ABI that can only be used with the rv32e architecture.  This ABI is
20357           not well specified at present, and is subject to change.
20358
20359       -mfdiv
20360       -mno-fdiv
20361           Do or don't use hardware floating-point divide and square root
20362           instructions.  This requires the F or D extensions for floating-
20363           point registers.  The default is to use them if the specified
20364           architecture has these instructions.
20365
20366       -mdiv
20367       -mno-div
20368           Do or don't use hardware instructions for integer division.  This
20369           requires the M extension.  The default is to use them if the
20370           specified architecture has these instructions.
20371
20372       -march=ISA-string
20373           Generate code for given RISC-V ISA (e.g. rv64im).  ISA strings must
20374           be lower-case.  Examples include rv64i, rv32g, rv32e, and rv32imaf.
20375
20376       -mtune=processor-string
20377           Optimize the output for the given processor, specified by
20378           microarchitecture name.  Permissible values for this option are:
20379           rocket, sifive-3-series, sifive-5-series, sifive-7-series, and
20380           size.
20381
20382           When -mtune= is not specified, the default is rocket.
20383
20384           The size choice is not intended for use by end-users.  This is used
20385           when -Os is specified.  It overrides the instruction cost info
20386           provided by -mtune=, but does not override the pipeline info.  This
20387           helps reduce code size while still giving good performance.
20388
20389       -mpreferred-stack-boundary=num
20390           Attempt to keep the stack boundary aligned to a 2 raised to num
20391           byte boundary.  If -mpreferred-stack-boundary is not specified, the
20392           default is 4 (16 bytes or 128-bits).
20393
20394           Warning: If you use this switch, then you must build all modules
20395           with the same value, including any libraries.  This includes the
20396           system libraries and startup modules.
20397
20398       -msmall-data-limit=n
20399           Put global and static data smaller than n bytes into a special
20400           section (on some targets).
20401
20402       -msave-restore
20403       -mno-save-restore
20404           Do or don't use smaller but slower prologue and epilogue code that
20405           uses library function calls.  The default is to use fast inline
20406           prologues and epilogues.
20407
20408       -mstrict-align
20409       -mno-strict-align
20410           Do not or do generate unaligned memory accesses.  The default is
20411           set depending on whether the processor we are optimizing for
20412           supports fast unaligned access or not.
20413
20414       -mcmodel=medlow
20415           Generate code for the medium-low code model. The program and its
20416           statically defined symbols must lie within a single 2 GiB address
20417           range and must lie between absolute addresses -2 GiB and +2 GiB.
20418           Programs can be statically or dynamically linked. This is the
20419           default code model.
20420
20421       -mcmodel=medany
20422           Generate code for the medium-any code model. The program and its
20423           statically defined symbols must be within any single 2 GiB address
20424           range. Programs can be statically or dynamically linked.
20425
20426       -mexplicit-relocs
20427       -mno-exlicit-relocs
20428           Use or do not use assembler relocation operators when dealing with
20429           symbolic addresses.  The alternative is to use assembler macros
20430           instead, which may limit optimization.
20431
20432       -mrelax
20433       -mno-relax
20434           Take advantage of linker relaxations to reduce the number of
20435           instructions required to materialize symbol addresses. The default
20436           is to take advantage of linker relaxations.
20437
20438       -memit-attribute
20439       -mno-emit-attribute
20440           Emit (do not emit) RISC-V attribute to record extra information
20441           into ELF objects.  This feature requires at least binutils 2.32.
20442
20443       RL78 Options
20444
20445       -msim
20446           Links in additional target libraries to support operation within a
20447           simulator.
20448
20449       -mmul=none
20450       -mmul=g10
20451       -mmul=g13
20452       -mmul=g14
20453       -mmul=rl78
20454           Specifies the type of hardware multiplication and division support
20455           to be used.  The simplest is "none", which uses software for both
20456           multiplication and division.  This is the default.  The "g13" value
20457           is for the hardware multiply/divide peripheral found on the
20458           RL78/G13 (S2 core) targets.  The "g14" value selects the use of the
20459           multiplication and division instructions supported by the RL78/G14
20460           (S3 core) parts.  The value "rl78" is an alias for "g14" and the
20461           value "mg10" is an alias for "none".
20462
20463           In addition a C preprocessor macro is defined, based upon the
20464           setting of this option.  Possible values are: "__RL78_MUL_NONE__",
20465           "__RL78_MUL_G13__" or "__RL78_MUL_G14__".
20466
20467       -mcpu=g10
20468       -mcpu=g13
20469       -mcpu=g14
20470       -mcpu=rl78
20471           Specifies the RL78 core to target.  The default is the G14 core,
20472           also known as an S3 core or just RL78.  The G13 or S2 core does not
20473           have multiply or divide instructions, instead it uses a hardware
20474           peripheral for these operations.  The G10 or S1 core does not have
20475           register banks, so it uses a different calling convention.
20476
20477           If this option is set it also selects the type of hardware multiply
20478           support to use, unless this is overridden by an explicit -mmul=none
20479           option on the command line.  Thus specifying -mcpu=g13 enables the
20480           use of the G13 hardware multiply peripheral and specifying
20481           -mcpu=g10 disables the use of hardware multiplications altogether.
20482
20483           Note, although the RL78/G14 core is the default target, specifying
20484           -mcpu=g14 or -mcpu=rl78 on the command line does change the
20485           behavior of the toolchain since it also enables G14 hardware
20486           multiply support.  If these options are not specified on the
20487           command line then software multiplication routines will be used
20488           even though the code targets the RL78 core.  This is for backwards
20489           compatibility with older toolchains which did not have hardware
20490           multiply and divide support.
20491
20492           In addition a C preprocessor macro is defined, based upon the
20493           setting of this option.  Possible values are: "__RL78_G10__",
20494           "__RL78_G13__" or "__RL78_G14__".
20495
20496       -mg10
20497       -mg13
20498       -mg14
20499       -mrl78
20500           These are aliases for the corresponding -mcpu= option.  They are
20501           provided for backwards compatibility.
20502
20503       -mallregs
20504           Allow the compiler to use all of the available registers.  By
20505           default registers "r24..r31" are reserved for use in interrupt
20506           handlers.  With this option enabled these registers can be used in
20507           ordinary functions as well.
20508
20509       -m64bit-doubles
20510       -m32bit-doubles
20511           Make the "double" data type be 64 bits (-m64bit-doubles) or 32 bits
20512           (-m32bit-doubles) in size.  The default is -m32bit-doubles.
20513
20514       -msave-mduc-in-interrupts
20515       -mno-save-mduc-in-interrupts
20516           Specifies that interrupt handler functions should preserve the MDUC
20517           registers.  This is only necessary if normal code might use the
20518           MDUC registers, for example because it performs multiplication and
20519           division operations.  The default is to ignore the MDUC registers
20520           as this makes the interrupt handlers faster.  The target option
20521           -mg13 needs to be passed for this to work as this feature is only
20522           available on the G13 target (S2 core).  The MDUC registers will
20523           only be saved if the interrupt handler performs a multiplication or
20524           division operation or it calls another function.
20525
20526       IBM RS/6000 and PowerPC Options
20527
20528       These -m options are defined for the IBM RS/6000 and PowerPC:
20529
20530       -mpowerpc-gpopt
20531       -mno-powerpc-gpopt
20532       -mpowerpc-gfxopt
20533       -mno-powerpc-gfxopt
20534       -mpowerpc64
20535       -mno-powerpc64
20536       -mmfcrf
20537       -mno-mfcrf
20538       -mpopcntb
20539       -mno-popcntb
20540       -mpopcntd
20541       -mno-popcntd
20542       -mfprnd
20543       -mno-fprnd
20544       -mcmpb
20545       -mno-cmpb
20546       -mmfpgpr
20547       -mno-mfpgpr
20548       -mhard-dfp
20549       -mno-hard-dfp
20550           You use these options to specify which instructions are available
20551           on the processor you are using.  The default value of these options
20552           is determined when configuring GCC.  Specifying the -mcpu=cpu_type
20553           overrides the specification of these options.  We recommend you use
20554           the -mcpu=cpu_type option rather than the options listed above.
20555
20556           Specifying -mpowerpc-gpopt allows GCC to use the optional PowerPC
20557           architecture instructions in the General Purpose group, including
20558           floating-point square root.  Specifying -mpowerpc-gfxopt allows GCC
20559           to use the optional PowerPC architecture instructions in the
20560           Graphics group, including floating-point select.
20561
20562           The -mmfcrf option allows GCC to generate the move from condition
20563           register field instruction implemented on the POWER4 processor and
20564           other processors that support the PowerPC V2.01 architecture.  The
20565           -mpopcntb option allows GCC to generate the popcount and double-
20566           precision FP reciprocal estimate instruction implemented on the
20567           POWER5 processor and other processors that support the PowerPC
20568           V2.02 architecture.  The -mpopcntd option allows GCC to generate
20569           the popcount instruction implemented on the POWER7 processor and
20570           other processors that support the PowerPC V2.06 architecture.  The
20571           -mfprnd option allows GCC to generate the FP round to integer
20572           instructions implemented on the POWER5+ processor and other
20573           processors that support the PowerPC V2.03 architecture.  The -mcmpb
20574           option allows GCC to generate the compare bytes instruction
20575           implemented on the POWER6 processor and other processors that
20576           support the PowerPC V2.05 architecture.  The -mmfpgpr option allows
20577           GCC to generate the FP move to/from general-purpose register
20578           instructions implemented on the POWER6X processor and other
20579           processors that support the extended PowerPC V2.05 architecture.
20580           The -mhard-dfp option allows GCC to generate the decimal floating-
20581           point instructions implemented on some POWER processors.
20582
20583           The -mpowerpc64 option allows GCC to generate the additional 64-bit
20584           instructions that are found in the full PowerPC64 architecture and
20585           to treat GPRs as 64-bit, doubleword quantities.  GCC defaults to
20586           -mno-powerpc64.
20587
20588       -mcpu=cpu_type
20589           Set architecture type, register usage, and instruction scheduling
20590           parameters for machine type cpu_type.  Supported values for
20591           cpu_type are 401, 403, 405, 405fp, 440, 440fp, 464, 464fp, 476,
20592           476fp, 505, 601, 602, 603, 603e, 604, 604e, 620, 630, 740, 7400,
20593           7450, 750, 801, 821, 823, 860, 970, 8540, a2, e300c2, e300c3,
20594           e500mc, e500mc64, e5500, e6500, ec603e, G3, G4, G5, titan, power3,
20595           power4, power5, power5+, power6, power6x, power7, power8, power9,
20596           powerpc, powerpc64, powerpc64le, rs64, and native.
20597
20598           -mcpu=powerpc, -mcpu=powerpc64, and -mcpu=powerpc64le specify pure
20599           32-bit PowerPC (either endian), 64-bit big endian PowerPC and
20600           64-bit little endian PowerPC architecture machine types, with an
20601           appropriate, generic processor model assumed for scheduling
20602           purposes.
20603
20604           Specifying native as cpu type detects and selects the architecture
20605           option that corresponds to the host processor of the system
20606           performing the compilation.  -mcpu=native has no effect if GCC does
20607           not recognize the processor.
20608
20609           The other options specify a specific processor.  Code generated
20610           under those options runs best on that processor, and may not run at
20611           all on others.
20612
20613           The -mcpu options automatically enable or disable the following
20614           options:
20615
20616           -maltivec  -mfprnd  -mhard-float  -mmfcrf  -mmultiple -mpopcntb
20617           -mpopcntd  -mpowerpc64 -mpowerpc-gpopt  -mpowerpc-gfxopt -mmulhw
20618           -mdlmzb  -mmfpgpr  -mvsx -mcrypto  -mhtm  -mpower8-fusion
20619           -mpower8-vector -mquad-memory  -mquad-memory-atomic  -mfloat128
20620           -mfloat128-hardware
20621
20622           The particular options set for any particular CPU varies between
20623           compiler versions, depending on what setting seems to produce
20624           optimal code for that CPU; it doesn't necessarily reflect the
20625           actual hardware's capabilities.  If you wish to set an individual
20626           option to a particular value, you may specify it after the -mcpu
20627           option, like -mcpu=970 -mno-altivec.
20628
20629           On AIX, the -maltivec and -mpowerpc64 options are not enabled or
20630           disabled by the -mcpu option at present because AIX does not have
20631           full support for these options.  You may still enable or disable
20632           them individually if you're sure it'll work in your environment.
20633
20634       -mtune=cpu_type
20635           Set the instruction scheduling parameters for machine type
20636           cpu_type, but do not set the architecture type or register usage,
20637           as -mcpu=cpu_type does.  The same values for cpu_type are used for
20638           -mtune as for -mcpu.  If both are specified, the code generated
20639           uses the architecture and registers set by -mcpu, but the
20640           scheduling parameters set by -mtune.
20641
20642       -mcmodel=small
20643           Generate PowerPC64 code for the small model: The TOC is limited to
20644           64k.
20645
20646       -mcmodel=medium
20647           Generate PowerPC64 code for the medium model: The TOC and other
20648           static data may be up to a total of 4G in size.  This is the
20649           default for 64-bit Linux.
20650
20651       -mcmodel=large
20652           Generate PowerPC64 code for the large model: The TOC may be up to
20653           4G in size.  Other data and code is only limited by the 64-bit
20654           address space.
20655
20656       -maltivec
20657       -mno-altivec
20658           Generate code that uses (does not use) AltiVec instructions, and
20659           also enable the use of built-in functions that allow more direct
20660           access to the AltiVec instruction set.  You may also need to set
20661           -mabi=altivec to adjust the current ABI with AltiVec ABI
20662           enhancements.
20663
20664           When -maltivec is used, the element order for AltiVec intrinsics
20665           such as "vec_splat", "vec_extract", and "vec_insert" match array
20666           element order corresponding to the endianness of the target.  That
20667           is, element zero identifies the leftmost element in a vector
20668           register when targeting a big-endian platform, and identifies the
20669           rightmost element in a vector register when targeting a little-
20670           endian platform.
20671
20672       -mvrsave
20673       -mno-vrsave
20674           Generate VRSAVE instructions when generating AltiVec code.
20675
20676       -msecure-plt
20677           Generate code that allows ld and ld.so to build executables and
20678           shared libraries with non-executable ".plt" and ".got" sections.
20679           This is a PowerPC 32-bit SYSV ABI option.
20680
20681       -mbss-plt
20682           Generate code that uses a BSS ".plt" section that ld.so fills in,
20683           and requires ".plt" and ".got" sections that are both writable and
20684           executable.  This is a PowerPC 32-bit SYSV ABI option.
20685
20686       -misel
20687       -mno-isel
20688           This switch enables or disables the generation of ISEL
20689           instructions.
20690
20691       -mvsx
20692       -mno-vsx
20693           Generate code that uses (does not use) vector/scalar (VSX)
20694           instructions, and also enable the use of built-in functions that
20695           allow more direct access to the VSX instruction set.
20696
20697       -mcrypto
20698       -mno-crypto
20699           Enable the use (disable) of the built-in functions that allow
20700           direct access to the cryptographic instructions that were added in
20701           version 2.07 of the PowerPC ISA.
20702
20703       -mhtm
20704       -mno-htm
20705           Enable (disable) the use of the built-in functions that allow
20706           direct access to the Hardware Transactional Memory (HTM)
20707           instructions that were added in version 2.07 of the PowerPC ISA.
20708
20709       -mpower8-fusion
20710       -mno-power8-fusion
20711           Generate code that keeps (does not keeps) some integer operations
20712           adjacent so that the instructions can be fused together on power8
20713           and later processors.
20714
20715       -mpower8-vector
20716       -mno-power8-vector
20717           Generate code that uses (does not use) the vector and scalar
20718           instructions that were added in version 2.07 of the PowerPC ISA.
20719           Also enable the use of built-in functions that allow more direct
20720           access to the vector instructions.
20721
20722       -mquad-memory
20723       -mno-quad-memory
20724           Generate code that uses (does not use) the non-atomic quad word
20725           memory instructions.  The -mquad-memory option requires use of
20726           64-bit mode.
20727
20728       -mquad-memory-atomic
20729       -mno-quad-memory-atomic
20730           Generate code that uses (does not use) the atomic quad word memory
20731           instructions.  The -mquad-memory-atomic option requires use of
20732           64-bit mode.
20733
20734       -mfloat128
20735       -mno-float128
20736           Enable/disable the __float128 keyword for IEEE 128-bit floating
20737           point and use either software emulation for IEEE 128-bit floating
20738           point or hardware instructions.
20739
20740           The VSX instruction set (-mvsx, -mcpu=power7, -mcpu=power8), or
20741           -mcpu=power9 must be enabled to use the IEEE 128-bit floating point
20742           support.  The IEEE 128-bit floating point support only works on
20743           PowerPC Linux systems.
20744
20745           The default for -mfloat128 is enabled on PowerPC Linux systems
20746           using the VSX instruction set, and disabled on other systems.
20747
20748           If you use the ISA 3.0 instruction set (-mpower9-vector or
20749           -mcpu=power9) on a 64-bit system, the IEEE 128-bit floating point
20750           support will also enable the generation of ISA 3.0 IEEE 128-bit
20751           floating point instructions.  Otherwise, if you do not specify to
20752           generate ISA 3.0 instructions or you are targeting a 32-bit big
20753           endian system, IEEE 128-bit floating point will be done with
20754           software emulation.
20755
20756       -mfloat128-hardware
20757       -mno-float128-hardware
20758           Enable/disable using ISA 3.0 hardware instructions to support the
20759           __float128 data type.
20760
20761           The default for -mfloat128-hardware is enabled on PowerPC Linux
20762           systems using the ISA 3.0 instruction set, and disabled on other
20763           systems.
20764
20765       -m32
20766       -m64
20767           Generate code for 32-bit or 64-bit environments of Darwin and SVR4
20768           targets (including GNU/Linux).  The 32-bit environment sets int,
20769           long and pointer to 32 bits and generates code that runs on any
20770           PowerPC variant.  The 64-bit environment sets int to 32 bits and
20771           long and pointer to 64 bits, and generates code for PowerPC64, as
20772           for -mpowerpc64.
20773
20774       -mfull-toc
20775       -mno-fp-in-toc
20776       -mno-sum-in-toc
20777       -mminimal-toc
20778           Modify generation of the TOC (Table Of Contents), which is created
20779           for every executable file.  The -mfull-toc option is selected by
20780           default.  In that case, GCC allocates at least one TOC entry for
20781           each unique non-automatic variable reference in your program.  GCC
20782           also places floating-point constants in the TOC.  However, only
20783           16,384 entries are available in the TOC.
20784
20785           If you receive a linker error message that saying you have
20786           overflowed the available TOC space, you can reduce the amount of
20787           TOC space used with the -mno-fp-in-toc and -mno-sum-in-toc options.
20788           -mno-fp-in-toc prevents GCC from putting floating-point constants
20789           in the TOC and -mno-sum-in-toc forces GCC to generate code to
20790           calculate the sum of an address and a constant at run time instead
20791           of putting that sum into the TOC.  You may specify one or both of
20792           these options.  Each causes GCC to produce very slightly slower and
20793           larger code at the expense of conserving TOC space.
20794
20795           If you still run out of space in the TOC even when you specify both
20796           of these options, specify -mminimal-toc instead.  This option
20797           causes GCC to make only one TOC entry for every file.  When you
20798           specify this option, GCC produces code that is slower and larger
20799           but which uses extremely little TOC space.  You may wish to use
20800           this option only on files that contain less frequently-executed
20801           code.
20802
20803       -maix64
20804       -maix32
20805           Enable 64-bit AIX ABI and calling convention: 64-bit pointers,
20806           64-bit "long" type, and the infrastructure needed to support them.
20807           Specifying -maix64 implies -mpowerpc64, while -maix32 disables the
20808           64-bit ABI and implies -mno-powerpc64.  GCC defaults to -maix32.
20809
20810       -mxl-compat
20811       -mno-xl-compat
20812           Produce code that conforms more closely to IBM XL compiler
20813           semantics when using AIX-compatible ABI.  Pass floating-point
20814           arguments to prototyped functions beyond the register save area
20815           (RSA) on the stack in addition to argument FPRs.  Do not assume
20816           that most significant double in 128-bit long double value is
20817           properly rounded when comparing values and converting to double.
20818           Use XL symbol names for long double support routines.
20819
20820           The AIX calling convention was extended but not initially
20821           documented to handle an obscure K&R C case of calling a function
20822           that takes the address of its arguments with fewer arguments than
20823           declared.  IBM XL compilers access floating-point arguments that do
20824           not fit in the RSA from the stack when a subroutine is compiled
20825           without optimization.  Because always storing floating-point
20826           arguments on the stack is inefficient and rarely needed, this
20827           option is not enabled by default and only is necessary when calling
20828           subroutines compiled by IBM XL compilers without optimization.
20829
20830       -mpe
20831           Support IBM RS/6000 SP Parallel Environment (PE).  Link an
20832           application written to use message passing with special startup
20833           code to enable the application to run.  The system must have PE
20834           installed in the standard location (/usr/lpp/ppe.poe/), or the
20835           specs file must be overridden with the -specs= option to specify
20836           the appropriate directory location.  The Parallel Environment does
20837           not support threads, so the -mpe option and the -pthread option are
20838           incompatible.
20839
20840       -malign-natural
20841       -malign-power
20842           On AIX, 32-bit Darwin, and 64-bit PowerPC GNU/Linux, the option
20843           -malign-natural overrides the ABI-defined alignment of larger
20844           types, such as floating-point doubles, on their natural size-based
20845           boundary.  The option -malign-power instructs GCC to follow the
20846           ABI-specified alignment rules.  GCC defaults to the standard
20847           alignment defined in the ABI.
20848
20849           On 64-bit Darwin, natural alignment is the default, and
20850           -malign-power is not supported.
20851
20852       -msoft-float
20853       -mhard-float
20854           Generate code that does not use (uses) the floating-point register
20855           set.  Software floating-point emulation is provided if you use the
20856           -msoft-float option, and pass the option to GCC when linking.
20857
20858       -mmultiple
20859       -mno-multiple
20860           Generate code that uses (does not use) the load multiple word
20861           instructions and the store multiple word instructions.  These
20862           instructions are generated by default on POWER systems, and not
20863           generated on PowerPC systems.  Do not use -mmultiple on little-
20864           endian PowerPC systems, since those instructions do not work when
20865           the processor is in little-endian mode.  The exceptions are PPC740
20866           and PPC750 which permit these instructions in little-endian mode.
20867
20868       -mupdate
20869       -mno-update
20870           Generate code that uses (does not use) the load or store
20871           instructions that update the base register to the address of the
20872           calculated memory location.  These instructions are generated by
20873           default.  If you use -mno-update, there is a small window between
20874           the time that the stack pointer is updated and the address of the
20875           previous frame is stored, which means code that walks the stack
20876           frame across interrupts or signals may get corrupted data.
20877
20878       -mavoid-indexed-addresses
20879       -mno-avoid-indexed-addresses
20880           Generate code that tries to avoid (not avoid) the use of indexed
20881           load or store instructions. These instructions can incur a
20882           performance penalty on Power6 processors in certain situations,
20883           such as when stepping through large arrays that cross a 16M
20884           boundary.  This option is enabled by default when targeting Power6
20885           and disabled otherwise.
20886
20887       -mfused-madd
20888       -mno-fused-madd
20889           Generate code that uses (does not use) the floating-point multiply
20890           and accumulate instructions.  These instructions are generated by
20891           default if hardware floating point is used.  The machine-dependent
20892           -mfused-madd option is now mapped to the machine-independent
20893           -ffp-contract=fast option, and -mno-fused-madd is mapped to
20894           -ffp-contract=off.
20895
20896       -mmulhw
20897       -mno-mulhw
20898           Generate code that uses (does not use) the half-word multiply and
20899           multiply-accumulate instructions on the IBM 405, 440, 464 and 476
20900           processors.  These instructions are generated by default when
20901           targeting those processors.
20902
20903       -mdlmzb
20904       -mno-dlmzb
20905           Generate code that uses (does not use) the string-search dlmzb
20906           instruction on the IBM 405, 440, 464 and 476 processors.  This
20907           instruction is generated by default when targeting those
20908           processors.
20909
20910       -mno-bit-align
20911       -mbit-align
20912           On System V.4 and embedded PowerPC systems do not (do) force
20913           structures and unions that contain bit-fields to be aligned to the
20914           base type of the bit-field.
20915
20916           For example, by default a structure containing nothing but 8
20917           "unsigned" bit-fields of length 1 is aligned to a 4-byte boundary
20918           and has a size of 4 bytes.  By using -mno-bit-align, the structure
20919           is aligned to a 1-byte boundary and is 1 byte in size.
20920
20921       -mno-strict-align
20922       -mstrict-align
20923           On System V.4 and embedded PowerPC systems do not (do) assume that
20924           unaligned memory references are handled by the system.
20925
20926       -mrelocatable
20927       -mno-relocatable
20928           Generate code that allows (does not allow) a static executable to
20929           be relocated to a different address at run time.  A simple embedded
20930           PowerPC system loader should relocate the entire contents of
20931           ".got2" and 4-byte locations listed in the ".fixup" section, a
20932           table of 32-bit addresses generated by this option.  For this to
20933           work, all objects linked together must be compiled with
20934           -mrelocatable or -mrelocatable-lib.  -mrelocatable code aligns the
20935           stack to an 8-byte boundary.
20936
20937       -mrelocatable-lib
20938       -mno-relocatable-lib
20939           Like -mrelocatable, -mrelocatable-lib generates a ".fixup" section
20940           to allow static executables to be relocated at run time, but
20941           -mrelocatable-lib does not use the smaller stack alignment of
20942           -mrelocatable.  Objects compiled with -mrelocatable-lib may be
20943           linked with objects compiled with any combination of the
20944           -mrelocatable options.
20945
20946       -mno-toc
20947       -mtoc
20948           On System V.4 and embedded PowerPC systems do not (do) assume that
20949           register 2 contains a pointer to a global area pointing to the
20950           addresses used in the program.
20951
20952       -mlittle
20953       -mlittle-endian
20954           On System V.4 and embedded PowerPC systems compile code for the
20955           processor in little-endian mode.  The -mlittle-endian option is the
20956           same as -mlittle.
20957
20958       -mbig
20959       -mbig-endian
20960           On System V.4 and embedded PowerPC systems compile code for the
20961           processor in big-endian mode.  The -mbig-endian option is the same
20962           as -mbig.
20963
20964       -mdynamic-no-pic
20965           On Darwin and Mac OS X systems, compile code so that it is not
20966           relocatable, but that its external references are relocatable.  The
20967           resulting code is suitable for applications, but not shared
20968           libraries.
20969
20970       -msingle-pic-base
20971           Treat the register used for PIC addressing as read-only, rather
20972           than loading it in the prologue for each function.  The runtime
20973           system is responsible for initializing this register with an
20974           appropriate value before execution begins.
20975
20976       -mprioritize-restricted-insns=priority
20977           This option controls the priority that is assigned to dispatch-slot
20978           restricted instructions during the second scheduling pass.  The
20979           argument priority takes the value 0, 1, or 2 to assign no, highest,
20980           or second-highest (respectively) priority to dispatch-slot
20981           restricted instructions.
20982
20983       -msched-costly-dep=dependence_type
20984           This option controls which dependences are considered costly by the
20985           target during instruction scheduling.  The argument dependence_type
20986           takes one of the following values:
20987
20988           no  No dependence is costly.
20989
20990           all All dependences are costly.
20991
20992           true_store_to_load
20993               A true dependence from store to load is costly.
20994
20995           store_to_load
20996               Any dependence from store to load is costly.
20997
20998           number
20999               Any dependence for which the latency is greater than or equal
21000               to number is costly.
21001
21002       -minsert-sched-nops=scheme
21003           This option controls which NOP insertion scheme is used during the
21004           second scheduling pass.  The argument scheme takes one of the
21005           following values:
21006
21007           no  Don't insert NOPs.
21008
21009           pad Pad with NOPs any dispatch group that has vacant issue slots,
21010               according to the scheduler's grouping.
21011
21012           regroup_exact
21013               Insert NOPs to force costly dependent insns into separate
21014               groups.  Insert exactly as many NOPs as needed to force an insn
21015               to a new group, according to the estimated processor grouping.
21016
21017           number
21018               Insert NOPs to force costly dependent insns into separate
21019               groups.  Insert number NOPs to force an insn to a new group.
21020
21021       -mcall-sysv
21022           On System V.4 and embedded PowerPC systems compile code using
21023           calling conventions that adhere to the March 1995 draft of the
21024           System V Application Binary Interface, PowerPC processor
21025           supplement.  This is the default unless you configured GCC using
21026           powerpc-*-eabiaix.
21027
21028       -mcall-sysv-eabi
21029       -mcall-eabi
21030           Specify both -mcall-sysv and -meabi options.
21031
21032       -mcall-sysv-noeabi
21033           Specify both -mcall-sysv and -mno-eabi options.
21034
21035       -mcall-aixdesc
21036           On System V.4 and embedded PowerPC systems compile code for the AIX
21037           operating system.
21038
21039       -mcall-linux
21040           On System V.4 and embedded PowerPC systems compile code for the
21041           Linux-based GNU system.
21042
21043       -mcall-freebsd
21044           On System V.4 and embedded PowerPC systems compile code for the
21045           FreeBSD operating system.
21046
21047       -mcall-netbsd
21048           On System V.4 and embedded PowerPC systems compile code for the
21049           NetBSD operating system.
21050
21051       -mcall-openbsd
21052           On System V.4 and embedded PowerPC systems compile code for the
21053           OpenBSD operating system.
21054
21055       -mtraceback=traceback_type
21056           Select the type of traceback table. Valid values for traceback_type
21057           are full, part, and no.
21058
21059       -maix-struct-return
21060           Return all structures in memory (as specified by the AIX ABI).
21061
21062       -msvr4-struct-return
21063           Return structures smaller than 8 bytes in registers (as specified
21064           by the SVR4 ABI).
21065
21066       -mabi=abi-type
21067           Extend the current ABI with a particular extension, or remove such
21068           extension.  Valid values are altivec, no-altivec, ibmlongdouble,
21069           ieeelongdouble, elfv1, elfv2.
21070
21071       -mabi=ibmlongdouble
21072           Change the current ABI to use IBM extended-precision long double.
21073           This is not likely to work if your system defaults to using IEEE
21074           extended-precision long double.  If you change the long double type
21075           from IEEE extended-precision, the compiler will issue a warning
21076           unless you use the -Wno-psabi option.  Requires -mlong-double-128
21077           to be enabled.
21078
21079       -mabi=ieeelongdouble
21080           Change the current ABI to use IEEE extended-precision long double.
21081           This is not likely to work if your system defaults to using IBM
21082           extended-precision long double.  If you change the long double type
21083           from IBM extended-precision, the compiler will issue a warning
21084           unless you use the -Wno-psabi option.  Requires -mlong-double-128
21085           to be enabled.
21086
21087       -mabi=elfv1
21088           Change the current ABI to use the ELFv1 ABI.  This is the default
21089           ABI for big-endian PowerPC 64-bit Linux.  Overriding the default
21090           ABI requires special system support and is likely to fail in
21091           spectacular ways.
21092
21093       -mabi=elfv2
21094           Change the current ABI to use the ELFv2 ABI.  This is the default
21095           ABI for little-endian PowerPC 64-bit Linux.  Overriding the default
21096           ABI requires special system support and is likely to fail in
21097           spectacular ways.
21098
21099       -mgnu-attribute
21100       -mno-gnu-attribute
21101           Emit .gnu_attribute assembly directives to set tag/value pairs in a
21102           .gnu.attributes section that specify ABI variations in function
21103           parameters or return values.
21104
21105       -mprototype
21106       -mno-prototype
21107           On System V.4 and embedded PowerPC systems assume that all calls to
21108           variable argument functions are properly prototyped.  Otherwise,
21109           the compiler must insert an instruction before every non-prototyped
21110           call to set or clear bit 6 of the condition code register ("CR") to
21111           indicate whether floating-point values are passed in the floating-
21112           point registers in case the function takes variable arguments.
21113           With -mprototype, only calls to prototyped variable argument
21114           functions set or clear the bit.
21115
21116       -msim
21117           On embedded PowerPC systems, assume that the startup module is
21118           called sim-crt0.o and that the standard C libraries are libsim.a
21119           and libc.a.  This is the default for powerpc-*-eabisim
21120           configurations.
21121
21122       -mmvme
21123           On embedded PowerPC systems, assume that the startup module is
21124           called crt0.o and the standard C libraries are libmvme.a and
21125           libc.a.
21126
21127       -mads
21128           On embedded PowerPC systems, assume that the startup module is
21129           called crt0.o and the standard C libraries are libads.a and libc.a.
21130
21131       -myellowknife
21132           On embedded PowerPC systems, assume that the startup module is
21133           called crt0.o and the standard C libraries are libyk.a and libc.a.
21134
21135       -mvxworks
21136           On System V.4 and embedded PowerPC systems, specify that you are
21137           compiling for a VxWorks system.
21138
21139       -memb
21140           On embedded PowerPC systems, set the "PPC_EMB" bit in the ELF flags
21141           header to indicate that eabi extended relocations are used.
21142
21143       -meabi
21144       -mno-eabi
21145           On System V.4 and embedded PowerPC systems do (do not) adhere to
21146           the Embedded Applications Binary Interface (EABI), which is a set
21147           of modifications to the System V.4 specifications.  Selecting
21148           -meabi means that the stack is aligned to an 8-byte boundary, a
21149           function "__eabi" is called from "main" to set up the EABI
21150           environment, and the -msdata option can use both "r2" and "r13" to
21151           point to two separate small data areas.  Selecting -mno-eabi means
21152           that the stack is aligned to a 16-byte boundary, no EABI
21153           initialization function is called from "main", and the -msdata
21154           option only uses "r13" to point to a single small data area.  The
21155           -meabi option is on by default if you configured GCC using one of
21156           the powerpc*-*-eabi* options.
21157
21158       -msdata=eabi
21159           On System V.4 and embedded PowerPC systems, put small initialized
21160           "const" global and static data in the ".sdata2" section, which is
21161           pointed to by register "r2".  Put small initialized non-"const"
21162           global and static data in the ".sdata" section, which is pointed to
21163           by register "r13".  Put small uninitialized global and static data
21164           in the ".sbss" section, which is adjacent to the ".sdata" section.
21165           The -msdata=eabi option is incompatible with the -mrelocatable
21166           option.  The -msdata=eabi option also sets the -memb option.
21167
21168       -msdata=sysv
21169           On System V.4 and embedded PowerPC systems, put small global and
21170           static data in the ".sdata" section, which is pointed to by
21171           register "r13".  Put small uninitialized global and static data in
21172           the ".sbss" section, which is adjacent to the ".sdata" section.
21173           The -msdata=sysv option is incompatible with the -mrelocatable
21174           option.
21175
21176       -msdata=default
21177       -msdata
21178           On System V.4 and embedded PowerPC systems, if -meabi is used,
21179           compile code the same as -msdata=eabi, otherwise compile code the
21180           same as -msdata=sysv.
21181
21182       -msdata=data
21183           On System V.4 and embedded PowerPC systems, put small global data
21184           in the ".sdata" section.  Put small uninitialized global data in
21185           the ".sbss" section.  Do not use register "r13" to address small
21186           data however.  This is the default behavior unless other -msdata
21187           options are used.
21188
21189       -msdata=none
21190       -mno-sdata
21191           On embedded PowerPC systems, put all initialized global and static
21192           data in the ".data" section, and all uninitialized data in the
21193           ".bss" section.
21194
21195       -mreadonly-in-sdata
21196           Put read-only objects in the ".sdata" section as well.  This is the
21197           default.
21198
21199       -mblock-move-inline-limit=num
21200           Inline all block moves (such as calls to "memcpy" or structure
21201           copies) less than or equal to num bytes.  The minimum value for num
21202           is 32 bytes on 32-bit targets and 64 bytes on 64-bit targets.  The
21203           default value is target-specific.
21204
21205       -mblock-compare-inline-limit=num
21206           Generate non-looping inline code for all block compares (such as
21207           calls to "memcmp" or structure compares) less than or equal to num
21208           bytes. If num is 0, all inline expansion (non-loop and loop) of
21209           block compare is disabled. The default value is target-specific.
21210
21211       -mblock-compare-inline-loop-limit=num
21212           Generate an inline expansion using loop code for all block compares
21213           that are less than or equal to num bytes, but greater than the
21214           limit for non-loop inline block compare expansion. If the block
21215           length is not constant, at most num bytes will be compared before
21216           "memcmp" is called to compare the remainder of the block. The
21217           default value is target-specific.
21218
21219       -mstring-compare-inline-limit=num
21220           Compare at most num string bytes with inline code.  If the
21221           difference or end of string is not found at the end of the inline
21222           compare a call to "strcmp" or "strncmp" will take care of the rest
21223           of the comparison. The default is 64 bytes.
21224
21225       -G num
21226           On embedded PowerPC systems, put global and static items less than
21227           or equal to num bytes into the small data or BSS sections instead
21228           of the normal data or BSS section.  By default, num is 8.  The -G
21229           num switch is also passed to the linker.  All modules should be
21230           compiled with the same -G num value.
21231
21232       -mregnames
21233       -mno-regnames
21234           On System V.4 and embedded PowerPC systems do (do not) emit
21235           register names in the assembly language output using symbolic
21236           forms.
21237
21238       -mlongcall
21239       -mno-longcall
21240           By default assume that all calls are far away so that a longer and
21241           more expensive calling sequence is required.  This is required for
21242           calls farther than 32 megabytes (33,554,432 bytes) from the current
21243           location.  A short call is generated if the compiler knows the call
21244           cannot be that far away.  This setting can be overridden by the
21245           "shortcall" function attribute, or by "#pragma longcall(0)".
21246
21247           Some linkers are capable of detecting out-of-range calls and
21248           generating glue code on the fly.  On these systems, long calls are
21249           unnecessary and generate slower code.  As of this writing, the AIX
21250           linker can do this, as can the GNU linker for PowerPC/64.  It is
21251           planned to add this feature to the GNU linker for 32-bit PowerPC
21252           systems as well.
21253
21254           On PowerPC64 ELFv2 and 32-bit PowerPC systems with newer GNU
21255           linkers, GCC can generate long calls using an inline PLT call
21256           sequence (see -mpltseq).  PowerPC with -mbss-plt and PowerPC64
21257           ELFv1 (big-endian) do not support inline PLT calls.
21258
21259           On Darwin/PPC systems, "#pragma longcall" generates "jbsr callee,
21260           L42", plus a branch island (glue code).  The two target addresses
21261           represent the callee and the branch island.  The Darwin/PPC linker
21262           prefers the first address and generates a "bl callee" if the PPC
21263           "bl" instruction reaches the callee directly; otherwise, the linker
21264           generates "bl L42" to call the branch island.  The branch island is
21265           appended to the body of the calling function; it computes the full
21266           32-bit address of the callee and jumps to it.
21267
21268           On Mach-O (Darwin) systems, this option directs the compiler emit
21269           to the glue for every direct call, and the Darwin linker decides
21270           whether to use or discard it.
21271
21272           In the future, GCC may ignore all longcall specifications when the
21273           linker is known to generate glue.
21274
21275       -mpltseq
21276       -mno-pltseq
21277           Implement (do not implement) -fno-plt and long calls using an
21278           inline PLT call sequence that supports lazy linking and long calls
21279           to functions in dlopen'd shared libraries.  Inline PLT calls are
21280           only supported on PowerPC64 ELFv2 and 32-bit PowerPC systems with
21281           newer GNU linkers, and are enabled by default if the support is
21282           detected when configuring GCC, and, in the case of 32-bit PowerPC,
21283           if GCC is configured with --enable-secureplt.  -mpltseq code and
21284           -mbss-plt 32-bit PowerPC relocatable objects may not be linked
21285           together.
21286
21287       -mtls-markers
21288       -mno-tls-markers
21289           Mark (do not mark) calls to "__tls_get_addr" with a relocation
21290           specifying the function argument.  The relocation allows the linker
21291           to reliably associate function call with argument setup
21292           instructions for TLS optimization, which in turn allows GCC to
21293           better schedule the sequence.
21294
21295       -mrecip
21296       -mno-recip
21297           This option enables use of the reciprocal estimate and reciprocal
21298           square root estimate instructions with additional Newton-Raphson
21299           steps to increase precision instead of doing a divide or square
21300           root and divide for floating-point arguments.  You should use the
21301           -ffast-math option when using -mrecip (or at least
21302           -funsafe-math-optimizations, -ffinite-math-only, -freciprocal-math
21303           and -fno-trapping-math).  Note that while the throughput of the
21304           sequence is generally higher than the throughput of the non-
21305           reciprocal instruction, the precision of the sequence can be
21306           decreased by up to 2 ulp (i.e. the inverse of 1.0 equals
21307           0.99999994) for reciprocal square roots.
21308
21309       -mrecip=opt
21310           This option controls which reciprocal estimate instructions may be
21311           used.  opt is a comma-separated list of options, which may be
21312           preceded by a "!" to invert the option:
21313
21314           all Enable all estimate instructions.
21315
21316           default
21317               Enable the default instructions, equivalent to -mrecip.
21318
21319           none
21320               Disable all estimate instructions, equivalent to -mno-recip.
21321
21322           div Enable the reciprocal approximation instructions for both
21323               single and double precision.
21324
21325           divf
21326               Enable the single-precision reciprocal approximation
21327               instructions.
21328
21329           divd
21330               Enable the double-precision reciprocal approximation
21331               instructions.
21332
21333           rsqrt
21334               Enable the reciprocal square root approximation instructions
21335               for both single and double precision.
21336
21337           rsqrtf
21338               Enable the single-precision reciprocal square root
21339               approximation instructions.
21340
21341           rsqrtd
21342               Enable the double-precision reciprocal square root
21343               approximation instructions.
21344
21345           So, for example, -mrecip=all,!rsqrtd enables all of the reciprocal
21346           estimate instructions, except for the "FRSQRTE", "XSRSQRTEDP", and
21347           "XVRSQRTEDP" instructions which handle the double-precision
21348           reciprocal square root calculations.
21349
21350       -mrecip-precision
21351       -mno-recip-precision
21352           Assume (do not assume) that the reciprocal estimate instructions
21353           provide higher-precision estimates than is mandated by the PowerPC
21354           ABI.  Selecting -mcpu=power6, -mcpu=power7 or -mcpu=power8
21355           automatically selects -mrecip-precision.  The double-precision
21356           square root estimate instructions are not generated by default on
21357           low-precision machines, since they do not provide an estimate that
21358           converges after three steps.
21359
21360       -mveclibabi=type
21361           Specifies the ABI type to use for vectorizing intrinsics using an
21362           external library.  The only type supported at present is mass,
21363           which specifies to use IBM's Mathematical Acceleration Subsystem
21364           (MASS) libraries for vectorizing intrinsics using external
21365           libraries.  GCC currently emits calls to "acosd2", "acosf4",
21366           "acoshd2", "acoshf4", "asind2", "asinf4", "asinhd2", "asinhf4",
21367           "atan2d2", "atan2f4", "atand2", "atanf4", "atanhd2", "atanhf4",
21368           "cbrtd2", "cbrtf4", "cosd2", "cosf4", "coshd2", "coshf4", "erfcd2",
21369           "erfcf4", "erfd2", "erff4", "exp2d2", "exp2f4", "expd2", "expf4",
21370           "expm1d2", "expm1f4", "hypotd2", "hypotf4", "lgammad2", "lgammaf4",
21371           "log10d2", "log10f4", "log1pd2", "log1pf4", "log2d2", "log2f4",
21372           "logd2", "logf4", "powd2", "powf4", "sind2", "sinf4", "sinhd2",
21373           "sinhf4", "sqrtd2", "sqrtf4", "tand2", "tanf4", "tanhd2", and
21374           "tanhf4" when generating code for power7.  Both -ftree-vectorize
21375           and -funsafe-math-optimizations must also be enabled.  The MASS
21376           libraries must be specified at link time.
21377
21378       -mfriz
21379       -mno-friz
21380           Generate (do not generate) the "friz" instruction when the
21381           -funsafe-math-optimizations option is used to optimize rounding of
21382           floating-point values to 64-bit integer and back to floating point.
21383           The "friz" instruction does not return the same value if the
21384           floating-point number is too large to fit in an integer.
21385
21386       -mpointers-to-nested-functions
21387       -mno-pointers-to-nested-functions
21388           Generate (do not generate) code to load up the static chain
21389           register ("r11") when calling through a pointer on AIX and 64-bit
21390           Linux systems where a function pointer points to a 3-word
21391           descriptor giving the function address, TOC value to be loaded in
21392           register "r2", and static chain value to be loaded in register
21393           "r11".  The -mpointers-to-nested-functions is on by default.  You
21394           cannot call through pointers to nested functions or pointers to
21395           functions compiled in other languages that use the static chain if
21396           you use -mno-pointers-to-nested-functions.
21397
21398       -msave-toc-indirect
21399       -mno-save-toc-indirect
21400           Generate (do not generate) code to save the TOC value in the
21401           reserved stack location in the function prologue if the function
21402           calls through a pointer on AIX and 64-bit Linux systems.  If the
21403           TOC value is not saved in the prologue, it is saved just before the
21404           call through the pointer.  The -mno-save-toc-indirect option is the
21405           default.
21406
21407       -mcompat-align-parm
21408       -mno-compat-align-parm
21409           Generate (do not generate) code to pass structure parameters with a
21410           maximum alignment of 64 bits, for compatibility with older versions
21411           of GCC.
21412
21413           Older versions of GCC (prior to 4.9.0) incorrectly did not align a
21414           structure parameter on a 128-bit boundary when that structure
21415           contained a member requiring 128-bit alignment.  This is corrected
21416           in more recent versions of GCC.  This option may be used to
21417           generate code that is compatible with functions compiled with older
21418           versions of GCC.
21419
21420           The -mno-compat-align-parm option is the default.
21421
21422       -mstack-protector-guard=guard
21423       -mstack-protector-guard-reg=reg
21424       -mstack-protector-guard-offset=offset
21425       -mstack-protector-guard-symbol=symbol
21426           Generate stack protection code using canary at guard.  Supported
21427           locations are global for global canary or tls for per-thread canary
21428           in the TLS block (the default with GNU libc version 2.4 or later).
21429
21430           With the latter choice the options -mstack-protector-guard-reg=reg
21431           and -mstack-protector-guard-offset=offset furthermore specify which
21432           register to use as base register for reading the canary, and from
21433           what offset from that base register. The default for those is as
21434           specified in the relevant ABI.
21435           -mstack-protector-guard-symbol=symbol overrides the offset with a
21436           symbol reference to a canary in the TLS block.
21437
21438       RX Options
21439
21440       These command-line options are defined for RX targets:
21441
21442       -m64bit-doubles
21443       -m32bit-doubles
21444           Make the "double" data type be 64 bits (-m64bit-doubles) or 32 bits
21445           (-m32bit-doubles) in size.  The default is -m32bit-doubles.  Note
21446           RX floating-point hardware only works on 32-bit values, which is
21447           why the default is -m32bit-doubles.
21448
21449       -fpu
21450       -nofpu
21451           Enables (-fpu) or disables (-nofpu) the use of RX floating-point
21452           hardware.  The default is enabled for the RX600 series and disabled
21453           for the RX200 series.
21454
21455           Floating-point instructions are only generated for 32-bit floating-
21456           point values, however, so the FPU hardware is not used for doubles
21457           if the -m64bit-doubles option is used.
21458
21459           Note If the -fpu option is enabled then -funsafe-math-optimizations
21460           is also enabled automatically.  This is because the RX FPU
21461           instructions are themselves unsafe.
21462
21463       -mcpu=name
21464           Selects the type of RX CPU to be targeted.  Currently three types
21465           are supported, the generic RX600 and RX200 series hardware and the
21466           specific RX610 CPU.  The default is RX600.
21467
21468           The only difference between RX600 and RX610 is that the RX610 does
21469           not support the "MVTIPL" instruction.
21470
21471           The RX200 series does not have a hardware floating-point unit and
21472           so -nofpu is enabled by default when this type is selected.
21473
21474       -mbig-endian-data
21475       -mlittle-endian-data
21476           Store data (but not code) in the big-endian format.  The default is
21477           -mlittle-endian-data, i.e. to store data in the little-endian
21478           format.
21479
21480       -msmall-data-limit=N
21481           Specifies the maximum size in bytes of global and static variables
21482           which can be placed into the small data area.  Using the small data
21483           area can lead to smaller and faster code, but the size of area is
21484           limited and it is up to the programmer to ensure that the area does
21485           not overflow.  Also when the small data area is used one of the
21486           RX's registers (usually "r13") is reserved for use pointing to this
21487           area, so it is no longer available for use by the compiler.  This
21488           could result in slower and/or larger code if variables are pushed
21489           onto the stack instead of being held in this register.
21490
21491           Note, common variables (variables that have not been initialized)
21492           and constants are not placed into the small data area as they are
21493           assigned to other sections in the output executable.
21494
21495           The default value is zero, which disables this feature.  Note, this
21496           feature is not enabled by default with higher optimization levels
21497           (-O2 etc) because of the potentially detrimental effects of
21498           reserving a register.  It is up to the programmer to experiment and
21499           discover whether this feature is of benefit to their program.  See
21500           the description of the -mpid option for a description of how the
21501           actual register to hold the small data area pointer is chosen.
21502
21503       -msim
21504       -mno-sim
21505           Use the simulator runtime.  The default is to use the libgloss
21506           board-specific runtime.
21507
21508       -mas100-syntax
21509       -mno-as100-syntax
21510           When generating assembler output use a syntax that is compatible
21511           with Renesas's AS100 assembler.  This syntax can also be handled by
21512           the GAS assembler, but it has some restrictions so it is not
21513           generated by default.
21514
21515       -mmax-constant-size=N
21516           Specifies the maximum size, in bytes, of a constant that can be
21517           used as an operand in a RX instruction.  Although the RX
21518           instruction set does allow constants of up to 4 bytes in length to
21519           be used in instructions, a longer value equates to a longer
21520           instruction.  Thus in some circumstances it can be beneficial to
21521           restrict the size of constants that are used in instructions.
21522           Constants that are too big are instead placed into a constant pool
21523           and referenced via register indirection.
21524
21525           The value N can be between 0 and 4.  A value of 0 (the default) or
21526           4 means that constants of any size are allowed.
21527
21528       -mrelax
21529           Enable linker relaxation.  Linker relaxation is a process whereby
21530           the linker attempts to reduce the size of a program by finding
21531           shorter versions of various instructions.  Disabled by default.
21532
21533       -mint-register=N
21534           Specify the number of registers to reserve for fast interrupt
21535           handler functions.  The value N can be between 0 and 4.  A value of
21536           1 means that register "r13" is reserved for the exclusive use of
21537           fast interrupt handlers.  A value of 2 reserves "r13" and "r12".  A
21538           value of 3 reserves "r13", "r12" and "r11", and a value of 4
21539           reserves "r13" through "r10".  A value of 0, the default, does not
21540           reserve any registers.
21541
21542       -msave-acc-in-interrupts
21543           Specifies that interrupt handler functions should preserve the
21544           accumulator register.  This is only necessary if normal code might
21545           use the accumulator register, for example because it performs
21546           64-bit multiplications.  The default is to ignore the accumulator
21547           as this makes the interrupt handlers faster.
21548
21549       -mpid
21550       -mno-pid
21551           Enables the generation of position independent data.  When enabled
21552           any access to constant data is done via an offset from a base
21553           address held in a register.  This allows the location of constant
21554           data to be determined at run time without requiring the executable
21555           to be relocated, which is a benefit to embedded applications with
21556           tight memory constraints.  Data that can be modified is not
21557           affected by this option.
21558
21559           Note, using this feature reserves a register, usually "r13", for
21560           the constant data base address.  This can result in slower and/or
21561           larger code, especially in complicated functions.
21562
21563           The actual register chosen to hold the constant data base address
21564           depends upon whether the -msmall-data-limit and/or the
21565           -mint-register command-line options are enabled.  Starting with
21566           register "r13" and proceeding downwards, registers are allocated
21567           first to satisfy the requirements of -mint-register, then -mpid and
21568           finally -msmall-data-limit.  Thus it is possible for the small data
21569           area register to be "r8" if both -mint-register=4 and -mpid are
21570           specified on the command line.
21571
21572           By default this feature is not enabled.  The default can be
21573           restored via the -mno-pid command-line option.
21574
21575       -mno-warn-multiple-fast-interrupts
21576       -mwarn-multiple-fast-interrupts
21577           Prevents GCC from issuing a warning message if it finds more than
21578           one fast interrupt handler when it is compiling a file.  The
21579           default is to issue a warning for each extra fast interrupt handler
21580           found, as the RX only supports one such interrupt.
21581
21582       -mallow-string-insns
21583       -mno-allow-string-insns
21584           Enables or disables the use of the string manipulation instructions
21585           "SMOVF", "SCMPU", "SMOVB", "SMOVU", "SUNTIL" "SWHILE" and also the
21586           "RMPA" instruction.  These instructions may prefetch data, which is
21587           not safe to do if accessing an I/O register.  (See section 12.2.7
21588           of the RX62N Group User's Manual for more information).
21589
21590           The default is to allow these instructions, but it is not possible
21591           for GCC to reliably detect all circumstances where a string
21592           instruction might be used to access an I/O register, so their use
21593           cannot be disabled automatically.  Instead it is reliant upon the
21594           programmer to use the -mno-allow-string-insns option if their
21595           program accesses I/O space.
21596
21597           When the instructions are enabled GCC defines the C preprocessor
21598           symbol "__RX_ALLOW_STRING_INSNS__", otherwise it defines the symbol
21599           "__RX_DISALLOW_STRING_INSNS__".
21600
21601       -mjsr
21602       -mno-jsr
21603           Use only (or not only) "JSR" instructions to access functions.
21604           This option can be used when code size exceeds the range of "BSR"
21605           instructions.  Note that -mno-jsr does not mean to not use "JSR"
21606           but instead means that any type of branch may be used.
21607
21608       Note: The generic GCC command-line option -ffixed-reg has special
21609       significance to the RX port when used with the "interrupt" function
21610       attribute.  This attribute indicates a function intended to process
21611       fast interrupts.  GCC ensures that it only uses the registers "r10",
21612       "r11", "r12" and/or "r13" and only provided that the normal use of the
21613       corresponding registers have been restricted via the -ffixed-reg or
21614       -mint-register command-line options.
21615
21616       S/390 and zSeries Options
21617
21618       These are the -m options defined for the S/390 and zSeries
21619       architecture.
21620
21621       -mhard-float
21622       -msoft-float
21623           Use (do not use) the hardware floating-point instructions and
21624           registers for floating-point operations.  When -msoft-float is
21625           specified, functions in libgcc.a are used to perform floating-point
21626           operations.  When -mhard-float is specified, the compiler generates
21627           IEEE floating-point instructions.  This is the default.
21628
21629       -mhard-dfp
21630       -mno-hard-dfp
21631           Use (do not use) the hardware decimal-floating-point instructions
21632           for decimal-floating-point operations.  When -mno-hard-dfp is
21633           specified, functions in libgcc.a are used to perform decimal-
21634           floating-point operations.  When -mhard-dfp is specified, the
21635           compiler generates decimal-floating-point hardware instructions.
21636           This is the default for -march=z9-ec or higher.
21637
21638       -mlong-double-64
21639       -mlong-double-128
21640           These switches control the size of "long double" type. A size of 64
21641           bits makes the "long double" type equivalent to the "double" type.
21642           This is the default.
21643
21644       -mbackchain
21645       -mno-backchain
21646           Store (do not store) the address of the caller's frame as backchain
21647           pointer into the callee's stack frame.  A backchain may be needed
21648           to allow debugging using tools that do not understand DWARF call
21649           frame information.  When -mno-packed-stack is in effect, the
21650           backchain pointer is stored at the bottom of the stack frame; when
21651           -mpacked-stack is in effect, the backchain is placed into the
21652           topmost word of the 96/160 byte register save area.
21653
21654           In general, code compiled with -mbackchain is call-compatible with
21655           code compiled with -mmo-backchain; however, use of the backchain
21656           for debugging purposes usually requires that the whole binary is
21657           built with -mbackchain.  Note that the combination of -mbackchain,
21658           -mpacked-stack and -mhard-float is not supported.  In order to
21659           build a linux kernel use -msoft-float.
21660
21661           The default is to not maintain the backchain.
21662
21663       -mpacked-stack
21664       -mno-packed-stack
21665           Use (do not use) the packed stack layout.  When -mno-packed-stack
21666           is specified, the compiler uses the all fields of the 96/160 byte
21667           register save area only for their default purpose; unused fields
21668           still take up stack space.  When -mpacked-stack is specified,
21669           register save slots are densely packed at the top of the register
21670           save area; unused space is reused for other purposes, allowing for
21671           more efficient use of the available stack space.  However, when
21672           -mbackchain is also in effect, the topmost word of the save area is
21673           always used to store the backchain, and the return address register
21674           is always saved two words below the backchain.
21675
21676           As long as the stack frame backchain is not used, code generated
21677           with -mpacked-stack is call-compatible with code generated with
21678           -mno-packed-stack.  Note that some non-FSF releases of GCC 2.95 for
21679           S/390 or zSeries generated code that uses the stack frame backchain
21680           at run time, not just for debugging purposes.  Such code is not
21681           call-compatible with code compiled with -mpacked-stack.  Also, note
21682           that the combination of -mbackchain, -mpacked-stack and
21683           -mhard-float is not supported.  In order to build a linux kernel
21684           use -msoft-float.
21685
21686           The default is to not use the packed stack layout.
21687
21688       -msmall-exec
21689       -mno-small-exec
21690           Generate (or do not generate) code using the "bras" instruction to
21691           do subroutine calls.  This only works reliably if the total
21692           executable size does not exceed 64k.  The default is to use the
21693           "basr" instruction instead, which does not have this limitation.
21694
21695       -m64
21696       -m31
21697           When -m31 is specified, generate code compliant to the GNU/Linux
21698           for S/390 ABI.  When -m64 is specified, generate code compliant to
21699           the GNU/Linux for zSeries ABI.  This allows GCC in particular to
21700           generate 64-bit instructions.  For the s390 targets, the default is
21701           -m31, while the s390x targets default to -m64.
21702
21703       -mzarch
21704       -mesa
21705           When -mzarch is specified, generate code using the instructions
21706           available on z/Architecture.  When -mesa is specified, generate
21707           code using the instructions available on ESA/390.  Note that -mesa
21708           is not possible with -m64.  When generating code compliant to the
21709           GNU/Linux for S/390 ABI, the default is -mesa.  When generating
21710           code compliant to the GNU/Linux for zSeries ABI, the default is
21711           -mzarch.
21712
21713       -mhtm
21714       -mno-htm
21715           The -mhtm option enables a set of builtins making use of
21716           instructions available with the transactional execution facility
21717           introduced with the IBM zEnterprise EC12 machine generation S/390
21718           System z Built-in Functions.  -mhtm is enabled by default when
21719           using -march=zEC12.
21720
21721       -mvx
21722       -mno-vx
21723           When -mvx is specified, generate code using the instructions
21724           available with the vector extension facility introduced with the
21725           IBM z13 machine generation.  This option changes the ABI for some
21726           vector type values with regard to alignment and calling
21727           conventions.  In case vector type values are being used in an ABI-
21728           relevant context a GAS .gnu_attribute command will be added to mark
21729           the resulting binary with the ABI used.  -mvx is enabled by default
21730           when using -march=z13.
21731
21732       -mzvector
21733       -mno-zvector
21734           The -mzvector option enables vector language extensions and
21735           builtins using instructions available with the vector extension
21736           facility introduced with the IBM z13 machine generation.  This
21737           option adds support for vector to be used as a keyword to define
21738           vector type variables and arguments.  vector is only available when
21739           GNU extensions are enabled.  It will not be expanded when
21740           requesting strict standard compliance e.g. with -std=c99.  In
21741           addition to the GCC low-level builtins -mzvector enables a set of
21742           builtins added for compatibility with AltiVec-style implementations
21743           like Power and Cell.  In order to make use of these builtins the
21744           header file vecintrin.h needs to be included.  -mzvector is
21745           disabled by default.
21746
21747       -mmvcle
21748       -mno-mvcle
21749           Generate (or do not generate) code using the "mvcle" instruction to
21750           perform block moves.  When -mno-mvcle is specified, use a "mvc"
21751           loop instead.  This is the default unless optimizing for size.
21752
21753       -mdebug
21754       -mno-debug
21755           Print (or do not print) additional debug information when
21756           compiling.  The default is to not print debug information.
21757
21758       -march=cpu-type
21759           Generate code that runs on cpu-type, which is the name of a system
21760           representing a certain processor type.  Possible values for cpu-
21761           type are z900/arch5, z990/arch6, z9-109, z9-ec/arch7, z10/arch8,
21762           z196/arch9, zEC12, z13/arch11, z14/arch12, and native.
21763
21764           The default is -march=z900.
21765
21766           Specifying native as cpu type can be used to select the best
21767           architecture option for the host processor.  -march=native has no
21768           effect if GCC does not recognize the processor.
21769
21770       -mtune=cpu-type
21771           Tune to cpu-type everything applicable about the generated code,
21772           except for the ABI and the set of available instructions.  The list
21773           of cpu-type values is the same as for -march.  The default is the
21774           value used for -march.
21775
21776       -mtpf-trace
21777       -mno-tpf-trace
21778           Generate code that adds (does not add) in TPF OS specific branches
21779           to trace routines in the operating system.  This option is off by
21780           default, even when compiling for the TPF OS.
21781
21782       -mfused-madd
21783       -mno-fused-madd
21784           Generate code that uses (does not use) the floating-point multiply
21785           and accumulate instructions.  These instructions are generated by
21786           default if hardware floating point is used.
21787
21788       -mwarn-framesize=framesize
21789           Emit a warning if the current function exceeds the given frame
21790           size.  Because this is a compile-time check it doesn't need to be a
21791           real problem when the program runs.  It is intended to identify
21792           functions that most probably cause a stack overflow.  It is useful
21793           to be used in an environment with limited stack size e.g. the linux
21794           kernel.
21795
21796       -mwarn-dynamicstack
21797           Emit a warning if the function calls "alloca" or uses dynamically-
21798           sized arrays.  This is generally a bad idea with a limited stack
21799           size.
21800
21801       -mstack-guard=stack-guard
21802       -mstack-size=stack-size
21803           If these options are provided the S/390 back end emits additional
21804           instructions in the function prologue that trigger a trap if the
21805           stack size is stack-guard bytes above the stack-size (remember that
21806           the stack on S/390 grows downward).  If the stack-guard option is
21807           omitted the smallest power of 2 larger than the frame size of the
21808           compiled function is chosen.  These options are intended to be used
21809           to help debugging stack overflow problems.  The additionally
21810           emitted code causes only little overhead and hence can also be used
21811           in production-like systems without greater performance degradation.
21812           The given values have to be exact powers of 2 and stack-size has to
21813           be greater than stack-guard without exceeding 64k.  In order to be
21814           efficient the extra code makes the assumption that the stack starts
21815           at an address aligned to the value given by stack-size.  The stack-
21816           guard option can only be used in conjunction with stack-size.
21817
21818       -mhotpatch=pre-halfwords,post-halfwords
21819           If the hotpatch option is enabled, a "hot-patching" function
21820           prologue is generated for all functions in the compilation unit.
21821           The funtion label is prepended with the given number of two-byte
21822           NOP instructions (pre-halfwords, maximum 1000000).  After the
21823           label, 2 * post-halfwords bytes are appended, using the largest NOP
21824           like instructions the architecture allows (maximum 1000000).
21825
21826           If both arguments are zero, hotpatching is disabled.
21827
21828           This option can be overridden for individual functions with the
21829           "hotpatch" attribute.
21830
21831       Score Options
21832
21833       These options are defined for Score implementations:
21834
21835       -meb
21836           Compile code for big-endian mode.  This is the default.
21837
21838       -mel
21839           Compile code for little-endian mode.
21840
21841       -mnhwloop
21842           Disable generation of "bcnz" instructions.
21843
21844       -muls
21845           Enable generation of unaligned load and store instructions.
21846
21847       -mmac
21848           Enable the use of multiply-accumulate instructions. Disabled by
21849           default.
21850
21851       -mscore5
21852           Specify the SCORE5 as the target architecture.
21853
21854       -mscore5u
21855           Specify the SCORE5U of the target architecture.
21856
21857       -mscore7
21858           Specify the SCORE7 as the target architecture. This is the default.
21859
21860       -mscore7d
21861           Specify the SCORE7D as the target architecture.
21862
21863       SH Options
21864
21865       These -m options are defined for the SH implementations:
21866
21867       -m1 Generate code for the SH1.
21868
21869       -m2 Generate code for the SH2.
21870
21871       -m2e
21872           Generate code for the SH2e.
21873
21874       -m2a-nofpu
21875           Generate code for the SH2a without FPU, or for a SH2a-FPU in such a
21876           way that the floating-point unit is not used.
21877
21878       -m2a-single-only
21879           Generate code for the SH2a-FPU, in such a way that no double-
21880           precision floating-point operations are used.
21881
21882       -m2a-single
21883           Generate code for the SH2a-FPU assuming the floating-point unit is
21884           in single-precision mode by default.
21885
21886       -m2a
21887           Generate code for the SH2a-FPU assuming the floating-point unit is
21888           in double-precision mode by default.
21889
21890       -m3 Generate code for the SH3.
21891
21892       -m3e
21893           Generate code for the SH3e.
21894
21895       -m4-nofpu
21896           Generate code for the SH4 without a floating-point unit.
21897
21898       -m4-single-only
21899           Generate code for the SH4 with a floating-point unit that only
21900           supports single-precision arithmetic.
21901
21902       -m4-single
21903           Generate code for the SH4 assuming the floating-point unit is in
21904           single-precision mode by default.
21905
21906       -m4 Generate code for the SH4.
21907
21908       -m4-100
21909           Generate code for SH4-100.
21910
21911       -m4-100-nofpu
21912           Generate code for SH4-100 in such a way that the floating-point
21913           unit is not used.
21914
21915       -m4-100-single
21916           Generate code for SH4-100 assuming the floating-point unit is in
21917           single-precision mode by default.
21918
21919       -m4-100-single-only
21920           Generate code for SH4-100 in such a way that no double-precision
21921           floating-point operations are used.
21922
21923       -m4-200
21924           Generate code for SH4-200.
21925
21926       -m4-200-nofpu
21927           Generate code for SH4-200 without in such a way that the floating-
21928           point unit is not used.
21929
21930       -m4-200-single
21931           Generate code for SH4-200 assuming the floating-point unit is in
21932           single-precision mode by default.
21933
21934       -m4-200-single-only
21935           Generate code for SH4-200 in such a way that no double-precision
21936           floating-point operations are used.
21937
21938       -m4-300
21939           Generate code for SH4-300.
21940
21941       -m4-300-nofpu
21942           Generate code for SH4-300 without in such a way that the floating-
21943           point unit is not used.
21944
21945       -m4-300-single
21946           Generate code for SH4-300 in such a way that no double-precision
21947           floating-point operations are used.
21948
21949       -m4-300-single-only
21950           Generate code for SH4-300 in such a way that no double-precision
21951           floating-point operations are used.
21952
21953       -m4-340
21954           Generate code for SH4-340 (no MMU, no FPU).
21955
21956       -m4-500
21957           Generate code for SH4-500 (no FPU).  Passes -isa=sh4-nofpu to the
21958           assembler.
21959
21960       -m4a-nofpu
21961           Generate code for the SH4al-dsp, or for a SH4a in such a way that
21962           the floating-point unit is not used.
21963
21964       -m4a-single-only
21965           Generate code for the SH4a, in such a way that no double-precision
21966           floating-point operations are used.
21967
21968       -m4a-single
21969           Generate code for the SH4a assuming the floating-point unit is in
21970           single-precision mode by default.
21971
21972       -m4a
21973           Generate code for the SH4a.
21974
21975       -m4al
21976           Same as -m4a-nofpu, except that it implicitly passes -dsp to the
21977           assembler.  GCC doesn't generate any DSP instructions at the
21978           moment.
21979
21980       -mb Compile code for the processor in big-endian mode.
21981
21982       -ml Compile code for the processor in little-endian mode.
21983
21984       -mdalign
21985           Align doubles at 64-bit boundaries.  Note that this changes the
21986           calling conventions, and thus some functions from the standard C
21987           library do not work unless you recompile it first with -mdalign.
21988
21989       -mrelax
21990           Shorten some address references at link time, when possible; uses
21991           the linker option -relax.
21992
21993       -mbigtable
21994           Use 32-bit offsets in "switch" tables.  The default is to use
21995           16-bit offsets.
21996
21997       -mbitops
21998           Enable the use of bit manipulation instructions on SH2A.
21999
22000       -mfmovd
22001           Enable the use of the instruction "fmovd".  Check -mdalign for
22002           alignment constraints.
22003
22004       -mrenesas
22005           Comply with the calling conventions defined by Renesas.
22006
22007       -mno-renesas
22008           Comply with the calling conventions defined for GCC before the
22009           Renesas conventions were available.  This option is the default for
22010           all targets of the SH toolchain.
22011
22012       -mnomacsave
22013           Mark the "MAC" register as call-clobbered, even if -mrenesas is
22014           given.
22015
22016       -mieee
22017       -mno-ieee
22018           Control the IEEE compliance of floating-point comparisons, which
22019           affects the handling of cases where the result of a comparison is
22020           unordered.  By default -mieee is implicitly enabled.  If
22021           -ffinite-math-only is enabled -mno-ieee is implicitly set, which
22022           results in faster floating-point greater-equal and less-equal
22023           comparisons.  The implicit settings can be overridden by specifying
22024           either -mieee or -mno-ieee.
22025
22026       -minline-ic_invalidate
22027           Inline code to invalidate instruction cache entries after setting
22028           up nested function trampolines.  This option has no effect if
22029           -musermode is in effect and the selected code generation option
22030           (e.g. -m4) does not allow the use of the "icbi" instruction.  If
22031           the selected code generation option does not allow the use of the
22032           "icbi" instruction, and -musermode is not in effect, the inlined
22033           code manipulates the instruction cache address array directly with
22034           an associative write.  This not only requires privileged mode at
22035           run time, but it also fails if the cache line had been mapped via
22036           the TLB and has become unmapped.
22037
22038       -misize
22039           Dump instruction size and location in the assembly code.
22040
22041       -mpadstruct
22042           This option is deprecated.  It pads structures to multiple of 4
22043           bytes, which is incompatible with the SH ABI.
22044
22045       -matomic-model=model
22046           Sets the model of atomic operations and additional parameters as a
22047           comma separated list.  For details on the atomic built-in functions
22048           see __atomic Builtins.  The following models and parameters are
22049           supported:
22050
22051           none
22052               Disable compiler generated atomic sequences and emit library
22053               calls for atomic operations.  This is the default if the target
22054               is not "sh*-*-linux*".
22055
22056           soft-gusa
22057               Generate GNU/Linux compatible gUSA software atomic sequences
22058               for the atomic built-in functions.  The generated atomic
22059               sequences require additional support from the
22060               interrupt/exception handling code of the system and are only
22061               suitable for SH3* and SH4* single-core systems.  This option is
22062               enabled by default when the target is "sh*-*-linux*" and SH3*
22063               or SH4*.  When the target is SH4A, this option also partially
22064               utilizes the hardware atomic instructions "movli.l" and
22065               "movco.l" to create more efficient code, unless strict is
22066               specified.
22067
22068           soft-tcb
22069               Generate software atomic sequences that use a variable in the
22070               thread control block.  This is a variation of the gUSA
22071               sequences which can also be used on SH1* and SH2* targets.  The
22072               generated atomic sequences require additional support from the
22073               interrupt/exception handling code of the system and are only
22074               suitable for single-core systems.  When using this model, the
22075               gbr-offset= parameter has to be specified as well.
22076
22077           soft-imask
22078               Generate software atomic sequences that temporarily disable
22079               interrupts by setting "SR.IMASK = 1111".  This model works only
22080               when the program runs in privileged mode and is only suitable
22081               for single-core systems.  Additional support from the
22082               interrupt/exception handling code of the system is not
22083               required.  This model is enabled by default when the target is
22084               "sh*-*-linux*" and SH1* or SH2*.
22085
22086           hard-llcs
22087               Generate hardware atomic sequences using the "movli.l" and
22088               "movco.l" instructions only.  This is only available on SH4A
22089               and is suitable for multi-core systems.  Since the hardware
22090               instructions support only 32 bit atomic variables access to 8
22091               or 16 bit variables is emulated with 32 bit accesses.  Code
22092               compiled with this option is also compatible with other
22093               software atomic model interrupt/exception handling systems if
22094               executed on an SH4A system.  Additional support from the
22095               interrupt/exception handling code of the system is not required
22096               for this model.
22097
22098           gbr-offset=
22099               This parameter specifies the offset in bytes of the variable in
22100               the thread control block structure that should be used by the
22101               generated atomic sequences when the soft-tcb model has been
22102               selected.  For other models this parameter is ignored.  The
22103               specified value must be an integer multiple of four and in the
22104               range 0-1020.
22105
22106           strict
22107               This parameter prevents mixed usage of multiple atomic models,
22108               even if they are compatible, and makes the compiler generate
22109               atomic sequences of the specified model only.
22110
22111       -mtas
22112           Generate the "tas.b" opcode for "__atomic_test_and_set".  Notice
22113           that depending on the particular hardware and software
22114           configuration this can degrade overall performance due to the
22115           operand cache line flushes that are implied by the "tas.b"
22116           instruction.  On multi-core SH4A processors the "tas.b" instruction
22117           must be used with caution since it can result in data corruption
22118           for certain cache configurations.
22119
22120       -mprefergot
22121           When generating position-independent code, emit function calls
22122           using the Global Offset Table instead of the Procedure Linkage
22123           Table.
22124
22125       -musermode
22126       -mno-usermode
22127           Don't allow (allow) the compiler generating privileged mode code.
22128           Specifying -musermode also implies -mno-inline-ic_invalidate if the
22129           inlined code would not work in user mode.  -musermode is the
22130           default when the target is "sh*-*-linux*".  If the target is SH1*
22131           or SH2* -musermode has no effect, since there is no user mode.
22132
22133       -multcost=number
22134           Set the cost to assume for a multiply insn.
22135
22136       -mdiv=strategy
22137           Set the division strategy to be used for integer division
22138           operations.  strategy can be one of:
22139
22140           call-div1
22141               Calls a library function that uses the single-step division
22142               instruction "div1" to perform the operation.  Division by zero
22143               calculates an unspecified result and does not trap.  This is
22144               the default except for SH4, SH2A and SHcompact.
22145
22146           call-fp
22147               Calls a library function that performs the operation in double
22148               precision floating point.  Division by zero causes a floating-
22149               point exception.  This is the default for SHcompact with FPU.
22150               Specifying this for targets that do not have a double precision
22151               FPU defaults to "call-div1".
22152
22153           call-table
22154               Calls a library function that uses a lookup table for small
22155               divisors and the "div1" instruction with case distinction for
22156               larger divisors.  Division by zero calculates an unspecified
22157               result and does not trap.  This is the default for SH4.
22158               Specifying this for targets that do not have dynamic shift
22159               instructions defaults to "call-div1".
22160
22161           When a division strategy has not been specified the default
22162           strategy is selected based on the current target.  For SH2A the
22163           default strategy is to use the "divs" and "divu" instructions
22164           instead of library function calls.
22165
22166       -maccumulate-outgoing-args
22167           Reserve space once for outgoing arguments in the function prologue
22168           rather than around each call.  Generally beneficial for performance
22169           and size.  Also needed for unwinding to avoid changing the stack
22170           frame around conditional code.
22171
22172       -mdivsi3_libfunc=name
22173           Set the name of the library function used for 32-bit signed
22174           division to name.  This only affects the name used in the call
22175           division strategies, and the compiler still expects the same sets
22176           of input/output/clobbered registers as if this option were not
22177           present.
22178
22179       -mfixed-range=register-range
22180           Generate code treating the given register range as fixed registers.
22181           A fixed register is one that the register allocator cannot use.
22182           This is useful when compiling kernel code.  A register range is
22183           specified as two registers separated by a dash.  Multiple register
22184           ranges can be specified separated by a comma.
22185
22186       -mbranch-cost=num
22187           Assume num to be the cost for a branch instruction.  Higher numbers
22188           make the compiler try to generate more branch-free code if
22189           possible.  If not specified the value is selected depending on the
22190           processor type that is being compiled for.
22191
22192       -mzdcbranch
22193       -mno-zdcbranch
22194           Assume (do not assume) that zero displacement conditional branch
22195           instructions "bt" and "bf" are fast.  If -mzdcbranch is specified,
22196           the compiler prefers zero displacement branch code sequences.  This
22197           is enabled by default when generating code for SH4 and SH4A.  It
22198           can be explicitly disabled by specifying -mno-zdcbranch.
22199
22200       -mcbranch-force-delay-slot
22201           Force the usage of delay slots for conditional branches, which
22202           stuffs the delay slot with a "nop" if a suitable instruction cannot
22203           be found.  By default this option is disabled.  It can be enabled
22204           to work around hardware bugs as found in the original SH7055.
22205
22206       -mfused-madd
22207       -mno-fused-madd
22208           Generate code that uses (does not use) the floating-point multiply
22209           and accumulate instructions.  These instructions are generated by
22210           default if hardware floating point is used.  The machine-dependent
22211           -mfused-madd option is now mapped to the machine-independent
22212           -ffp-contract=fast option, and -mno-fused-madd is mapped to
22213           -ffp-contract=off.
22214
22215       -mfsca
22216       -mno-fsca
22217           Allow or disallow the compiler to emit the "fsca" instruction for
22218           sine and cosine approximations.  The option -mfsca must be used in
22219           combination with -funsafe-math-optimizations.  It is enabled by
22220           default when generating code for SH4A.  Using -mno-fsca disables
22221           sine and cosine approximations even if -funsafe-math-optimizations
22222           is in effect.
22223
22224       -mfsrra
22225       -mno-fsrra
22226           Allow or disallow the compiler to emit the "fsrra" instruction for
22227           reciprocal square root approximations.  The option -mfsrra must be
22228           used in combination with -funsafe-math-optimizations and
22229           -ffinite-math-only.  It is enabled by default when generating code
22230           for SH4A.  Using -mno-fsrra disables reciprocal square root
22231           approximations even if -funsafe-math-optimizations and
22232           -ffinite-math-only are in effect.
22233
22234       -mpretend-cmove
22235           Prefer zero-displacement conditional branches for conditional move
22236           instruction patterns.  This can result in faster code on the SH4
22237           processor.
22238
22239       -mfdpic
22240           Generate code using the FDPIC ABI.
22241
22242       Solaris 2 Options
22243
22244       These -m options are supported on Solaris 2:
22245
22246       -mclear-hwcap
22247           -mclear-hwcap tells the compiler to remove the hardware
22248           capabilities generated by the Solaris assembler.  This is only
22249           necessary when object files use ISA extensions not supported by the
22250           current machine, but check at runtime whether or not to use them.
22251
22252       -mimpure-text
22253           -mimpure-text, used in addition to -shared, tells the compiler to
22254           not pass -z text to the linker when linking a shared object.  Using
22255           this option, you can link position-dependent code into a shared
22256           object.
22257
22258           -mimpure-text suppresses the "relocations remain against
22259           allocatable but non-writable sections" linker error message.
22260           However, the necessary relocations trigger copy-on-write, and the
22261           shared object is not actually shared across processes.  Instead of
22262           using -mimpure-text, you should compile all source code with -fpic
22263           or -fPIC.
22264
22265       These switches are supported in addition to the above on Solaris 2:
22266
22267       -pthreads
22268           This is a synonym for -pthread.
22269
22270       SPARC Options
22271
22272       These -m options are supported on the SPARC:
22273
22274       -mno-app-regs
22275       -mapp-regs
22276           Specify -mapp-regs to generate output using the global registers 2
22277           through 4, which the SPARC SVR4 ABI reserves for applications.
22278           Like the global register 1, each global register 2 through 4 is
22279           then treated as an allocable register that is clobbered by function
22280           calls.  This is the default.
22281
22282           To be fully SVR4 ABI-compliant at the cost of some performance
22283           loss, specify -mno-app-regs.  You should compile libraries and
22284           system software with this option.
22285
22286       -mflat
22287       -mno-flat
22288           With -mflat, the compiler does not generate save/restore
22289           instructions and uses a "flat" or single register window model.
22290           This model is compatible with the regular register window model.
22291           The local registers and the input registers (0--5) are still
22292           treated as "call-saved" registers and are saved on the stack as
22293           needed.
22294
22295           With -mno-flat (the default), the compiler generates save/restore
22296           instructions (except for leaf functions).  This is the normal
22297           operating mode.
22298
22299       -mfpu
22300       -mhard-float
22301           Generate output containing floating-point instructions.  This is
22302           the default.
22303
22304       -mno-fpu
22305       -msoft-float
22306           Generate output containing library calls for floating point.
22307           Warning: the requisite libraries are not available for all SPARC
22308           targets.  Normally the facilities of the machine's usual C compiler
22309           are used, but this cannot be done directly in cross-compilation.
22310           You must make your own arrangements to provide suitable library
22311           functions for cross-compilation.  The embedded targets sparc-*-aout
22312           and sparclite-*-* do provide software floating-point support.
22313
22314           -msoft-float changes the calling convention in the output file;
22315           therefore, it is only useful if you compile all of a program with
22316           this option.  In particular, you need to compile libgcc.a, the
22317           library that comes with GCC, with -msoft-float in order for this to
22318           work.
22319
22320       -mhard-quad-float
22321           Generate output containing quad-word (long double) floating-point
22322           instructions.
22323
22324       -msoft-quad-float
22325           Generate output containing library calls for quad-word (long
22326           double) floating-point instructions.  The functions called are
22327           those specified in the SPARC ABI.  This is the default.
22328
22329           As of this writing, there are no SPARC implementations that have
22330           hardware support for the quad-word floating-point instructions.
22331           They all invoke a trap handler for one of these instructions, and
22332           then the trap handler emulates the effect of the instruction.
22333           Because of the trap handler overhead, this is much slower than
22334           calling the ABI library routines.  Thus the -msoft-quad-float
22335           option is the default.
22336
22337       -mno-unaligned-doubles
22338       -munaligned-doubles
22339           Assume that doubles have 8-byte alignment.  This is the default.
22340
22341           With -munaligned-doubles, GCC assumes that doubles have 8-byte
22342           alignment only if they are contained in another type, or if they
22343           have an absolute address.  Otherwise, it assumes they have 4-byte
22344           alignment.  Specifying this option avoids some rare compatibility
22345           problems with code generated by other compilers.  It is not the
22346           default because it results in a performance loss, especially for
22347           floating-point code.
22348
22349       -muser-mode
22350       -mno-user-mode
22351           Do not generate code that can only run in supervisor mode.  This is
22352           relevant only for the "casa" instruction emitted for the LEON3
22353           processor.  This is the default.
22354
22355       -mfaster-structs
22356       -mno-faster-structs
22357           With -mfaster-structs, the compiler assumes that structures should
22358           have 8-byte alignment.  This enables the use of pairs of "ldd" and
22359           "std" instructions for copies in structure assignment, in place of
22360           twice as many "ld" and "st" pairs.  However, the use of this
22361           changed alignment directly violates the SPARC ABI.  Thus, it's
22362           intended only for use on targets where the developer acknowledges
22363           that their resulting code is not directly in line with the rules of
22364           the ABI.
22365
22366       -mstd-struct-return
22367       -mno-std-struct-return
22368           With -mstd-struct-return, the compiler generates checking code in
22369           functions returning structures or unions to detect size mismatches
22370           between the two sides of function calls, as per the 32-bit ABI.
22371
22372           The default is -mno-std-struct-return.  This option has no effect
22373           in 64-bit mode.
22374
22375       -mlra
22376       -mno-lra
22377           Enable Local Register Allocation.  This is the default for SPARC
22378           since GCC 7 so -mno-lra needs to be passed to get old Reload.
22379
22380       -mcpu=cpu_type
22381           Set the instruction set, register set, and instruction scheduling
22382           parameters for machine type cpu_type.  Supported values for
22383           cpu_type are v7, cypress, v8, supersparc, hypersparc, leon, leon3,
22384           leon3v7, sparclite, f930, f934, sparclite86x, sparclet, tsc701, v9,
22385           ultrasparc, ultrasparc3, niagara, niagara2, niagara3, niagara4,
22386           niagara7 and m8.
22387
22388           Native Solaris and GNU/Linux toolchains also support the value
22389           native, which selects the best architecture option for the host
22390           processor.  -mcpu=native has no effect if GCC does not recognize
22391           the processor.
22392
22393           Default instruction scheduling parameters are used for values that
22394           select an architecture and not an implementation.  These are v7,
22395           v8, sparclite, sparclet, v9.
22396
22397           Here is a list of each supported architecture and their supported
22398           implementations.
22399
22400           v7  cypress, leon3v7
22401
22402           v8  supersparc, hypersparc, leon, leon3
22403
22404           sparclite
22405               f930, f934, sparclite86x
22406
22407           sparclet
22408               tsc701
22409
22410           v9  ultrasparc, ultrasparc3, niagara, niagara2, niagara3, niagara4,
22411               niagara7, m8
22412
22413           By default (unless configured otherwise), GCC generates code for
22414           the V7 variant of the SPARC architecture.  With -mcpu=cypress, the
22415           compiler additionally optimizes it for the Cypress CY7C602 chip, as
22416           used in the SPARCStation/SPARCServer 3xx series.  This is also
22417           appropriate for the older SPARCStation 1, 2, IPX etc.
22418
22419           With -mcpu=v8, GCC generates code for the V8 variant of the SPARC
22420           architecture.  The only difference from V7 code is that the
22421           compiler emits the integer multiply and integer divide instructions
22422           which exist in SPARC-V8 but not in SPARC-V7.  With
22423           -mcpu=supersparc, the compiler additionally optimizes it for the
22424           SuperSPARC chip, as used in the SPARCStation 10, 1000 and 2000
22425           series.
22426
22427           With -mcpu=sparclite, GCC generates code for the SPARClite variant
22428           of the SPARC architecture.  This adds the integer multiply, integer
22429           divide step and scan ("ffs") instructions which exist in SPARClite
22430           but not in SPARC-V7.  With -mcpu=f930, the compiler additionally
22431           optimizes it for the Fujitsu MB86930 chip, which is the original
22432           SPARClite, with no FPU.  With -mcpu=f934, the compiler additionally
22433           optimizes it for the Fujitsu MB86934 chip, which is the more recent
22434           SPARClite with FPU.
22435
22436           With -mcpu=sparclet, GCC generates code for the SPARClet variant of
22437           the SPARC architecture.  This adds the integer multiply,
22438           multiply/accumulate, integer divide step and scan ("ffs")
22439           instructions which exist in SPARClet but not in SPARC-V7.  With
22440           -mcpu=tsc701, the compiler additionally optimizes it for the TEMIC
22441           SPARClet chip.
22442
22443           With -mcpu=v9, GCC generates code for the V9 variant of the SPARC
22444           architecture.  This adds 64-bit integer and floating-point move
22445           instructions, 3 additional floating-point condition code registers
22446           and conditional move instructions.  With -mcpu=ultrasparc, the
22447           compiler additionally optimizes it for the Sun UltraSPARC I/II/IIi
22448           chips.  With -mcpu=ultrasparc3, the compiler additionally optimizes
22449           it for the Sun UltraSPARC III/III+/IIIi/IIIi+/IV/IV+ chips.  With
22450           -mcpu=niagara, the compiler additionally optimizes it for Sun
22451           UltraSPARC T1 chips.  With -mcpu=niagara2, the compiler
22452           additionally optimizes it for Sun UltraSPARC T2 chips. With
22453           -mcpu=niagara3, the compiler additionally optimizes it for Sun
22454           UltraSPARC T3 chips.  With -mcpu=niagara4, the compiler
22455           additionally optimizes it for Sun UltraSPARC T4 chips.  With
22456           -mcpu=niagara7, the compiler additionally optimizes it for Oracle
22457           SPARC M7 chips.  With -mcpu=m8, the compiler additionally optimizes
22458           it for Oracle M8 chips.
22459
22460       -mtune=cpu_type
22461           Set the instruction scheduling parameters for machine type
22462           cpu_type, but do not set the instruction set or register set that
22463           the option -mcpu=cpu_type does.
22464
22465           The same values for -mcpu=cpu_type can be used for -mtune=cpu_type,
22466           but the only useful values are those that select a particular CPU
22467           implementation.  Those are cypress, supersparc, hypersparc, leon,
22468           leon3, leon3v7, f930, f934, sparclite86x, tsc701, ultrasparc,
22469           ultrasparc3, niagara, niagara2, niagara3, niagara4, niagara7 and
22470           m8.  With native Solaris and GNU/Linux toolchains, native can also
22471           be used.
22472
22473       -mv8plus
22474       -mno-v8plus
22475           With -mv8plus, GCC generates code for the SPARC-V8+ ABI.  The
22476           difference from the V8 ABI is that the global and out registers are
22477           considered 64 bits wide.  This is enabled by default on Solaris in
22478           32-bit mode for all SPARC-V9 processors.
22479
22480       -mvis
22481       -mno-vis
22482           With -mvis, GCC generates code that takes advantage of the
22483           UltraSPARC Visual Instruction Set extensions.  The default is
22484           -mno-vis.
22485
22486       -mvis2
22487       -mno-vis2
22488           With -mvis2, GCC generates code that takes advantage of version 2.0
22489           of the UltraSPARC Visual Instruction Set extensions.  The default
22490           is -mvis2 when targeting a cpu that supports such instructions,
22491           such as UltraSPARC-III and later.  Setting -mvis2 also sets -mvis.
22492
22493       -mvis3
22494       -mno-vis3
22495           With -mvis3, GCC generates code that takes advantage of version 3.0
22496           of the UltraSPARC Visual Instruction Set extensions.  The default
22497           is -mvis3 when targeting a cpu that supports such instructions,
22498           such as niagara-3 and later.  Setting -mvis3 also sets -mvis2 and
22499           -mvis.
22500
22501       -mvis4
22502       -mno-vis4
22503           With -mvis4, GCC generates code that takes advantage of version 4.0
22504           of the UltraSPARC Visual Instruction Set extensions.  The default
22505           is -mvis4 when targeting a cpu that supports such instructions,
22506           such as niagara-7 and later.  Setting -mvis4 also sets -mvis3,
22507           -mvis2 and -mvis.
22508
22509       -mvis4b
22510       -mno-vis4b
22511           With -mvis4b, GCC generates code that takes advantage of version
22512           4.0 of the UltraSPARC Visual Instruction Set extensions, plus the
22513           additional VIS instructions introduced in the Oracle SPARC
22514           Architecture 2017.  The default is -mvis4b when targeting a cpu
22515           that supports such instructions, such as m8 and later.  Setting
22516           -mvis4b also sets -mvis4, -mvis3, -mvis2 and -mvis.
22517
22518       -mcbcond
22519       -mno-cbcond
22520           With -mcbcond, GCC generates code that takes advantage of the
22521           UltraSPARC Compare-and-Branch-on-Condition instructions.  The
22522           default is -mcbcond when targeting a CPU that supports such
22523           instructions, such as Niagara-4 and later.
22524
22525       -mfmaf
22526       -mno-fmaf
22527           With -mfmaf, GCC generates code that takes advantage of the
22528           UltraSPARC Fused Multiply-Add Floating-point instructions.  The
22529           default is -mfmaf when targeting a CPU that supports such
22530           instructions, such as Niagara-3 and later.
22531
22532       -mfsmuld
22533       -mno-fsmuld
22534           With -mfsmuld, GCC generates code that takes advantage of the
22535           Floating-point Multiply Single to Double (FsMULd) instruction.  The
22536           default is -mfsmuld when targeting a CPU supporting the
22537           architecture versions V8 or V9 with FPU except -mcpu=leon.
22538
22539       -mpopc
22540       -mno-popc
22541           With -mpopc, GCC generates code that takes advantage of the
22542           UltraSPARC Population Count instruction.  The default is -mpopc
22543           when targeting a CPU that supports such an instruction, such as
22544           Niagara-2 and later.
22545
22546       -msubxc
22547       -mno-subxc
22548           With -msubxc, GCC generates code that takes advantage of the
22549           UltraSPARC Subtract-Extended-with-Carry instruction.  The default
22550           is -msubxc when targeting a CPU that supports such an instruction,
22551           such as Niagara-7 and later.
22552
22553       -mfix-at697f
22554           Enable the documented workaround for the single erratum of the
22555           Atmel AT697F processor (which corresponds to erratum #13 of the
22556           AT697E processor).
22557
22558       -mfix-ut699
22559           Enable the documented workarounds for the floating-point errata and
22560           the data cache nullify errata of the UT699 processor.
22561
22562       -mfix-ut700
22563           Enable the documented workaround for the back-to-back store errata
22564           of the UT699E/UT700 processor.
22565
22566       -mfix-gr712rc
22567           Enable the documented workaround for the back-to-back store errata
22568           of the GR712RC processor.
22569
22570       These -m options are supported in addition to the above on SPARC-V9
22571       processors in 64-bit environments:
22572
22573       -m32
22574       -m64
22575           Generate code for a 32-bit or 64-bit environment.  The 32-bit
22576           environment sets int, long and pointer to 32 bits.  The 64-bit
22577           environment sets int to 32 bits and long and pointer to 64 bits.
22578
22579       -mcmodel=which
22580           Set the code model to one of
22581
22582           medlow
22583               The Medium/Low code model: 64-bit addresses, programs must be
22584               linked in the low 32 bits of memory.  Programs can be
22585               statically or dynamically linked.
22586
22587           medmid
22588               The Medium/Middle code model: 64-bit addresses, programs must
22589               be linked in the low 44 bits of memory, the text and data
22590               segments must be less than 2GB in size and the data segment
22591               must be located within 2GB of the text segment.
22592
22593           medany
22594               The Medium/Anywhere code model: 64-bit addresses, programs may
22595               be linked anywhere in memory, the text and data segments must
22596               be less than 2GB in size and the data segment must be located
22597               within 2GB of the text segment.
22598
22599           embmedany
22600               The Medium/Anywhere code model for embedded systems: 64-bit
22601               addresses, the text and data segments must be less than 2GB in
22602               size, both starting anywhere in memory (determined at link
22603               time).  The global register %g4 points to the base of the data
22604               segment.  Programs are statically linked and PIC is not
22605               supported.
22606
22607       -mmemory-model=mem-model
22608           Set the memory model in force on the processor to one of
22609
22610           default
22611               The default memory model for the processor and operating
22612               system.
22613
22614           rmo Relaxed Memory Order
22615
22616           pso Partial Store Order
22617
22618           tso Total Store Order
22619
22620           sc  Sequential Consistency
22621
22622           These memory models are formally defined in Appendix D of the
22623           SPARC-V9 architecture manual, as set in the processor's "PSTATE.MM"
22624           field.
22625
22626       -mstack-bias
22627       -mno-stack-bias
22628           With -mstack-bias, GCC assumes that the stack pointer, and frame
22629           pointer if present, are offset by -2047 which must be added back
22630           when making stack frame references.  This is the default in 64-bit
22631           mode.  Otherwise, assume no such offset is present.
22632
22633       SPU Options
22634
22635       These -m options are supported on the SPU:
22636
22637       -mwarn-reloc
22638       -merror-reloc
22639           The loader for SPU does not handle dynamic relocations.  By
22640           default, GCC gives an error when it generates code that requires a
22641           dynamic relocation.  -mno-error-reloc disables the error,
22642           -mwarn-reloc generates a warning instead.
22643
22644       -msafe-dma
22645       -munsafe-dma
22646           Instructions that initiate or test completion of DMA must not be
22647           reordered with respect to loads and stores of the memory that is
22648           being accessed.  With -munsafe-dma you must use the "volatile"
22649           keyword to protect memory accesses, but that can lead to
22650           inefficient code in places where the memory is known to not change.
22651           Rather than mark the memory as volatile, you can use -msafe-dma to
22652           tell the compiler to treat the DMA instructions as potentially
22653           affecting all memory.
22654
22655       -mbranch-hints
22656           By default, GCC generates a branch hint instruction to avoid
22657           pipeline stalls for always-taken or probably-taken branches.  A
22658           hint is not generated closer than 8 instructions away from its
22659           branch.  There is little reason to disable them, except for
22660           debugging purposes, or to make an object a little bit smaller.
22661
22662       -msmall-mem
22663       -mlarge-mem
22664           By default, GCC generates code assuming that addresses are never
22665           larger than 18 bits.  With -mlarge-mem code is generated that
22666           assumes a full 32-bit address.
22667
22668       -mstdmain
22669           By default, GCC links against startup code that assumes the SPU-
22670           style main function interface (which has an unconventional
22671           parameter list).  With -mstdmain, GCC links your program against
22672           startup code that assumes a C99-style interface to "main",
22673           including a local copy of "argv" strings.
22674
22675       -mfixed-range=register-range
22676           Generate code treating the given register range as fixed registers.
22677           A fixed register is one that the register allocator cannot use.
22678           This is useful when compiling kernel code.  A register range is
22679           specified as two registers separated by a dash.  Multiple register
22680           ranges can be specified separated by a comma.
22681
22682       -mea32
22683       -mea64
22684           Compile code assuming that pointers to the PPU address space
22685           accessed via the "__ea" named address space qualifier are either 32
22686           or 64 bits wide.  The default is 32 bits.  As this is an ABI-
22687           changing option, all object code in an executable must be compiled
22688           with the same setting.
22689
22690       -maddress-space-conversion
22691       -mno-address-space-conversion
22692           Allow/disallow treating the "__ea" address space as superset of the
22693           generic address space.  This enables explicit type casts between
22694           "__ea" and generic pointer as well as implicit conversions of
22695           generic pointers to "__ea" pointers.  The default is to allow
22696           address space pointer conversions.
22697
22698       -mcache-size=cache-size
22699           This option controls the version of libgcc that the compiler links
22700           to an executable and selects a software-managed cache for accessing
22701           variables in the "__ea" address space with a particular cache size.
22702           Possible options for cache-size are 8, 16, 32, 64 and 128.  The
22703           default cache size is 64KB.
22704
22705       -matomic-updates
22706       -mno-atomic-updates
22707           This option controls the version of libgcc that the compiler links
22708           to an executable and selects whether atomic updates to the
22709           software-managed cache of PPU-side variables are used.  If you use
22710           atomic updates, changes to a PPU variable from SPU code using the
22711           "__ea" named address space qualifier do not interfere with changes
22712           to other PPU variables residing in the same cache line from PPU
22713           code.  If you do not use atomic updates, such interference may
22714           occur; however, writing back cache lines is more efficient.  The
22715           default behavior is to use atomic updates.
22716
22717       -mdual-nops
22718       -mdual-nops=n
22719           By default, GCC inserts NOPs to increase dual issue when it expects
22720           it to increase performance.  n can be a value from 0 to 10.  A
22721           smaller n inserts fewer NOPs.  10 is the default, 0 is the same as
22722           -mno-dual-nops.  Disabled with -Os.
22723
22724       -mhint-max-nops=n
22725           Maximum number of NOPs to insert for a branch hint.  A branch hint
22726           must be at least 8 instructions away from the branch it is
22727           affecting.  GCC inserts up to n NOPs to enforce this, otherwise it
22728           does not generate the branch hint.
22729
22730       -mhint-max-distance=n
22731           The encoding of the branch hint instruction limits the hint to be
22732           within 256 instructions of the branch it is affecting.  By default,
22733           GCC makes sure it is within 125.
22734
22735       -msafe-hints
22736           Work around a hardware bug that causes the SPU to stall
22737           indefinitely.  By default, GCC inserts the "hbrp" instruction to
22738           make sure this stall won't happen.
22739
22740       Options for System V
22741
22742       These additional options are available on System V Release 4 for
22743       compatibility with other compilers on those systems:
22744
22745       -G  Create a shared object.  It is recommended that -symbolic or
22746           -shared be used instead.
22747
22748       -Qy Identify the versions of each tool used by the compiler, in a
22749           ".ident" assembler directive in the output.
22750
22751       -Qn Refrain from adding ".ident" directives to the output file (this is
22752           the default).
22753
22754       -YP,dirs
22755           Search the directories dirs, and no others, for libraries specified
22756           with -l.
22757
22758       -Ym,dir
22759           Look in the directory dir to find the M4 preprocessor.  The
22760           assembler uses this option.
22761
22762       TILE-Gx Options
22763
22764       These -m options are supported on the TILE-Gx:
22765
22766       -mcmodel=small
22767           Generate code for the small model.  The distance for direct calls
22768           is limited to 500M in either direction.  PC-relative addresses are
22769           32 bits.  Absolute addresses support the full address range.
22770
22771       -mcmodel=large
22772           Generate code for the large model.  There is no limitation on call
22773           distance, pc-relative addresses, or absolute addresses.
22774
22775       -mcpu=name
22776           Selects the type of CPU to be targeted.  Currently the only
22777           supported type is tilegx.
22778
22779       -m32
22780       -m64
22781           Generate code for a 32-bit or 64-bit environment.  The 32-bit
22782           environment sets int, long, and pointer to 32 bits.  The 64-bit
22783           environment sets int to 32 bits and long and pointer to 64 bits.
22784
22785       -mbig-endian
22786       -mlittle-endian
22787           Generate code in big/little endian mode, respectively.
22788
22789       TILEPro Options
22790
22791       These -m options are supported on the TILEPro:
22792
22793       -mcpu=name
22794           Selects the type of CPU to be targeted.  Currently the only
22795           supported type is tilepro.
22796
22797       -m32
22798           Generate code for a 32-bit environment, which sets int, long, and
22799           pointer to 32 bits.  This is the only supported behavior so the
22800           flag is essentially ignored.
22801
22802       V850 Options
22803
22804       These -m options are defined for V850 implementations:
22805
22806       -mlong-calls
22807       -mno-long-calls
22808           Treat all calls as being far away (near).  If calls are assumed to
22809           be far away, the compiler always loads the function's address into
22810           a register, and calls indirect through the pointer.
22811
22812       -mno-ep
22813       -mep
22814           Do not optimize (do optimize) basic blocks that use the same index
22815           pointer 4 or more times to copy pointer into the "ep" register, and
22816           use the shorter "sld" and "sst" instructions.  The -mep option is
22817           on by default if you optimize.
22818
22819       -mno-prolog-function
22820       -mprolog-function
22821           Do not use (do use) external functions to save and restore
22822           registers at the prologue and epilogue of a function.  The external
22823           functions are slower, but use less code space if more than one
22824           function saves the same number of registers.  The -mprolog-function
22825           option is on by default if you optimize.
22826
22827       -mspace
22828           Try to make the code as small as possible.  At present, this just
22829           turns on the -mep and -mprolog-function options.
22830
22831       -mtda=n
22832           Put static or global variables whose size is n bytes or less into
22833           the tiny data area that register "ep" points to.  The tiny data
22834           area can hold up to 256 bytes in total (128 bytes for byte
22835           references).
22836
22837       -msda=n
22838           Put static or global variables whose size is n bytes or less into
22839           the small data area that register "gp" points to.  The small data
22840           area can hold up to 64 kilobytes.
22841
22842       -mzda=n
22843           Put static or global variables whose size is n bytes or less into
22844           the first 32 kilobytes of memory.
22845
22846       -mv850
22847           Specify that the target processor is the V850.
22848
22849       -mv850e3v5
22850           Specify that the target processor is the V850E3V5.  The
22851           preprocessor constant "__v850e3v5__" is defined if this option is
22852           used.
22853
22854       -mv850e2v4
22855           Specify that the target processor is the V850E3V5.  This is an
22856           alias for the -mv850e3v5 option.
22857
22858       -mv850e2v3
22859           Specify that the target processor is the V850E2V3.  The
22860           preprocessor constant "__v850e2v3__" is defined if this option is
22861           used.
22862
22863       -mv850e2
22864           Specify that the target processor is the V850E2.  The preprocessor
22865           constant "__v850e2__" is defined if this option is used.
22866
22867       -mv850e1
22868           Specify that the target processor is the V850E1.  The preprocessor
22869           constants "__v850e1__" and "__v850e__" are defined if this option
22870           is used.
22871
22872       -mv850es
22873           Specify that the target processor is the V850ES.  This is an alias
22874           for the -mv850e1 option.
22875
22876       -mv850e
22877           Specify that the target processor is the V850E.  The preprocessor
22878           constant "__v850e__" is defined if this option is used.
22879
22880           If neither -mv850 nor -mv850e nor -mv850e1 nor -mv850e2 nor
22881           -mv850e2v3 nor -mv850e3v5 are defined then a default target
22882           processor is chosen and the relevant __v850*__ preprocessor
22883           constant is defined.
22884
22885           The preprocessor constants "__v850" and "__v851__" are always
22886           defined, regardless of which processor variant is the target.
22887
22888       -mdisable-callt
22889       -mno-disable-callt
22890           This option suppresses generation of the "CALLT" instruction for
22891           the v850e, v850e1, v850e2, v850e2v3 and v850e3v5 flavors of the
22892           v850 architecture.
22893
22894           This option is enabled by default when the RH850 ABI is in use (see
22895           -mrh850-abi), and disabled by default when the GCC ABI is in use.
22896           If "CALLT" instructions are being generated then the C preprocessor
22897           symbol "__V850_CALLT__" is defined.
22898
22899       -mrelax
22900       -mno-relax
22901           Pass on (or do not pass on) the -mrelax command-line option to the
22902           assembler.
22903
22904       -mlong-jumps
22905       -mno-long-jumps
22906           Disable (or re-enable) the generation of PC-relative jump
22907           instructions.
22908
22909       -msoft-float
22910       -mhard-float
22911           Disable (or re-enable) the generation of hardware floating point
22912           instructions.  This option is only significant when the target
22913           architecture is V850E2V3 or higher.  If hardware floating point
22914           instructions are being generated then the C preprocessor symbol
22915           "__FPU_OK__" is defined, otherwise the symbol "__NO_FPU__" is
22916           defined.
22917
22918       -mloop
22919           Enables the use of the e3v5 LOOP instruction.  The use of this
22920           instruction is not enabled by default when the e3v5 architecture is
22921           selected because its use is still experimental.
22922
22923       -mrh850-abi
22924       -mghs
22925           Enables support for the RH850 version of the V850 ABI.  This is the
22926           default.  With this version of the ABI the following rules apply:
22927
22928           *   Integer sized structures and unions are returned via a memory
22929               pointer rather than a register.
22930
22931           *   Large structures and unions (more than 8 bytes in size) are
22932               passed by value.
22933
22934           *   Functions are aligned to 16-bit boundaries.
22935
22936           *   The -m8byte-align command-line option is supported.
22937
22938           *   The -mdisable-callt command-line option is enabled by default.
22939               The -mno-disable-callt command-line option is not supported.
22940
22941           When this version of the ABI is enabled the C preprocessor symbol
22942           "__V850_RH850_ABI__" is defined.
22943
22944       -mgcc-abi
22945           Enables support for the old GCC version of the V850 ABI.  With this
22946           version of the ABI the following rules apply:
22947
22948           *   Integer sized structures and unions are returned in register
22949               "r10".
22950
22951           *   Large structures and unions (more than 8 bytes in size) are
22952               passed by reference.
22953
22954           *   Functions are aligned to 32-bit boundaries, unless optimizing
22955               for size.
22956
22957           *   The -m8byte-align command-line option is not supported.
22958
22959           *   The -mdisable-callt command-line option is supported but not
22960               enabled by default.
22961
22962           When this version of the ABI is enabled the C preprocessor symbol
22963           "__V850_GCC_ABI__" is defined.
22964
22965       -m8byte-align
22966       -mno-8byte-align
22967           Enables support for "double" and "long long" types to be aligned on
22968           8-byte boundaries.  The default is to restrict the alignment of all
22969           objects to at most 4-bytes.  When -m8byte-align is in effect the C
22970           preprocessor symbol "__V850_8BYTE_ALIGN__" is defined.
22971
22972       -mbig-switch
22973           Generate code suitable for big switch tables.  Use this option only
22974           if the assembler/linker complain about out of range branches within
22975           a switch table.
22976
22977       -mapp-regs
22978           This option causes r2 and r5 to be used in the code generated by
22979           the compiler.  This setting is the default.
22980
22981       -mno-app-regs
22982           This option causes r2 and r5 to be treated as fixed registers.
22983
22984       VAX Options
22985
22986       These -m options are defined for the VAX:
22987
22988       -munix
22989           Do not output certain jump instructions ("aobleq" and so on) that
22990           the Unix assembler for the VAX cannot handle across long ranges.
22991
22992       -mgnu
22993           Do output those jump instructions, on the assumption that the GNU
22994           assembler is being used.
22995
22996       -mg Output code for G-format floating-point numbers instead of
22997           D-format.
22998
22999       Visium Options
23000
23001       -mdebug
23002           A program which performs file I/O and is destined to run on an MCM
23003           target should be linked with this option.  It causes the libraries
23004           libc.a and libdebug.a to be linked.  The program should be run on
23005           the target under the control of the GDB remote debugging stub.
23006
23007       -msim
23008           A program which performs file I/O and is destined to run on the
23009           simulator should be linked with option.  This causes libraries
23010           libc.a and libsim.a to be linked.
23011
23012       -mfpu
23013       -mhard-float
23014           Generate code containing floating-point instructions.  This is the
23015           default.
23016
23017       -mno-fpu
23018       -msoft-float
23019           Generate code containing library calls for floating-point.
23020
23021           -msoft-float changes the calling convention in the output file;
23022           therefore, it is only useful if you compile all of a program with
23023           this option.  In particular, you need to compile libgcc.a, the
23024           library that comes with GCC, with -msoft-float in order for this to
23025           work.
23026
23027       -mcpu=cpu_type
23028           Set the instruction set, register set, and instruction scheduling
23029           parameters for machine type cpu_type.  Supported values for
23030           cpu_type are mcm, gr5 and gr6.
23031
23032           mcm is a synonym of gr5 present for backward compatibility.
23033
23034           By default (unless configured otherwise), GCC generates code for
23035           the GR5 variant of the Visium architecture.
23036
23037           With -mcpu=gr6, GCC generates code for the GR6 variant of the
23038           Visium architecture.  The only difference from GR5 code is that the
23039           compiler will generate block move instructions.
23040
23041       -mtune=cpu_type
23042           Set the instruction scheduling parameters for machine type
23043           cpu_type, but do not set the instruction set or register set that
23044           the option -mcpu=cpu_type would.
23045
23046       -msv-mode
23047           Generate code for the supervisor mode, where there are no
23048           restrictions on the access to general registers.  This is the
23049           default.
23050
23051       -muser-mode
23052           Generate code for the user mode, where the access to some general
23053           registers is forbidden: on the GR5, registers r24 to r31 cannot be
23054           accessed in this mode; on the GR6, only registers r29 to r31 are
23055           affected.
23056
23057       VMS Options
23058
23059       These -m options are defined for the VMS implementations:
23060
23061       -mvms-return-codes
23062           Return VMS condition codes from "main". The default is to return
23063           POSIX-style condition (e.g. error) codes.
23064
23065       -mdebug-main=prefix
23066           Flag the first routine whose name starts with prefix as the main
23067           routine for the debugger.
23068
23069       -mmalloc64
23070           Default to 64-bit memory allocation routines.
23071
23072       -mpointer-size=size
23073           Set the default size of pointers. Possible options for size are 32
23074           or short for 32 bit pointers, 64 or long for 64 bit pointers, and
23075           no for supporting only 32 bit pointers.  The later option disables
23076           "pragma pointer_size".
23077
23078       VxWorks Options
23079
23080       The options in this section are defined for all VxWorks targets.
23081       Options specific to the target hardware are listed with the other
23082       options for that target.
23083
23084       -mrtp
23085           GCC can generate code for both VxWorks kernels and real time
23086           processes (RTPs).  This option switches from the former to the
23087           latter.  It also defines the preprocessor macro "__RTP__".
23088
23089       -non-static
23090           Link an RTP executable against shared libraries rather than static
23091           libraries.  The options -static and -shared can also be used for
23092           RTPs; -static is the default.
23093
23094       -Bstatic
23095       -Bdynamic
23096           These options are passed down to the linker.  They are defined for
23097           compatibility with Diab.
23098
23099       -Xbind-lazy
23100           Enable lazy binding of function calls.  This option is equivalent
23101           to -Wl,-z,now and is defined for compatibility with Diab.
23102
23103       -Xbind-now
23104           Disable lazy binding of function calls.  This option is the default
23105           and is defined for compatibility with Diab.
23106
23107       x86 Options
23108
23109       These -m options are defined for the x86 family of computers.
23110
23111       -march=cpu-type
23112           Generate instructions for the machine type cpu-type.  In contrast
23113           to -mtune=cpu-type, which merely tunes the generated code for the
23114           specified cpu-type, -march=cpu-type allows GCC to generate code
23115           that may not run at all on processors other than the one indicated.
23116           Specifying -march=cpu-type implies -mtune=cpu-type.
23117
23118           The choices for cpu-type are:
23119
23120           native
23121               This selects the CPU to generate code for at compilation time
23122               by determining the processor type of the compiling machine.
23123               Using -march=native enables all instruction subsets supported
23124               by the local machine (hence the result might not run on
23125               different machines).  Using -mtune=native produces code
23126               optimized for the local machine under the constraints of the
23127               selected instruction set.
23128
23129           x86-64
23130               A generic CPU with 64-bit extensions.
23131
23132           i386
23133               Original Intel i386 CPU.
23134
23135           i486
23136               Intel i486 CPU.  (No scheduling is implemented for this chip.)
23137
23138           i586
23139           pentium
23140               Intel Pentium CPU with no MMX support.
23141
23142           lakemont
23143               Intel Lakemont MCU, based on Intel Pentium CPU.
23144
23145           pentium-mmx
23146               Intel Pentium MMX CPU, based on Pentium core with MMX
23147               instruction set support.
23148
23149           pentiumpro
23150               Intel Pentium Pro CPU.
23151
23152           i686
23153               When used with -march, the Pentium Pro instruction set is used,
23154               so the code runs on all i686 family chips.  When used with
23155               -mtune, it has the same meaning as generic.
23156
23157           pentium2
23158               Intel Pentium II CPU, based on Pentium Pro core with MMX
23159               instruction set support.
23160
23161           pentium3
23162           pentium3m
23163               Intel Pentium III CPU, based on Pentium Pro core with MMX and
23164               SSE instruction set support.
23165
23166           pentium-m
23167               Intel Pentium M; low-power version of Intel Pentium III CPU
23168               with MMX, SSE and SSE2 instruction set support.  Used by
23169               Centrino notebooks.
23170
23171           pentium4
23172           pentium4m
23173               Intel Pentium 4 CPU with MMX, SSE and SSE2 instruction set
23174               support.
23175
23176           prescott
23177               Improved version of Intel Pentium 4 CPU with MMX, SSE, SSE2 and
23178               SSE3 instruction set support.
23179
23180           nocona
23181               Improved version of Intel Pentium 4 CPU with 64-bit extensions,
23182               MMX, SSE, SSE2 and SSE3 instruction set support.
23183
23184           core2
23185               Intel Core 2 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3
23186               and SSSE3 instruction set support.
23187
23188           nehalem
23189               Intel Nehalem CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3,
23190               SSSE3, SSE4.1, SSE4.2 and POPCNT instruction set support.
23191
23192           westmere
23193               Intel Westmere CPU with 64-bit extensions, MMX, SSE, SSE2,
23194               SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES and PCLMUL instruction
23195               set support.
23196
23197           sandybridge
23198               Intel Sandy Bridge CPU with 64-bit extensions, MMX, SSE, SSE2,
23199               SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES and PCLMUL
23200               instruction set support.
23201
23202           ivybridge
23203               Intel Ivy Bridge CPU with 64-bit extensions, MMX, SSE, SSE2,
23204               SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES, PCLMUL,
23205               FSGSBASE, RDRND and F16C instruction set support.
23206
23207           haswell
23208               Intel Haswell CPU with 64-bit extensions, MOVBE, MMX, SSE,
23209               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
23210               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2 and F16C instruction
23211               set support.
23212
23213           broadwell
23214               Intel Broadwell CPU with 64-bit extensions, MOVBE, MMX, SSE,
23215               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
23216               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX and
23217               PREFETCHW instruction set support.
23218
23219           skylake
23220               Intel Skylake CPU with 64-bit extensions, MOVBE, MMX, SSE,
23221               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
23222               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
23223               PREFETCHW, CLFLUSHOPT, XSAVEC and XSAVES instruction set
23224               support.
23225
23226           bonnell
23227               Intel Bonnell CPU with 64-bit extensions, MOVBE, MMX, SSE,
23228               SSE2, SSE3 and SSSE3 instruction set support.
23229
23230           silvermont
23231               Intel Silvermont CPU with 64-bit extensions, MOVBE, MMX, SSE,
23232               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PCLMUL and
23233               RDRND instruction set support.
23234
23235           goldmont
23236               Intel Goldmont CPU with 64-bit extensions, MOVBE, MMX, SSE,
23237               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PCLMUL, RDRND,
23238               XSAVE, XSAVEOPT and FSGSBASE instruction set support.
23239
23240           goldmont-plus
23241               Intel Goldmont Plus CPU with 64-bit extensions, MOVBE, MMX,
23242               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PCLMUL,
23243               RDRND, XSAVE, XSAVEOPT, FSGSBASE, PTWRITE, RDPID, SGX and UMIP
23244               instruction set support.
23245
23246           tremont
23247               Intel Tremont CPU with 64-bit extensions, MOVBE, MMX, SSE,
23248               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AES, PCLMUL, RDRND,
23249               XSAVE, XSAVEOPT, FSGSBASE, PTWRITE, RDPID, SGX, UMIP, GFNI-SSE,
23250               CLWB and ENCLV instruction set support.
23251
23252           knl Intel Knight's Landing CPU with 64-bit extensions, MOVBE, MMX,
23253               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
23254               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
23255               PREFETCHW, AVX512F, AVX512PF, AVX512ER and AVX512CD instruction
23256               set support.
23257
23258           knm Intel Knights Mill CPU with 64-bit extensions, MOVBE, MMX, SSE,
23259               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES,
23260               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
23261               PREFETCHW, AVX512F, AVX512PF, AVX512ER, AVX512CD, AVX5124VNNIW,
23262               AVX5124FMAPS and AVX512VPOPCNTDQ instruction set support.
23263
23264           skylake-avx512
23265               Intel Skylake Server CPU with 64-bit extensions, MOVBE, MMX,
23266               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
23267               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
23268               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, CLWB,
23269               AVX512VL, AVX512BW, AVX512DQ and AVX512CD instruction set
23270               support.
23271
23272           cannonlake
23273               Intel Cannonlake Server CPU with 64-bit extensions, MOVBE, MMX,
23274               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
23275               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
23276               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
23277               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA and
23278               UMIP instruction set support.
23279
23280           icelake-client
23281               Intel Icelake Client CPU with 64-bit extensions, MOVBE, MMX,
23282               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
23283               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
23284               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
23285               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA,
23286               CLWB, UMIP, RDPID, GFNI, AVX512VBMI2, AVX512VPOPCNTDQ,
23287               AVX512BITALG, AVX512VNNI, VPCLMULQDQ, VAES instruction set
23288               support.
23289
23290           icelake-server
23291               Intel Icelake Server CPU with 64-bit extensions, MOVBE, MMX,
23292               SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2,
23293               AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED,
23294               ADCX, PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, AVX512VL,
23295               AVX512BW, AVX512DQ, AVX512CD, AVX512VBMI, AVX512IFMA, SHA,
23296               CLWB, UMIP, RDPID, GFNI, AVX512VBMI2, AVX512VPOPCNTDQ,
23297               AVX512BITALG, AVX512VNNI, VPCLMULQDQ, VAES, PCONFIG and
23298               WBNOINVD instruction set support.
23299
23300           cascadelake
23301               Intel Cascadelake CPU with 64-bit extensions, MOVBE, MMX, SSE,
23302               SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, PKU, AVX, AVX2, AES,
23303               PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C, RDSEED, ADCX,
23304               PREFETCHW, CLFLUSHOPT, XSAVEC, XSAVES, AVX512F, CLWB, AVX512VL,
23305               AVX512BW, AVX512DQ, AVX512CD and AVX512VNNI instruction set
23306               support.
23307
23308           k6  AMD K6 CPU with MMX instruction set support.
23309
23310           k6-2
23311           k6-3
23312               Improved versions of AMD K6 CPU with MMX and 3DNow! instruction
23313               set support.
23314
23315           athlon
23316           athlon-tbird
23317               AMD Athlon CPU with MMX, 3dNOW!, enhanced 3DNow! and SSE
23318               prefetch instructions support.
23319
23320           athlon-4
23321           athlon-xp
23322           athlon-mp
23323               Improved AMD Athlon CPU with MMX, 3DNow!, enhanced 3DNow! and
23324               full SSE instruction set support.
23325
23326           k8
23327           opteron
23328           athlon64
23329           athlon-fx
23330               Processors based on the AMD K8 core with x86-64 instruction set
23331               support, including the AMD Opteron, Athlon 64, and Athlon 64 FX
23332               processors.  (This supersets MMX, SSE, SSE2, 3DNow!, enhanced
23333               3DNow! and 64-bit instruction set extensions.)
23334
23335           k8-sse3
23336           opteron-sse3
23337           athlon64-sse3
23338               Improved versions of AMD K8 cores with SSE3 instruction set
23339               support.
23340
23341           amdfam10
23342           barcelona
23343               CPUs based on AMD Family 10h cores with x86-64 instruction set
23344               support.  (This supersets MMX, SSE, SSE2, SSE3, SSE4A, 3DNow!,
23345               enhanced 3DNow!, ABM and 64-bit instruction set extensions.)
23346
23347           bdver1
23348               CPUs based on AMD Family 15h cores with x86-64 instruction set
23349               support.  (This supersets FMA4, AVX, XOP, LWP, AES, PCL_MUL,
23350               CX16, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1, SSE4.2, ABM
23351               and 64-bit instruction set extensions.)
23352
23353           bdver2
23354               AMD Family 15h core based CPUs with x86-64 instruction set
23355               support.  (This supersets BMI, TBM, F16C, FMA, FMA4, AVX, XOP,
23356               LWP, AES, PCL_MUL, CX16, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3,
23357               SSE4.1, SSE4.2, ABM and 64-bit instruction set extensions.)
23358
23359           bdver3
23360               AMD Family 15h core based CPUs with x86-64 instruction set
23361               support.  (This supersets BMI, TBM, F16C, FMA, FMA4, FSGSBASE,
23362               AVX, XOP, LWP, AES, PCL_MUL, CX16, MMX, SSE, SSE2, SSE3, SSE4A,
23363               SSSE3, SSE4.1, SSE4.2, ABM and 64-bit instruction set
23364               extensions.
23365
23366           bdver4
23367               AMD Family 15h core based CPUs with x86-64 instruction set
23368               support.  (This supersets BMI, BMI2, TBM, F16C, FMA, FMA4,
23369               FSGSBASE, AVX, AVX2, XOP, LWP, AES, PCL_MUL, CX16, MOVBE, MMX,
23370               SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1, SSE4.2, ABM and 64-bit
23371               instruction set extensions.
23372
23373           znver1
23374               AMD Family 17h core based CPUs with x86-64 instruction set
23375               support.  (This supersets BMI, BMI2, F16C, FMA, FSGSBASE, AVX,
23376               AVX2, ADCX, RDSEED, MWAITX, SHA, CLZERO, AES, PCL_MUL, CX16,
23377               MOVBE, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1, SSE4.2, ABM,
23378               XSAVEC, XSAVES, CLFLUSHOPT, POPCNT, and 64-bit instruction set
23379               extensions.
23380
23381           znver2
23382               AMD Family 17h core based CPUs with x86-64 instruction set
23383               support. (This supersets BMI, BMI2, ,CLWB, F16C, FMA, FSGSBASE,
23384               AVX, AVX2, ADCX, RDSEED, MWAITX, SHA, CLZERO, AES, PCL_MUL,
23385               CX16, MOVBE, MMX, SSE, SSE2, SSE3, SSE4A, SSSE3, SSE4.1,
23386               SSE4.2, ABM, XSAVEC, XSAVES, CLFLUSHOPT, POPCNT, and 64-bit
23387               instruction set extensions.)
23388
23389           btver1
23390               CPUs based on AMD Family 14h cores with x86-64 instruction set
23391               support.  (This supersets MMX, SSE, SSE2, SSE3, SSSE3, SSE4A,
23392               CX16, ABM and 64-bit instruction set extensions.)
23393
23394           btver2
23395               CPUs based on AMD Family 16h cores with x86-64 instruction set
23396               support. This includes MOVBE, F16C, BMI, AVX, PCL_MUL, AES,
23397               SSE4.2, SSE4.1, CX16, ABM, SSE4A, SSSE3, SSE3, SSE2, SSE, MMX
23398               and 64-bit instruction set extensions.
23399
23400           winchip-c6
23401               IDT WinChip C6 CPU, dealt in same way as i486 with additional
23402               MMX instruction set support.
23403
23404           winchip2
23405               IDT WinChip 2 CPU, dealt in same way as i486 with additional
23406               MMX and 3DNow!  instruction set support.
23407
23408           c3  VIA C3 CPU with MMX and 3DNow! instruction set support.  (No
23409               scheduling is implemented for this chip.)
23410
23411           c3-2
23412               VIA C3-2 (Nehemiah/C5XL) CPU with MMX and SSE instruction set
23413               support.  (No scheduling is implemented for this chip.)
23414
23415           c7  VIA C7 (Esther) CPU with MMX, SSE, SSE2 and SSE3 instruction
23416               set support.  (No scheduling is implemented for this chip.)
23417
23418           samuel-2
23419               VIA Eden Samuel 2 CPU with MMX and 3DNow! instruction set
23420               support.  (No scheduling is implemented for this chip.)
23421
23422           nehemiah
23423               VIA Eden Nehemiah CPU with MMX and SSE instruction set support.
23424               (No scheduling is implemented for this chip.)
23425
23426           esther
23427               VIA Eden Esther CPU with MMX, SSE, SSE2 and SSE3 instruction
23428               set support.  (No scheduling is implemented for this chip.)
23429
23430           eden-x2
23431               VIA Eden X2 CPU with x86-64, MMX, SSE, SSE2 and SSE3
23432               instruction set support.  (No scheduling is implemented for
23433               this chip.)
23434
23435           eden-x4
23436               VIA Eden X4 CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3,
23437               SSE4.1, SSE4.2, AVX and AVX2 instruction set support.  (No
23438               scheduling is implemented for this chip.)
23439
23440           nano
23441               Generic VIA Nano CPU with x86-64, MMX, SSE, SSE2, SSE3 and
23442               SSSE3 instruction set support.  (No scheduling is implemented
23443               for this chip.)
23444
23445           nano-1000
23446               VIA Nano 1xxx CPU with x86-64, MMX, SSE, SSE2, SSE3 and SSSE3
23447               instruction set support.  (No scheduling is implemented for
23448               this chip.)
23449
23450           nano-2000
23451               VIA Nano 2xxx CPU with x86-64, MMX, SSE, SSE2, SSE3 and SSSE3
23452               instruction set support.  (No scheduling is implemented for
23453               this chip.)
23454
23455           nano-3000
23456               VIA Nano 3xxx CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3 and
23457               SSE4.1 instruction set support.  (No scheduling is implemented
23458               for this chip.)
23459
23460           nano-x2
23461               VIA Nano Dual Core CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3
23462               and SSE4.1 instruction set support.  (No scheduling is
23463               implemented for this chip.)
23464
23465           nano-x4
23466               VIA Nano Quad Core CPU with x86-64, MMX, SSE, SSE2, SSE3, SSSE3
23467               and SSE4.1 instruction set support.  (No scheduling is
23468               implemented for this chip.)
23469
23470           geode
23471               AMD Geode embedded processor with MMX and 3DNow! instruction
23472               set support.
23473
23474       -mtune=cpu-type
23475           Tune to cpu-type everything applicable about the generated code,
23476           except for the ABI and the set of available instructions.  While
23477           picking a specific cpu-type schedules things appropriately for that
23478           particular chip, the compiler does not generate any code that
23479           cannot run on the default machine type unless you use a -march=cpu-
23480           type option.  For example, if GCC is configured for
23481           i686-pc-linux-gnu then -mtune=pentium4 generates code that is tuned
23482           for Pentium 4 but still runs on i686 machines.
23483
23484           The choices for cpu-type are the same as for -march.  In addition,
23485           -mtune supports 2 extra choices for cpu-type:
23486
23487           generic
23488               Produce code optimized for the most common IA32/AMD64/EM64T
23489               processors.  If you know the CPU on which your code will run,
23490               then you should use the corresponding -mtune or -march option
23491               instead of -mtune=generic.  But, if you do not know exactly
23492               what CPU users of your application will have, then you should
23493               use this option.
23494
23495               As new processors are deployed in the marketplace, the behavior
23496               of this option will change.  Therefore, if you upgrade to a
23497               newer version of GCC, code generation controlled by this option
23498               will change to reflect the processors that are most common at
23499               the time that version of GCC is released.
23500
23501               There is no -march=generic option because -march indicates the
23502               instruction set the compiler can use, and there is no generic
23503               instruction set applicable to all processors.  In contrast,
23504               -mtune indicates the processor (or, in this case, collection of
23505               processors) for which the code is optimized.
23506
23507           intel
23508               Produce code optimized for the most current Intel processors,
23509               which are Haswell and Silvermont for this version of GCC.  If
23510               you know the CPU on which your code will run, then you should
23511               use the corresponding -mtune or -march option instead of
23512               -mtune=intel.  But, if you want your application performs
23513               better on both Haswell and Silvermont, then you should use this
23514               option.
23515
23516               As new Intel processors are deployed in the marketplace, the
23517               behavior of this option will change.  Therefore, if you upgrade
23518               to a newer version of GCC, code generation controlled by this
23519               option will change to reflect the most current Intel processors
23520               at the time that version of GCC is released.
23521
23522               There is no -march=intel option because -march indicates the
23523               instruction set the compiler can use, and there is no common
23524               instruction set applicable to all processors.  In contrast,
23525               -mtune indicates the processor (or, in this case, collection of
23526               processors) for which the code is optimized.
23527
23528       -mcpu=cpu-type
23529           A deprecated synonym for -mtune.
23530
23531       -mfpmath=unit
23532           Generate floating-point arithmetic for selected unit unit.  The
23533           choices for unit are:
23534
23535           387 Use the standard 387 floating-point coprocessor present on the
23536               majority of chips and emulated otherwise.  Code compiled with
23537               this option runs almost everywhere.  The temporary results are
23538               computed in 80-bit precision instead of the precision specified
23539               by the type, resulting in slightly different results compared
23540               to most of other chips.  See -ffloat-store for more detailed
23541               description.
23542
23543               This is the default choice for non-Darwin x86-32 targets.
23544
23545           sse Use scalar floating-point instructions present in the SSE
23546               instruction set.  This instruction set is supported by Pentium
23547               III and newer chips, and in the AMD line by Athlon-4, Athlon XP
23548               and Athlon MP chips.  The earlier version of the SSE
23549               instruction set supports only single-precision arithmetic, thus
23550               the double and extended-precision arithmetic are still done
23551               using 387.  A later version, present only in Pentium 4 and AMD
23552               x86-64 chips, supports double-precision arithmetic too.
23553
23554               For the x86-32 compiler, you must use -march=cpu-type, -msse or
23555               -msse2 switches to enable SSE extensions and make this option
23556               effective.  For the x86-64 compiler, these extensions are
23557               enabled by default.
23558
23559               The resulting code should be considerably faster in the
23560               majority of cases and avoid the numerical instability problems
23561               of 387 code, but may break some existing code that expects
23562               temporaries to be 80 bits.
23563
23564               This is the default choice for the x86-64 compiler, Darwin
23565               x86-32 targets, and the default choice for x86-32 targets with
23566               the SSE2 instruction set when -ffast-math is enabled.
23567
23568           sse,387
23569           sse+387
23570           both
23571               Attempt to utilize both instruction sets at once.  This
23572               effectively doubles the amount of available registers, and on
23573               chips with separate execution units for 387 and SSE the
23574               execution resources too.  Use this option with care, as it is
23575               still experimental, because the GCC register allocator does not
23576               model separate functional units well, resulting in unstable
23577               performance.
23578
23579       -masm=dialect
23580           Output assembly instructions using selected dialect.  Also affects
23581           which dialect is used for basic "asm" and extended "asm". Supported
23582           choices (in dialect order) are att or intel. The default is att.
23583           Darwin does not support intel.
23584
23585       -mieee-fp
23586       -mno-ieee-fp
23587           Control whether or not the compiler uses IEEE floating-point
23588           comparisons.  These correctly handle the case where the result of a
23589           comparison is unordered.
23590
23591       -m80387
23592       -mhard-float
23593           Generate output containing 80387 instructions for floating point.
23594
23595       -mno-80387
23596       -msoft-float
23597           Generate output containing library calls for floating point.
23598
23599           Warning: the requisite libraries are not part of GCC.  Normally the
23600           facilities of the machine's usual C compiler are used, but this
23601           cannot be done directly in cross-compilation.  You must make your
23602           own arrangements to provide suitable library functions for cross-
23603           compilation.
23604
23605           On machines where a function returns floating-point results in the
23606           80387 register stack, some floating-point opcodes may be emitted
23607           even if -msoft-float is used.
23608
23609       -mno-fp-ret-in-387
23610           Do not use the FPU registers for return values of functions.
23611
23612           The usual calling convention has functions return values of types
23613           "float" and "double" in an FPU register, even if there is no FPU.
23614           The idea is that the operating system should emulate an FPU.
23615
23616           The option -mno-fp-ret-in-387 causes such values to be returned in
23617           ordinary CPU registers instead.
23618
23619       -mno-fancy-math-387
23620           Some 387 emulators do not support the "sin", "cos" and "sqrt"
23621           instructions for the 387.  Specify this option to avoid generating
23622           those instructions.  This option is overridden when -march
23623           indicates that the target CPU always has an FPU and so the
23624           instruction does not need emulation.  These instructions are not
23625           generated unless you also use the -funsafe-math-optimizations
23626           switch.
23627
23628       -malign-double
23629       -mno-align-double
23630           Control whether GCC aligns "double", "long double", and "long long"
23631           variables on a two-word boundary or a one-word boundary.  Aligning
23632           "double" variables on a two-word boundary produces code that runs
23633           somewhat faster on a Pentium at the expense of more memory.
23634
23635           On x86-64, -malign-double is enabled by default.
23636
23637           Warning: if you use the -malign-double switch, structures
23638           containing the above types are aligned differently than the
23639           published application binary interface specifications for the
23640           x86-32 and are not binary compatible with structures in code
23641           compiled without that switch.
23642
23643       -m96bit-long-double
23644       -m128bit-long-double
23645           These switches control the size of "long double" type.  The x86-32
23646           application binary interface specifies the size to be 96 bits, so
23647           -m96bit-long-double is the default in 32-bit mode.
23648
23649           Modern architectures (Pentium and newer) prefer "long double" to be
23650           aligned to an 8- or 16-byte boundary.  In arrays or structures
23651           conforming to the ABI, this is not possible.  So specifying
23652           -m128bit-long-double aligns "long double" to a 16-byte boundary by
23653           padding the "long double" with an additional 32-bit zero.
23654
23655           In the x86-64 compiler, -m128bit-long-double is the default choice
23656           as its ABI specifies that "long double" is aligned on 16-byte
23657           boundary.
23658
23659           Notice that neither of these options enable any extra precision
23660           over the x87 standard of 80 bits for a "long double".
23661
23662           Warning: if you override the default value for your target ABI,
23663           this changes the size of structures and arrays containing "long
23664           double" variables, as well as modifying the function calling
23665           convention for functions taking "long double".  Hence they are not
23666           binary-compatible with code compiled without that switch.
23667
23668       -mlong-double-64
23669       -mlong-double-80
23670       -mlong-double-128
23671           These switches control the size of "long double" type. A size of 64
23672           bits makes the "long double" type equivalent to the "double" type.
23673           This is the default for 32-bit Bionic C library.  A size of 128
23674           bits makes the "long double" type equivalent to the "__float128"
23675           type. This is the default for 64-bit Bionic C library.
23676
23677           Warning: if you override the default value for your target ABI,
23678           this changes the size of structures and arrays containing "long
23679           double" variables, as well as modifying the function calling
23680           convention for functions taking "long double".  Hence they are not
23681           binary-compatible with code compiled without that switch.
23682
23683       -malign-data=type
23684           Control how GCC aligns variables.  Supported values for type are
23685           compat uses increased alignment value compatible uses GCC 4.8 and
23686           earlier, abi uses alignment value as specified by the psABI, and
23687           cacheline uses increased alignment value to match the cache line
23688           size.  compat is the default.
23689
23690       -mlarge-data-threshold=threshold
23691           When -mcmodel=medium is specified, data objects larger than
23692           threshold are placed in the large data section.  This value must be
23693           the same across all objects linked into the binary, and defaults to
23694           65535.
23695
23696       -mrtd
23697           Use a different function-calling convention, in which functions
23698           that take a fixed number of arguments return with the "ret num"
23699           instruction, which pops their arguments while returning.  This
23700           saves one instruction in the caller since there is no need to pop
23701           the arguments there.
23702
23703           You can specify that an individual function is called with this
23704           calling sequence with the function attribute "stdcall".  You can
23705           also override the -mrtd option by using the function attribute
23706           "cdecl".
23707
23708           Warning: this calling convention is incompatible with the one
23709           normally used on Unix, so you cannot use it if you need to call
23710           libraries compiled with the Unix compiler.
23711
23712           Also, you must provide function prototypes for all functions that
23713           take variable numbers of arguments (including "printf"); otherwise
23714           incorrect code is generated for calls to those functions.
23715
23716           In addition, seriously incorrect code results if you call a
23717           function with too many arguments.  (Normally, extra arguments are
23718           harmlessly ignored.)
23719
23720       -mregparm=num
23721           Control how many registers are used to pass integer arguments.  By
23722           default, no registers are used to pass arguments, and at most 3
23723           registers can be used.  You can control this behavior for a
23724           specific function by using the function attribute "regparm".
23725
23726           Warning: if you use this switch, and num is nonzero, then you must
23727           build all modules with the same value, including any libraries.
23728           This includes the system libraries and startup modules.
23729
23730       -msseregparm
23731           Use SSE register passing conventions for float and double arguments
23732           and return values.  You can control this behavior for a specific
23733           function by using the function attribute "sseregparm".
23734
23735           Warning: if you use this switch then you must build all modules
23736           with the same value, including any libraries.  This includes the
23737           system libraries and startup modules.
23738
23739       -mvect8-ret-in-mem
23740           Return 8-byte vectors in memory instead of MMX registers.  This is
23741           the default on Solaris@tie{}8 and 9 and VxWorks to match the ABI of
23742           the Sun Studio compilers until version 12.  Later compiler versions
23743           (starting with Studio 12 Update@tie{}1) follow the ABI used by
23744           other x86 targets, which is the default on Solaris@tie{}10 and
23745           later.  Only use this option if you need to remain compatible with
23746           existing code produced by those previous compiler versions or older
23747           versions of GCC.
23748
23749       -mpc32
23750       -mpc64
23751       -mpc80
23752           Set 80387 floating-point precision to 32, 64 or 80 bits.  When
23753           -mpc32 is specified, the significands of results of floating-point
23754           operations are rounded to 24 bits (single precision); -mpc64 rounds
23755           the significands of results of floating-point operations to 53 bits
23756           (double precision) and -mpc80 rounds the significands of results of
23757           floating-point operations to 64 bits (extended double precision),
23758           which is the default.  When this option is used, floating-point
23759           operations in higher precisions are not available to the programmer
23760           without setting the FPU control word explicitly.
23761
23762           Setting the rounding of floating-point operations to less than the
23763           default 80 bits can speed some programs by 2% or more.  Note that
23764           some mathematical libraries assume that extended-precision (80-bit)
23765           floating-point operations are enabled by default; routines in such
23766           libraries could suffer significant loss of accuracy, typically
23767           through so-called "catastrophic cancellation", when this option is
23768           used to set the precision to less than extended precision.
23769
23770       -mstackrealign
23771           Realign the stack at entry.  On the x86, the -mstackrealign option
23772           generates an alternate prologue and epilogue that realigns the run-
23773           time stack if necessary.  This supports mixing legacy codes that
23774           keep 4-byte stack alignment with modern codes that keep 16-byte
23775           stack alignment for SSE compatibility.  See also the attribute
23776           "force_align_arg_pointer", applicable to individual functions.
23777
23778       -mpreferred-stack-boundary=num
23779           Attempt to keep the stack boundary aligned to a 2 raised to num
23780           byte boundary.  If -mpreferred-stack-boundary is not specified, the
23781           default is 4 (16 bytes or 128 bits).
23782
23783           Warning: When generating code for the x86-64 architecture with SSE
23784           extensions disabled, -mpreferred-stack-boundary=3 can be used to
23785           keep the stack boundary aligned to 8 byte boundary.  Since x86-64
23786           ABI require 16 byte stack alignment, this is ABI incompatible and
23787           intended to be used in controlled environment where stack space is
23788           important limitation.  This option leads to wrong code when
23789           functions compiled with 16 byte stack alignment (such as functions
23790           from a standard library) are called with misaligned stack.  In this
23791           case, SSE instructions may lead to misaligned memory access traps.
23792           In addition, variable arguments are handled incorrectly for 16 byte
23793           aligned objects (including x87 long double and __int128), leading
23794           to wrong results.  You must build all modules with
23795           -mpreferred-stack-boundary=3, including any libraries.  This
23796           includes the system libraries and startup modules.
23797
23798       -mincoming-stack-boundary=num
23799           Assume the incoming stack is aligned to a 2 raised to num byte
23800           boundary.  If -mincoming-stack-boundary is not specified, the one
23801           specified by -mpreferred-stack-boundary is used.
23802
23803           On Pentium and Pentium Pro, "double" and "long double" values
23804           should be aligned to an 8-byte boundary (see -malign-double) or
23805           suffer significant run time performance penalties.  On Pentium III,
23806           the Streaming SIMD Extension (SSE) data type "__m128" may not work
23807           properly if it is not 16-byte aligned.
23808
23809           To ensure proper alignment of this values on the stack, the stack
23810           boundary must be as aligned as that required by any value stored on
23811           the stack.  Further, every function must be generated such that it
23812           keeps the stack aligned.  Thus calling a function compiled with a
23813           higher preferred stack boundary from a function compiled with a
23814           lower preferred stack boundary most likely misaligns the stack.  It
23815           is recommended that libraries that use callbacks always use the
23816           default setting.
23817
23818           This extra alignment does consume extra stack space, and generally
23819           increases code size.  Code that is sensitive to stack space usage,
23820           such as embedded systems and operating system kernels, may want to
23821           reduce the preferred alignment to -mpreferred-stack-boundary=2.
23822
23823       -mmmx
23824       -msse
23825       -msse2
23826       -msse3
23827       -mssse3
23828       -msse4
23829       -msse4a
23830       -msse4.1
23831       -msse4.2
23832       -mavx
23833       -mavx2
23834       -mavx512f
23835       -mavx512pf
23836       -mavx512er
23837       -mavx512cd
23838       -mavx512vl
23839       -mavx512bw
23840       -mavx512dq
23841       -mavx512ifma
23842       -mavx512vbmi
23843       -msha
23844       -maes
23845       -mpclmul
23846       -mclflushopt
23847       -mclwb
23848       -mfsgsbase
23849       -mptwrite
23850       -mrdrnd
23851       -mf16c
23852       -mfma
23853       -mpconfig
23854       -mwbnoinvd
23855       -mfma4
23856       -mprfchw
23857       -mrdpid
23858       -mprefetchwt1
23859       -mrdseed
23860       -msgx
23861       -mxop
23862       -mlwp
23863       -m3dnow
23864       -m3dnowa
23865       -mpopcnt
23866       -mabm
23867       -madx
23868       -mbmi
23869       -mbmi2
23870       -mlzcnt
23871       -mfxsr
23872       -mxsave
23873       -mxsaveopt
23874       -mxsavec
23875       -mxsaves
23876       -mrtm
23877       -mhle
23878       -mtbm
23879       -mmwaitx
23880       -mclzero
23881       -mpku
23882       -mavx512vbmi2
23883       -mgfni
23884       -mvaes
23885       -mwaitpkg
23886       -mvpclmulqdq
23887       -mavx512bitalg
23888       -mmovdiri
23889       -mmovdir64b
23890       -mavx512vpopcntdq
23891       -mavx5124fmaps
23892       -mavx512vnni
23893       -mavx5124vnniw
23894       -mcldemote
23895           These switches enable the use of instructions in the MMX, SSE,
23896           SSE2, SSE3, SSSE3, SSE4, SSE4A, SSE4.1, SSE4.2, AVX, AVX2, AVX512F,
23897           AVX512PF, AVX512ER, AVX512CD, AVX512VL, AVX512BW, AVX512DQ,
23898           AVX512IFMA, AVX512VBMI, SHA, AES, PCLMUL, CLFLUSHOPT, CLWB,
23899           FSGSBASE, PTWRITE, RDRND, F16C, FMA, PCONFIG, WBNOINVD, FMA4,
23900           PREFETCHW, RDPID, PREFETCHWT1, RDSEED, SGX, XOP, LWP, 3DNow!,
23901           enhanced 3DNow!, POPCNT, ABM, ADX, BMI, BMI2, LZCNT, FXSR, XSAVE,
23902           XSAVEOPT, XSAVEC, XSAVES, RTM, HLE, TBM, MWAITX, CLZERO, PKU,
23903           AVX512VBMI2, GFNI, VAES, WAITPKG, VPCLMULQDQ, AVX512BITALG,
23904           MOVDIRI, MOVDIR64B, AVX512VPOPCNTDQ, AVX5124FMAPS, AVX512VNNI,
23905           AVX5124VNNIW, or CLDEMOTE extended instruction sets.  Each has a
23906           corresponding -mno- option to disable use of these instructions.
23907
23908           These extensions are also available as built-in functions: see x86
23909           Built-in Functions, for details of the functions enabled and
23910           disabled by these switches.
23911
23912           To generate SSE/SSE2 instructions automatically from floating-point
23913           code (as opposed to 387 instructions), see -mfpmath=sse.
23914
23915           GCC depresses SSEx instructions when -mavx is used. Instead, it
23916           generates new AVX instructions or AVX equivalence for all SSEx
23917           instructions when needed.
23918
23919           These options enable GCC to use these extended instructions in
23920           generated code, even without -mfpmath=sse.  Applications that
23921           perform run-time CPU detection must compile separate files for each
23922           supported architecture, using the appropriate flags.  In
23923           particular, the file containing the CPU detection code should be
23924           compiled without these options.
23925
23926       -mdump-tune-features
23927           This option instructs GCC to dump the names of the x86 performance
23928           tuning features and default settings. The names can be used in
23929           -mtune-ctrl=feature-list.
23930
23931       -mtune-ctrl=feature-list
23932           This option is used to do fine grain control of x86 code generation
23933           features.  feature-list is a comma separated list of feature names.
23934           See also -mdump-tune-features. When specified, the feature is
23935           turned on if it is not preceded with ^, otherwise, it is turned
23936           off.  -mtune-ctrl=feature-list is intended to be used by GCC
23937           developers. Using it may lead to code paths not covered by testing
23938           and can potentially result in compiler ICEs or runtime errors.
23939
23940       -mno-default
23941           This option instructs GCC to turn off all tunable features. See
23942           also -mtune-ctrl=feature-list and -mdump-tune-features.
23943
23944       -mcld
23945           This option instructs GCC to emit a "cld" instruction in the
23946           prologue of functions that use string instructions.  String
23947           instructions depend on the DF flag to select between autoincrement
23948           or autodecrement mode.  While the ABI specifies the DF flag to be
23949           cleared on function entry, some operating systems violate this
23950           specification by not clearing the DF flag in their exception
23951           dispatchers.  The exception handler can be invoked with the DF flag
23952           set, which leads to wrong direction mode when string instructions
23953           are used.  This option can be enabled by default on 32-bit x86
23954           targets by configuring GCC with the --enable-cld configure option.
23955           Generation of "cld" instructions can be suppressed with the
23956           -mno-cld compiler option in this case.
23957
23958       -mvzeroupper
23959           This option instructs GCC to emit a "vzeroupper" instruction before
23960           a transfer of control flow out of the function to minimize the AVX
23961           to SSE transition penalty as well as remove unnecessary "zeroupper"
23962           intrinsics.
23963
23964       -mprefer-avx128
23965           This option instructs GCC to use 128-bit AVX instructions instead
23966           of 256-bit AVX instructions in the auto-vectorizer.
23967
23968       -mprefer-vector-width=opt
23969           This option instructs GCC to use opt-bit vector width in
23970           instructions instead of default on the selected platform.
23971
23972           none
23973               No extra limitations applied to GCC other than defined by the
23974               selected platform.
23975
23976           128 Prefer 128-bit vector width for instructions.
23977
23978           256 Prefer 256-bit vector width for instructions.
23979
23980           512 Prefer 512-bit vector width for instructions.
23981
23982       -mcx16
23983           This option enables GCC to generate "CMPXCHG16B" instructions in
23984           64-bit code to implement compare-and-exchange operations on 16-byte
23985           aligned 128-bit objects.  This is useful for atomic updates of data
23986           structures exceeding one machine word in size.  The compiler uses
23987           this instruction to implement __sync Builtins.  However, for
23988           __atomic Builtins operating on 128-bit integers, a library call is
23989           always used.
23990
23991       -msahf
23992           This option enables generation of "SAHF" instructions in 64-bit
23993           code.  Early Intel Pentium 4 CPUs with Intel 64 support, prior to
23994           the introduction of Pentium 4 G1 step in December 2005, lacked the
23995           "LAHF" and "SAHF" instructions which are supported by AMD64.  These
23996           are load and store instructions, respectively, for certain status
23997           flags.  In 64-bit mode, the "SAHF" instruction is used to optimize
23998           "fmod", "drem", and "remainder" built-in functions; see Other
23999           Builtins for details.
24000
24001       -mmovbe
24002           This option enables use of the "movbe" instruction to implement
24003           "__builtin_bswap32" and "__builtin_bswap64".
24004
24005       -mshstk
24006           The -mshstk option enables shadow stack built-in functions from x86
24007           Control-flow Enforcement Technology (CET).
24008
24009       -mcrc32
24010           This option enables built-in functions "__builtin_ia32_crc32qi",
24011           "__builtin_ia32_crc32hi", "__builtin_ia32_crc32si" and
24012           "__builtin_ia32_crc32di" to generate the "crc32" machine
24013           instruction.
24014
24015       -mrecip
24016           This option enables use of "RCPSS" and "RSQRTSS" instructions (and
24017           their vectorized variants "RCPPS" and "RSQRTPS") with an additional
24018           Newton-Raphson step to increase precision instead of "DIVSS" and
24019           "SQRTSS" (and their vectorized variants) for single-precision
24020           floating-point arguments.  These instructions are generated only
24021           when -funsafe-math-optimizations is enabled together with
24022           -ffinite-math-only and -fno-trapping-math.  Note that while the
24023           throughput of the sequence is higher than the throughput of the
24024           non-reciprocal instruction, the precision of the sequence can be
24025           decreased by up to 2 ulp (i.e. the inverse of 1.0 equals
24026           0.99999994).
24027
24028           Note that GCC implements "1.0f/sqrtf(x)" in terms of "RSQRTSS" (or
24029           "RSQRTPS") already with -ffast-math (or the above option
24030           combination), and doesn't need -mrecip.
24031
24032           Also note that GCC emits the above sequence with additional Newton-
24033           Raphson step for vectorized single-float division and vectorized
24034           "sqrtf(x)" already with -ffast-math (or the above option
24035           combination), and doesn't need -mrecip.
24036
24037       -mrecip=opt
24038           This option controls which reciprocal estimate instructions may be
24039           used.  opt is a comma-separated list of options, which may be
24040           preceded by a ! to invert the option:
24041
24042           all Enable all estimate instructions.
24043
24044           default
24045               Enable the default instructions, equivalent to -mrecip.
24046
24047           none
24048               Disable all estimate instructions, equivalent to -mno-recip.
24049
24050           div Enable the approximation for scalar division.
24051
24052           vec-div
24053               Enable the approximation for vectorized division.
24054
24055           sqrt
24056               Enable the approximation for scalar square root.
24057
24058           vec-sqrt
24059               Enable the approximation for vectorized square root.
24060
24061           So, for example, -mrecip=all,!sqrt enables all of the reciprocal
24062           approximations, except for square root.
24063
24064       -mveclibabi=type
24065           Specifies the ABI type to use for vectorizing intrinsics using an
24066           external library.  Supported values for type are svml for the Intel
24067           short vector math library and acml for the AMD math core library.
24068           To use this option, both -ftree-vectorize and
24069           -funsafe-math-optimizations have to be enabled, and an SVML or ACML
24070           ABI-compatible library must be specified at link time.
24071
24072           GCC currently emits calls to "vmldExp2", "vmldLn2", "vmldLog102",
24073           "vmldPow2", "vmldTanh2", "vmldTan2", "vmldAtan2", "vmldAtanh2",
24074           "vmldCbrt2", "vmldSinh2", "vmldSin2", "vmldAsinh2", "vmldAsin2",
24075           "vmldCosh2", "vmldCos2", "vmldAcosh2", "vmldAcos2", "vmlsExp4",
24076           "vmlsLn4", "vmlsLog104", "vmlsPow4", "vmlsTanh4", "vmlsTan4",
24077           "vmlsAtan4", "vmlsAtanh4", "vmlsCbrt4", "vmlsSinh4", "vmlsSin4",
24078           "vmlsAsinh4", "vmlsAsin4", "vmlsCosh4", "vmlsCos4", "vmlsAcosh4"
24079           and "vmlsAcos4" for corresponding function type when
24080           -mveclibabi=svml is used, and "__vrd2_sin", "__vrd2_cos",
24081           "__vrd2_exp", "__vrd2_log", "__vrd2_log2", "__vrd2_log10",
24082           "__vrs4_sinf", "__vrs4_cosf", "__vrs4_expf", "__vrs4_logf",
24083           "__vrs4_log2f", "__vrs4_log10f" and "__vrs4_powf" for the
24084           corresponding function type when -mveclibabi=acml is used.
24085
24086       -mabi=name
24087           Generate code for the specified calling convention.  Permissible
24088           values are sysv for the ABI used on GNU/Linux and other systems,
24089           and ms for the Microsoft ABI.  The default is to use the Microsoft
24090           ABI when targeting Microsoft Windows and the SysV ABI on all other
24091           systems.  You can control this behavior for specific functions by
24092           using the function attributes "ms_abi" and "sysv_abi".
24093
24094       -mforce-indirect-call
24095           Force all calls to functions to be indirect. This is useful when
24096           using Intel Processor Trace where it generates more precise timing
24097           information for function calls.
24098
24099       -mmanual-endbr
24100           Insert ENDBR instruction at function entry only via the "cf_check"
24101           function attribute. This is useful when used with the option
24102           -fcf-protection=branch to control ENDBR insertion at the function
24103           entry.
24104
24105       -mcall-ms2sysv-xlogues
24106           Due to differences in 64-bit ABIs, any Microsoft ABI function that
24107           calls a System V ABI function must consider RSI, RDI and XMM6-15 as
24108           clobbered.  By default, the code for saving and restoring these
24109           registers is emitted inline, resulting in fairly lengthy prologues
24110           and epilogues.  Using -mcall-ms2sysv-xlogues emits prologues and
24111           epilogues that use stubs in the static portion of libgcc to perform
24112           these saves and restores, thus reducing function size at the cost
24113           of a few extra instructions.
24114
24115       -mtls-dialect=type
24116           Generate code to access thread-local storage using the gnu or gnu2
24117           conventions.  gnu is the conservative default; gnu2 is more
24118           efficient, but it may add compile- and run-time requirements that
24119           cannot be satisfied on all systems.
24120
24121       -mpush-args
24122       -mno-push-args
24123           Use PUSH operations to store outgoing parameters.  This method is
24124           shorter and usually equally fast as method using SUB/MOV operations
24125           and is enabled by default.  In some cases disabling it may improve
24126           performance because of improved scheduling and reduced
24127           dependencies.
24128
24129       -maccumulate-outgoing-args
24130           If enabled, the maximum amount of space required for outgoing
24131           arguments is computed in the function prologue.  This is faster on
24132           most modern CPUs because of reduced dependencies, improved
24133           scheduling and reduced stack usage when the preferred stack
24134           boundary is not equal to 2.  The drawback is a notable increase in
24135           code size.  This switch implies -mno-push-args.
24136
24137       -mthreads
24138           Support thread-safe exception handling on MinGW.  Programs that
24139           rely on thread-safe exception handling must compile and link all
24140           code with the -mthreads option.  When compiling, -mthreads defines
24141           -D_MT; when linking, it links in a special thread helper library
24142           -lmingwthrd which cleans up per-thread exception-handling data.
24143
24144       -mms-bitfields
24145       -mno-ms-bitfields
24146           Enable/disable bit-field layout compatible with the native
24147           Microsoft Windows compiler.
24148
24149           If "packed" is used on a structure, or if bit-fields are used, it
24150           may be that the Microsoft ABI lays out the structure differently
24151           than the way GCC normally does.  Particularly when moving packed
24152           data between functions compiled with GCC and the native Microsoft
24153           compiler (either via function call or as data in a file), it may be
24154           necessary to access either format.
24155
24156           This option is enabled by default for Microsoft Windows targets.
24157           This behavior can also be controlled locally by use of variable or
24158           type attributes.  For more information, see x86 Variable Attributes
24159           and x86 Type Attributes.
24160
24161           The Microsoft structure layout algorithm is fairly simple with the
24162           exception of the bit-field packing.  The padding and alignment of
24163           members of structures and whether a bit-field can straddle a
24164           storage-unit boundary are determine by these rules:
24165
24166           1. Structure members are stored sequentially in the order in which
24167           they are
24168               declared: the first member has the lowest memory address and
24169               the last member the highest.
24170
24171           2. Every data object has an alignment requirement.  The alignment
24172           requirement
24173               for all data except structures, unions, and arrays is either
24174               the size of the object or the current packing size (specified
24175               with either the "aligned" attribute or the "pack" pragma),
24176               whichever is less.  For structures, unions, and arrays, the
24177               alignment requirement is the largest alignment requirement of
24178               its members.  Every object is allocated an offset so that:
24179
24180                       offset % alignment_requirement == 0
24181
24182           3. Adjacent bit-fields are packed into the same 1-, 2-, or 4-byte
24183           allocation
24184               unit if the integral types are the same size and if the next
24185               bit-field fits into the current allocation unit without
24186               crossing the boundary imposed by the common alignment
24187               requirements of the bit-fields.
24188
24189           MSVC interprets zero-length bit-fields in the following ways:
24190
24191           1. If a zero-length bit-field is inserted between two bit-fields
24192           that
24193               are normally coalesced, the bit-fields are not coalesced.
24194
24195               For example:
24196
24197                       struct
24198                        {
24199                          unsigned long bf_1 : 12;
24200                          unsigned long : 0;
24201                          unsigned long bf_2 : 12;
24202                        } t1;
24203
24204               The size of "t1" is 8 bytes with the zero-length bit-field.  If
24205               the zero-length bit-field were removed, "t1"'s size would be 4
24206               bytes.
24207
24208           2. If a zero-length bit-field is inserted after a bit-field, "foo",
24209           and the
24210               alignment of the zero-length bit-field is greater than the
24211               member that follows it, "bar", "bar" is aligned as the type of
24212               the zero-length bit-field.
24213
24214               For example:
24215
24216                       struct
24217                        {
24218                          char foo : 4;
24219                          short : 0;
24220                          char bar;
24221                        } t2;
24222
24223                       struct
24224                        {
24225                          char foo : 4;
24226                          short : 0;
24227                          double bar;
24228                        } t3;
24229
24230               For "t2", "bar" is placed at offset 2, rather than offset 1.
24231               Accordingly, the size of "t2" is 4.  For "t3", the zero-length
24232               bit-field does not affect the alignment of "bar" or, as a
24233               result, the size of the structure.
24234
24235               Taking this into account, it is important to note the
24236               following:
24237
24238               1. If a zero-length bit-field follows a normal bit-field, the
24239               type of the
24240                   zero-length bit-field may affect the alignment of the
24241                   structure as whole. For example, "t2" has a size of 4
24242                   bytes, since the zero-length bit-field follows a normal
24243                   bit-field, and is of type short.
24244
24245               2. Even if a zero-length bit-field is not followed by a normal
24246               bit-field, it may
24247                   still affect the alignment of the structure:
24248
24249                           struct
24250                            {
24251                              char foo : 6;
24252                              long : 0;
24253                            } t4;
24254
24255                   Here, "t4" takes up 4 bytes.
24256
24257           3. Zero-length bit-fields following non-bit-field members are
24258           ignored:
24259                       struct
24260                        {
24261                          char foo;
24262                          long : 0;
24263                          char bar;
24264                        } t5;
24265
24266               Here, "t5" takes up 2 bytes.
24267
24268       -mno-align-stringops
24269           Do not align the destination of inlined string operations.  This
24270           switch reduces code size and improves performance in case the
24271           destination is already aligned, but GCC doesn't know about it.
24272
24273       -minline-all-stringops
24274           By default GCC inlines string operations only when the destination
24275           is known to be aligned to least a 4-byte boundary.  This enables
24276           more inlining and increases code size, but may improve performance
24277           of code that depends on fast "memcpy", "strlen", and "memset" for
24278           short lengths.
24279
24280       -minline-stringops-dynamically
24281           For string operations of unknown size, use run-time checks with
24282           inline code for small blocks and a library call for large blocks.
24283
24284       -mstringop-strategy=alg
24285           Override the internal decision heuristic for the particular
24286           algorithm to use for inlining string operations.  The allowed
24287           values for alg are:
24288
24289           rep_byte
24290           rep_4byte
24291           rep_8byte
24292               Expand using i386 "rep" prefix of the specified size.
24293
24294           byte_loop
24295           loop
24296           unrolled_loop
24297               Expand into an inline loop.
24298
24299           libcall
24300               Always use a library call.
24301
24302       -mmemcpy-strategy=strategy
24303           Override the internal decision heuristic to decide if
24304           "__builtin_memcpy" should be inlined and what inline algorithm to
24305           use when the expected size of the copy operation is known. strategy
24306           is a comma-separated list of alg:max_size:dest_align triplets.  alg
24307           is specified in -mstringop-strategy, max_size specifies the max
24308           byte size with which inline algorithm alg is allowed.  For the last
24309           triplet, the max_size must be "-1". The max_size of the triplets in
24310           the list must be specified in increasing order.  The minimal byte
24311           size for alg is 0 for the first triplet and "max_size + 1" of the
24312           preceding range.
24313
24314       -mmemset-strategy=strategy
24315           The option is similar to -mmemcpy-strategy= except that it is to
24316           control "__builtin_memset" expansion.
24317
24318       -momit-leaf-frame-pointer
24319           Don't keep the frame pointer in a register for leaf functions.
24320           This avoids the instructions to save, set up, and restore frame
24321           pointers and makes an extra register available in leaf functions.
24322           The option -fomit-leaf-frame-pointer removes the frame pointer for
24323           leaf functions, which might make debugging harder.
24324
24325       -mtls-direct-seg-refs
24326       -mno-tls-direct-seg-refs
24327           Controls whether TLS variables may be accessed with offsets from
24328           the TLS segment register (%gs for 32-bit, %fs for 64-bit), or
24329           whether the thread base pointer must be added.  Whether or not this
24330           is valid depends on the operating system, and whether it maps the
24331           segment to cover the entire TLS area.
24332
24333           For systems that use the GNU C Library, the default is on.
24334
24335       -msse2avx
24336       -mno-sse2avx
24337           Specify that the assembler should encode SSE instructions with VEX
24338           prefix.  The option -mavx turns this on by default.
24339
24340       -mfentry
24341       -mno-fentry
24342           If profiling is active (-pg), put the profiling counter call before
24343           the prologue.  Note: On x86 architectures the attribute
24344           "ms_hook_prologue" isn't possible at the moment for -mfentry and
24345           -pg.
24346
24347       -mrecord-mcount
24348       -mno-record-mcount
24349           If profiling is active (-pg), generate a __mcount_loc section that
24350           contains pointers to each profiling call. This is useful for
24351           automatically patching and out calls.
24352
24353       -mnop-mcount
24354       -mno-nop-mcount
24355           If profiling is active (-pg), generate the calls to the profiling
24356           functions as NOPs. This is useful when they should be patched in
24357           later dynamically. This is likely only useful together with
24358           -mrecord-mcount.
24359
24360       -minstrument-return=type
24361           Instrument function exit in -pg -mfentry instrumented functions
24362           with call to specified function. This only instruments true returns
24363           ending with ret, but not sibling calls ending with jump. Valid
24364           types are none to not instrument, call to generate a call to
24365           __return__, or nop5 to generate a 5 byte nop.
24366
24367       -mrecord-return
24368       -mno-record-return
24369           Generate a __return_loc section pointing to all return
24370           instrumentation code.
24371
24372       -mfentry-name=name
24373           Set name of __fentry__ symbol called at function entry for -pg
24374           -mfentry functions.
24375
24376       -mfentry-section=name
24377           Set name of section to record -mrecord-mcount calls (default
24378           __mcount_loc).
24379
24380       -mskip-rax-setup
24381       -mno-skip-rax-setup
24382           When generating code for the x86-64 architecture with SSE
24383           extensions disabled, -mskip-rax-setup can be used to skip setting
24384           up RAX register when there are no variable arguments passed in
24385           vector registers.
24386
24387           Warning: Since RAX register is used to avoid unnecessarily saving
24388           vector registers on stack when passing variable arguments, the
24389           impacts of this option are callees may waste some stack space,
24390           misbehave or jump to a random location.  GCC 4.4 or newer don't
24391           have those issues, regardless the RAX register value.
24392
24393       -m8bit-idiv
24394       -mno-8bit-idiv
24395           On some processors, like Intel Atom, 8-bit unsigned integer divide
24396           is much faster than 32-bit/64-bit integer divide.  This option
24397           generates a run-time check.  If both dividend and divisor are
24398           within range of 0 to 255, 8-bit unsigned integer divide is used
24399           instead of 32-bit/64-bit integer divide.
24400
24401       -mavx256-split-unaligned-load
24402       -mavx256-split-unaligned-store
24403           Split 32-byte AVX unaligned load and store.
24404
24405       -mstack-protector-guard=guard
24406       -mstack-protector-guard-reg=reg
24407       -mstack-protector-guard-offset=offset
24408           Generate stack protection code using canary at guard.  Supported
24409           locations are global for global canary or tls for per-thread canary
24410           in the TLS block (the default).  This option has effect only when
24411           -fstack-protector or -fstack-protector-all is specified.
24412
24413           With the latter choice the options -mstack-protector-guard-reg=reg
24414           and -mstack-protector-guard-offset=offset furthermore specify which
24415           segment register (%fs or %gs) to use as base register for reading
24416           the canary, and from what offset from that base register.  The
24417           default for those is as specified in the relevant ABI.
24418
24419       -mgeneral-regs-only
24420           Generate code that uses only the general-purpose registers.  This
24421           prevents the compiler from using floating-point, vector, mask and
24422           bound registers.
24423
24424       -mindirect-branch=choice
24425           Convert indirect call and jump with choice.  The default is keep,
24426           which keeps indirect call and jump unmodified.  thunk converts
24427           indirect call and jump to call and return thunk.  thunk-inline
24428           converts indirect call and jump to inlined call and return thunk.
24429           thunk-extern converts indirect call and jump to external call and
24430           return thunk provided in a separate object file.  You can control
24431           this behavior for a specific function by using the function
24432           attribute "indirect_branch".
24433
24434           Note that -mcmodel=large is incompatible with
24435           -mindirect-branch=thunk and -mindirect-branch=thunk-extern since
24436           the thunk function may not be reachable in the large code model.
24437
24438           Note that -mindirect-branch=thunk-extern is incompatible with
24439           -fcf-protection=branch since the external thunk cannot be modified
24440           to disable control-flow check.
24441
24442       -mfunction-return=choice
24443           Convert function return with choice.  The default is keep, which
24444           keeps function return unmodified.  thunk converts function return
24445           to call and return thunk.  thunk-inline converts function return to
24446           inlined call and return thunk.  thunk-extern converts function
24447           return to external call and return thunk provided in a separate
24448           object file.  You can control this behavior for a specific function
24449           by using the function attribute "function_return".
24450
24451           Note that -mcmodel=large is incompatible with
24452           -mfunction-return=thunk and -mfunction-return=thunk-extern since
24453           the thunk function may not be reachable in the large code model.
24454
24455       -mindirect-branch-register
24456           Force indirect call and jump via register.
24457
24458       These -m switches are supported in addition to the above on x86-64
24459       processors in 64-bit environments.
24460
24461       -m32
24462       -m64
24463       -mx32
24464       -m16
24465       -miamcu
24466           Generate code for a 16-bit, 32-bit or 64-bit environment.  The -m32
24467           option sets "int", "long", and pointer types to 32 bits, and
24468           generates code that runs on any i386 system.
24469
24470           The -m64 option sets "int" to 32 bits and "long" and pointer types
24471           to 64 bits, and generates code for the x86-64 architecture.  For
24472           Darwin only the -m64 option also turns off the -fno-pic and
24473           -mdynamic-no-pic options.
24474
24475           The -mx32 option sets "int", "long", and pointer types to 32 bits,
24476           and generates code for the x86-64 architecture.
24477
24478           The -m16 option is the same as -m32, except for that it outputs the
24479           ".code16gcc" assembly directive at the beginning of the assembly
24480           output so that the binary can run in 16-bit mode.
24481
24482           The -miamcu option generates code which conforms to Intel MCU
24483           psABI.  It requires the -m32 option to be turned on.
24484
24485       -mno-red-zone
24486           Do not use a so-called "red zone" for x86-64 code.  The red zone is
24487           mandated by the x86-64 ABI; it is a 128-byte area beyond the
24488           location of the stack pointer that is not modified by signal or
24489           interrupt handlers and therefore can be used for temporary data
24490           without adjusting the stack pointer.  The flag -mno-red-zone
24491           disables this red zone.
24492
24493       -mcmodel=small
24494           Generate code for the small code model: the program and its symbols
24495           must be linked in the lower 2 GB of the address space.  Pointers
24496           are 64 bits.  Programs can be statically or dynamically linked.
24497           This is the default code model.
24498
24499       -mcmodel=kernel
24500           Generate code for the kernel code model.  The kernel runs in the
24501           negative 2 GB of the address space.  This model has to be used for
24502           Linux kernel code.
24503
24504       -mcmodel=medium
24505           Generate code for the medium model: the program is linked in the
24506           lower 2 GB of the address space.  Small symbols are also placed
24507           there.  Symbols with sizes larger than -mlarge-data-threshold are
24508           put into large data or BSS sections and can be located above 2GB.
24509           Programs can be statically or dynamically linked.
24510
24511       -mcmodel=large
24512           Generate code for the large model.  This model makes no assumptions
24513           about addresses and sizes of sections.
24514
24515       -maddress-mode=long
24516           Generate code for long address mode.  This is only supported for
24517           64-bit and x32 environments.  It is the default address mode for
24518           64-bit environments.
24519
24520       -maddress-mode=short
24521           Generate code for short address mode.  This is only supported for
24522           32-bit and x32 environments.  It is the default address mode for
24523           32-bit and x32 environments.
24524
24525       x86 Windows Options
24526
24527       These additional options are available for Microsoft Windows targets:
24528
24529       -mconsole
24530           This option specifies that a console application is to be
24531           generated, by instructing the linker to set the PE header subsystem
24532           type required for console applications.  This option is available
24533           for Cygwin and MinGW targets and is enabled by default on those
24534           targets.
24535
24536       -mdll
24537           This option is available for Cygwin and MinGW targets.  It
24538           specifies that a DLL---a dynamic link library---is to be generated,
24539           enabling the selection of the required runtime startup object and
24540           entry point.
24541
24542       -mnop-fun-dllimport
24543           This option is available for Cygwin and MinGW targets.  It
24544           specifies that the "dllimport" attribute should be ignored.
24545
24546       -mthread
24547           This option is available for MinGW targets. It specifies that
24548           MinGW-specific thread support is to be used.
24549
24550       -municode
24551           This option is available for MinGW-w64 targets.  It causes the
24552           "UNICODE" preprocessor macro to be predefined, and chooses Unicode-
24553           capable runtime startup code.
24554
24555       -mwin32
24556           This option is available for Cygwin and MinGW targets.  It
24557           specifies that the typical Microsoft Windows predefined macros are
24558           to be set in the pre-processor, but does not influence the choice
24559           of runtime library/startup code.
24560
24561       -mwindows
24562           This option is available for Cygwin and MinGW targets.  It
24563           specifies that a GUI application is to be generated by instructing
24564           the linker to set the PE header subsystem type appropriately.
24565
24566       -fno-set-stack-executable
24567           This option is available for MinGW targets. It specifies that the
24568           executable flag for the stack used by nested functions isn't set.
24569           This is necessary for binaries running in kernel mode of Microsoft
24570           Windows, as there the User32 API, which is used to set executable
24571           privileges, isn't available.
24572
24573       -fwritable-relocated-rdata
24574           This option is available for MinGW and Cygwin targets.  It
24575           specifies that relocated-data in read-only section is put into the
24576           ".data" section.  This is a necessary for older runtimes not
24577           supporting modification of ".rdata" sections for pseudo-relocation.
24578
24579       -mpe-aligned-commons
24580           This option is available for Cygwin and MinGW targets.  It
24581           specifies that the GNU extension to the PE file format that permits
24582           the correct alignment of COMMON variables should be used when
24583           generating code.  It is enabled by default if GCC detects that the
24584           target assembler found during configuration supports the feature.
24585
24586       See also under x86 Options for standard options.
24587
24588       Xstormy16 Options
24589
24590       These options are defined for Xstormy16:
24591
24592       -msim
24593           Choose startup files and linker script suitable for the simulator.
24594
24595       Xtensa Options
24596
24597       These options are supported for Xtensa targets:
24598
24599       -mconst16
24600       -mno-const16
24601           Enable or disable use of "CONST16" instructions for loading
24602           constant values.  The "CONST16" instruction is currently not a
24603           standard option from Tensilica.  When enabled, "CONST16"
24604           instructions are always used in place of the standard "L32R"
24605           instructions.  The use of "CONST16" is enabled by default only if
24606           the "L32R" instruction is not available.
24607
24608       -mfused-madd
24609       -mno-fused-madd
24610           Enable or disable use of fused multiply/add and multiply/subtract
24611           instructions in the floating-point option.  This has no effect if
24612           the floating-point option is not also enabled.  Disabling fused
24613           multiply/add and multiply/subtract instructions forces the compiler
24614           to use separate instructions for the multiply and add/subtract
24615           operations.  This may be desirable in some cases where strict IEEE
24616           754-compliant results are required: the fused multiply add/subtract
24617           instructions do not round the intermediate result, thereby
24618           producing results with more bits of precision than specified by the
24619           IEEE standard.  Disabling fused multiply add/subtract instructions
24620           also ensures that the program output is not sensitive to the
24621           compiler's ability to combine multiply and add/subtract operations.
24622
24623       -mserialize-volatile
24624       -mno-serialize-volatile
24625           When this option is enabled, GCC inserts "MEMW" instructions before
24626           "volatile" memory references to guarantee sequential consistency.
24627           The default is -mserialize-volatile.  Use -mno-serialize-volatile
24628           to omit the "MEMW" instructions.
24629
24630       -mforce-no-pic
24631           For targets, like GNU/Linux, where all user-mode Xtensa code must
24632           be position-independent code (PIC), this option disables PIC for
24633           compiling kernel code.
24634
24635       -mtext-section-literals
24636       -mno-text-section-literals
24637           These options control the treatment of literal pools.  The default
24638           is -mno-text-section-literals, which places literals in a separate
24639           section in the output file.  This allows the literal pool to be
24640           placed in a data RAM/ROM, and it also allows the linker to combine
24641           literal pools from separate object files to remove redundant
24642           literals and improve code size.  With -mtext-section-literals, the
24643           literals are interspersed in the text section in order to keep them
24644           as close as possible to their references.  This may be necessary
24645           for large assembly files.  Literals for each function are placed
24646           right before that function.
24647
24648       -mauto-litpools
24649       -mno-auto-litpools
24650           These options control the treatment of literal pools.  The default
24651           is -mno-auto-litpools, which places literals in a separate section
24652           in the output file unless -mtext-section-literals is used.  With
24653           -mauto-litpools the literals are interspersed in the text section
24654           by the assembler.  Compiler does not produce explicit ".literal"
24655           directives and loads literals into registers with "MOVI"
24656           instructions instead of "L32R" to let the assembler do relaxation
24657           and place literals as necessary.  This option allows assembler to
24658           create several literal pools per function and assemble very big
24659           functions, which may not be possible with -mtext-section-literals.
24660
24661       -mtarget-align
24662       -mno-target-align
24663           When this option is enabled, GCC instructs the assembler to
24664           automatically align instructions to reduce branch penalties at the
24665           expense of some code density.  The assembler attempts to widen
24666           density instructions to align branch targets and the instructions
24667           following call instructions.  If there are not enough preceding
24668           safe density instructions to align a target, no widening is
24669           performed.  The default is -mtarget-align.  These options do not
24670           affect the treatment of auto-aligned instructions like "LOOP",
24671           which the assembler always aligns, either by widening density
24672           instructions or by inserting NOP instructions.
24673
24674       -mlongcalls
24675       -mno-longcalls
24676           When this option is enabled, GCC instructs the assembler to
24677           translate direct calls to indirect calls unless it can determine
24678           that the target of a direct call is in the range allowed by the
24679           call instruction.  This translation typically occurs for calls to
24680           functions in other source files.  Specifically, the assembler
24681           translates a direct "CALL" instruction into an "L32R" followed by a
24682           "CALLX" instruction.  The default is -mno-longcalls.  This option
24683           should be used in programs where the call target can potentially be
24684           out of range.  This option is implemented in the assembler, not the
24685           compiler, so the assembly code generated by GCC still shows direct
24686           call instructions---look at the disassembled object code to see the
24687           actual instructions.  Note that the assembler uses an indirect call
24688           for every cross-file call, not just those that really are out of
24689           range.
24690
24691       zSeries Options
24692
24693       These are listed under
24694

ENVIRONMENT

24696       This section describes several environment variables that affect how
24697       GCC operates.  Some of them work by specifying directories or prefixes
24698       to use when searching for various kinds of files.  Some are used to
24699       specify other aspects of the compilation environment.
24700
24701       Note that you can also specify places to search using options such as
24702       -B, -I and -L.  These take precedence over places specified using
24703       environment variables, which in turn take precedence over those
24704       specified by the configuration of GCC.
24705
24706       LANG
24707       LC_CTYPE
24708       LC_MESSAGES
24709       LC_ALL
24710           These environment variables control the way that GCC uses
24711           localization information which allows GCC to work with different
24712           national conventions.  GCC inspects the locale categories LC_CTYPE
24713           and LC_MESSAGES if it has been configured to do so.  These locale
24714           categories can be set to any value supported by your installation.
24715           A typical value is en_GB.UTF-8 for English in the United Kingdom
24716           encoded in UTF-8.
24717
24718           The LC_CTYPE environment variable specifies character
24719           classification.  GCC uses it to determine the character boundaries
24720           in a string; this is needed for some multibyte encodings that
24721           contain quote and escape characters that are otherwise interpreted
24722           as a string end or escape.
24723
24724           The LC_MESSAGES environment variable specifies the language to use
24725           in diagnostic messages.
24726
24727           If the LC_ALL environment variable is set, it overrides the value
24728           of LC_CTYPE and LC_MESSAGES; otherwise, LC_CTYPE and LC_MESSAGES
24729           default to the value of the LANG environment variable.  If none of
24730           these variables are set, GCC defaults to traditional C English
24731           behavior.
24732
24733       TMPDIR
24734           If TMPDIR is set, it specifies the directory to use for temporary
24735           files.  GCC uses temporary files to hold the output of one stage of
24736           compilation which is to be used as input to the next stage: for
24737           example, the output of the preprocessor, which is the input to the
24738           compiler proper.
24739
24740       GCC_COMPARE_DEBUG
24741           Setting GCC_COMPARE_DEBUG is nearly equivalent to passing
24742           -fcompare-debug to the compiler driver.  See the documentation of
24743           this option for more details.
24744
24745       GCC_EXEC_PREFIX
24746           If GCC_EXEC_PREFIX is set, it specifies a prefix to use in the
24747           names of the subprograms executed by the compiler.  No slash is
24748           added when this prefix is combined with the name of a subprogram,
24749           but you can specify a prefix that ends with a slash if you wish.
24750
24751           If GCC_EXEC_PREFIX is not set, GCC attempts to figure out an
24752           appropriate prefix to use based on the pathname it is invoked with.
24753
24754           If GCC cannot find the subprogram using the specified prefix, it
24755           tries looking in the usual places for the subprogram.
24756
24757           The default value of GCC_EXEC_PREFIX is prefix/lib/gcc/ where
24758           prefix is the prefix to the installed compiler. In many cases
24759           prefix is the value of "prefix" when you ran the configure script.
24760
24761           Other prefixes specified with -B take precedence over this prefix.
24762
24763           This prefix is also used for finding files such as crt0.o that are
24764           used for linking.
24765
24766           In addition, the prefix is used in an unusual way in finding the
24767           directories to search for header files.  For each of the standard
24768           directories whose name normally begins with /usr/local/lib/gcc
24769           (more precisely, with the value of GCC_INCLUDE_DIR), GCC tries
24770           replacing that beginning with the specified prefix to produce an
24771           alternate directory name.  Thus, with -Bfoo/, GCC searches foo/bar
24772           just before it searches the standard directory /usr/local/lib/bar.
24773           If a standard directory begins with the configured prefix then the
24774           value of prefix is replaced by GCC_EXEC_PREFIX when looking for
24775           header files.
24776
24777       COMPILER_PATH
24778           The value of COMPILER_PATH is a colon-separated list of
24779           directories, much like PATH.  GCC tries the directories thus
24780           specified when searching for subprograms, if it cannot find the
24781           subprograms using GCC_EXEC_PREFIX.
24782
24783       LIBRARY_PATH
24784           The value of LIBRARY_PATH is a colon-separated list of directories,
24785           much like PATH.  When configured as a native compiler, GCC tries
24786           the directories thus specified when searching for special linker
24787           files, if it cannot find them using GCC_EXEC_PREFIX.  Linking using
24788           GCC also uses these directories when searching for ordinary
24789           libraries for the -l option (but directories specified with -L come
24790           first).
24791
24792       LANG
24793           This variable is used to pass locale information to the compiler.
24794           One way in which this information is used is to determine the
24795           character set to be used when character literals, string literals
24796           and comments are parsed in C and C++.  When the compiler is
24797           configured to allow multibyte characters, the following values for
24798           LANG are recognized:
24799
24800           C-JIS
24801               Recognize JIS characters.
24802
24803           C-SJIS
24804               Recognize SJIS characters.
24805
24806           C-EUCJP
24807               Recognize EUCJP characters.
24808
24809           If LANG is not defined, or if it has some other value, then the
24810           compiler uses "mblen" and "mbtowc" as defined by the default locale
24811           to recognize and translate multibyte characters.
24812
24813       Some additional environment variables affect the behavior of the
24814       preprocessor.
24815
24816       CPATH
24817       C_INCLUDE_PATH
24818       CPLUS_INCLUDE_PATH
24819       OBJC_INCLUDE_PATH
24820           Each variable's value is a list of directories separated by a
24821           special character, much like PATH, in which to look for header
24822           files.  The special character, "PATH_SEPARATOR", is target-
24823           dependent and determined at GCC build time.  For Microsoft Windows-
24824           based targets it is a semicolon, and for almost all other targets
24825           it is a colon.
24826
24827           CPATH specifies a list of directories to be searched as if
24828           specified with -I, but after any paths given with -I options on the
24829           command line.  This environment variable is used regardless of
24830           which language is being preprocessed.
24831
24832           The remaining environment variables apply only when preprocessing
24833           the particular language indicated.  Each specifies a list of
24834           directories to be searched as if specified with -isystem, but after
24835           any paths given with -isystem options on the command line.
24836
24837           In all these variables, an empty element instructs the compiler to
24838           search its current working directory.  Empty elements can appear at
24839           the beginning or end of a path.  For instance, if the value of
24840           CPATH is ":/special/include", that has the same effect as
24841           -I. -I/special/include.
24842
24843       DEPENDENCIES_OUTPUT
24844           If this variable is set, its value specifies how to output
24845           dependencies for Make based on the non-system header files
24846           processed by the compiler.  System header files are ignored in the
24847           dependency output.
24848
24849           The value of DEPENDENCIES_OUTPUT can be just a file name, in which
24850           case the Make rules are written to that file, guessing the target
24851           name from the source file name.  Or the value can have the form
24852           file target, in which case the rules are written to file file using
24853           target as the target name.
24854
24855           In other words, this environment variable is equivalent to
24856           combining the options -MM and -MF, with an optional -MT switch too.
24857
24858       SUNPRO_DEPENDENCIES
24859           This variable is the same as DEPENDENCIES_OUTPUT (see above),
24860           except that system header files are not ignored, so it implies -M
24861           rather than -MM.  However, the dependence on the main input file is
24862           omitted.
24863
24864       SOURCE_DATE_EPOCH
24865           If this variable is set, its value specifies a UNIX timestamp to be
24866           used in replacement of the current date and time in the "__DATE__"
24867           and "__TIME__" macros, so that the embedded timestamps become
24868           reproducible.
24869
24870           The value of SOURCE_DATE_EPOCH must be a UNIX timestamp, defined as
24871           the number of seconds (excluding leap seconds) since 01 Jan 1970
24872           00:00:00 represented in ASCII; identical to the output of
24873           @command{date +%s} on GNU/Linux and other systems that support the
24874           %s extension in the "date" command.
24875
24876           The value should be a known timestamp such as the last modification
24877           time of the source or package and it should be set by the build
24878           process.
24879

BUGS

24881       For instructions on reporting bugs, see <https://bugzilla.redhat.com/>.
24882

FOOTNOTES

24884       1.  On some systems, gcc -shared needs to build supplementary stub code
24885           for constructors to work.  On multi-libbed systems, gcc -shared
24886           must select the correct support libraries to link against.  Failing
24887           to supply the correct flags may lead to subtle defects.  Supplying
24888           them in cases where they are not necessary is innocuous.
24889

SEE ALSO

24891       gpl(7), gfdl(7), fsf-funding(7), cpp(1), gcov(1), as(1), ld(1), gdb(1),
24892       dbx(1) and the Info entries for gcc, cpp, as, ld, binutils and gdb.
24893

AUTHOR

24895       See the Info entry for gcc, or
24896       <http://gcc.gnu.org/onlinedocs/gcc/Contributors.html>, for contributors
24897       to GCC.
24898
24900       Copyright (c) 1988-2019 Free Software Foundation, Inc.
24901
24902       Permission is granted to copy, distribute and/or modify this document
24903       under the terms of the GNU Free Documentation License, Version 1.3 or
24904       any later version published by the Free Software Foundation; with the
24905       Invariant Sections being "GNU General Public License" and "Funding Free
24906       Software", the Front-Cover texts being (a) (see below), and with the
24907       Back-Cover Texts being (b) (see below).  A copy of the license is
24908       included in the gfdl(7) man page.
24909
24910       (a) The FSF's Front-Cover Text is:
24911
24912            A GNU Manual
24913
24914       (b) The FSF's Back-Cover Text is:
24915
24916            You have freedom to copy and modify this GNU Manual, like GNU
24917            software.  Copies published by the Free Software Foundation raise
24918            funds for GNU development.
24919
24920
24921
24922gcc-9.2.0                         2019-08-12                            GCC(1)
Impressum