1FILECHECK(1) LLVM FILECHECK(1)
2
6 FileCheck - Flexible pattern matching file verifier
7
9 FileCheck match-filename [--check-prefix=XXX] [--strict-whitespace]
10
12 FileCheck reads two files (one from standard input, and one specified
13 on the command line) and uses one to verify the other. This behavior
14 is particularly useful for the testsuite, which wants to verify that
15 the output of some tool (e.g. llc) contains the expected information
16 (for example, a movsd from esp or whatever is interesting). This is
17 similar to using grep, but it is optimized for matching multiple dif‐
18 ferent inputs in one file in a specific order.
19 The match-filename file specifies the file that contains the patterns
21 to match. The file to verify is read from standard input unless the
22 --input-file option is used.
23
25 -help Print a summary of command line options.
26 --check-prefix prefix
28 FileCheck searches the contents of match-filename for patterns
29 to match. By default, these patterns are prefixed with
30 "CHECK:". If you'd like to use a different prefix (e.g. because
31 the same input file is checking multiple different tool or
32 options), the --check-prefix argument allows you to specify one
33 or more prefixes to match. Multiple prefixes are useful for
34 tests which might change for different run options, but most
35 lines remain the same.
36 --check-prefixes prefix1,prefix2,...
38 An alias of --check-prefix that allows multiple prefixes to be
39 specified as a comma separated list.
40 --input-file filename
42 File to check (defaults to stdin).
43 --match-full-lines
45 By default, FileCheck allows matches of anywhere on a line. This
46 option will require all positive matches to cover an entire
47 line. Leading and trailing whitespace is ignored, unless
48 --strict-whitespace is also specified. (Note: negative matches
49 from CHECK-NOT are not affected by this option!)
50 Passing this option is equivalent to inserting {{^ *}} or {{^}}
52 before, and {{ *$}} or {{$}} after every positive check pattern.
53 --strict-whitespace
55 By default, FileCheck canonicalizes input horizontal whitespace
56 (spaces and tabs) which causes it to ignore these differences (a
57 space will match a tab). The --strict-whitespace argument dis‐
58 ables this behavior. End-of-line sequences are canonicalized to
59 UNIX-style \n in all modes.
60 --implicit-check-not check-pattern
62 Adds implicit negative checks for the specified patterns between
63 positive checks. The option allows writing stricter tests with‐
64 out stuffing them with CHECK-NOTs.
65 For example, "--implicit-check-not warning:" can be useful when
67 testing diagnostic messages from tools that don't have an option
68 similar to clang -verify. With this option FileCheck will verify
69 that input does not contain warnings not covered by any CHECK:
70 patterns.
71 --enable-var-scope
73 Enables scope for regex variables.
74 Variables with names that start with $ are considered global and
76 remain set throughout the file.
77 All other variables get undefined after each encountered
79 CHECK-LABEL.
80 -version
82 Show the version number of this program.
83
85 If FileCheck verifies that the file matches the expected contents, it
86 exits with 0. Otherwise, if not, or if an error occurs, it will exit
87 with a non-zero value.
88
90 FileCheck is typically used from LLVM regression tests, being invoked
91 on the RUN line of the test. A simple example of using FileCheck from
92 a RUN line looks like this:
93 ; RUN: llvm-as < %s | llc -march=x86-64 | FileCheck %s
95 This syntax says to pipe the current file ("%s") into llvm-as, pipe
97 that into llc, then pipe the output of llc into FileCheck. This means
98 that FileCheck will be verifying its standard input (the llc output)
99 against the filename argument specified (the original .ll file speci‐
100 fied by "%s"). To see how this works, let's look at the rest of the
101 .ll file (after the RUN line):
102 define void @sub1(i32* %p, i32 %v) {
104 entry:
105 ; CHECK: sub1:
106 ; CHECK: subl
107 %0 = tail call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %p, i32 %v)
108 ret void
109 }
110 define void @inc4(i64* %p) {
112 entry:
113 ; CHECK: inc4:
114 ; CHECK: incq
115 %0 = tail call i64 @llvm.atomic.load.add.i64.p0i64(i64* %p, i64 1)
116 ret void
117 }
118 Here you can see some "CHECK:" lines specified in comments. Now you
120 can see how the file is piped into llvm-as, then llc, and the machine
121 code output is what we are verifying. FileCheck checks the machine
122 code output to verify that it matches what the "CHECK:" lines specify.
123 The syntax of the "CHECK:" lines is very simple: they are fixed strings
125 that must occur in order. FileCheck defaults to ignoring horizontal
126 whitespace differences (e.g. a space is allowed to match a tab) but
127 otherwise, the contents of the "CHECK:" line is required to match some
128 thing in the test file exactly.
129 One nice thing about FileCheck (compared to grep) is that it allows
131 merging test cases together into logical groups. For example, because
132 the test above is checking for the "sub1:" and "inc4:" labels, it will
133 not match unless there is a "subl" in between those labels. If it
134 existed somewhere else in the file, that would not count: "grep subl"
135 matches if "subl" exists anywhere in the file.
136 The FileCheck -check-prefix option
138 The FileCheck -check-prefix option allows multiple test configurations
139 to be driven from one .ll file. This is useful in many circumstances,
140 for example, testing different architectural variants with llc. Here's
141 a simple example:
142 ; RUN: llvm-as < %s | llc -mtriple=i686-apple-darwin9 -mattr=sse41 \
144 ; RUN: | FileCheck %s -check-prefix=X32
145 ; RUN: llvm-as < %s | llc -mtriple=x86_64-apple-darwin9 -mattr=sse41 \
146 ; RUN: | FileCheck %s -check-prefix=X64
147 define <4 x i32> @pinsrd_1(i32 %s, <4 x i32> %tmp) nounwind {
149 %tmp1 = insertelement <4 x i32>; %tmp, i32 %s, i32 1
150 ret <4 x i32> %tmp1
151 ; X32: pinsrd_1:
152 ; X32: pinsrd $1, 4(%esp), %xmm0
153 ; X64: pinsrd_1:
155 ; X64: pinsrd $1, %edi, %xmm0
156 }
157 In this case, we're testing that we get the expected code generation
159 with both 32-bit and 64-bit code generation.
160 The CHECK-NEXT: directive
162 Sometimes you want to match lines and would like to verify that matches
163 happen on exactly consecutive lines with no other lines in between
164 them. In this case, you can use "CHECK:" and "CHECK-NEXT:" directives
165 to specify this. If you specified a custom check prefix, just use
166 "<PREFIX>-NEXT:". For example, something like this works as you'd
167 expect:
168 define void @t2(<2 x double>* %r, <2 x double>* %A, double %B) {
170 %tmp3 = load <2 x double>* %A, align 16
171 %tmp7 = insertelement <2 x double> undef, double %B, i32 0
172 %tmp9 = shufflevector <2 x double> %tmp3,
173 <2 x double> %tmp7,
174 <2 x i32> < i32 0, i32 2 >
175 store <2 x double> %tmp9, <2 x double>* %r, align 16
176 ret void
177 ; CHECK: t2:
179 ; CHECK: movl 8(%esp), %eax
180 ; CHECK-NEXT: movapd (%eax), %xmm0
181 ; CHECK-NEXT: movhpd 12(%esp), %xmm0
182 ; CHECK-NEXT: movl 4(%esp), %eax
183 ; CHECK-NEXT: movapd %xmm0, (%eax)
184 ; CHECK-NEXT: ret
185 }
186 "CHECK-NEXT:" directives reject the input unless there is exactly one
188 newline between it and the previous directive. A "CHECK-NEXT:" cannot
189 be the first directive in a file.
190 The CHECK-SAME: directive
192 Sometimes you want to match lines and would like to verify that matches
193 happen on the same line as the previous match. In this case, you can
194 use "CHECK:" and "CHECK-SAME:" directives to specify this. If you
195 specified a custom check prefix, just use "<PREFIX>-SAME:".
196 "CHECK-SAME:" is particularly powerful in conjunction with "CHECK-NOT:"
198 (described below).
199 For example, the following works like you'd expect:
201 !0 = !DILocation(line: 5, scope: !1, inlinedAt: !2)
203 ; CHECK: !DILocation(line: 5,
205 ; CHECK-NOT: column:
206 ; CHECK-SAME: scope: ![[SCOPE:[0-9]+]]
207 "CHECK-SAME:" directives reject the input if there are any newlines
209 between it and the previous directive. A "CHECK-SAME:" cannot be the
210 first directive in a file.
211 The CHECK-NOT: directive
213 The "CHECK-NOT:" directive is used to verify that a string doesn't
214 occur between two matches (or before the first match, or after the last
215 match). For example, to verify that a load is removed by a transforma‐
216 tion, a test like this can be used:
217 define i8 @coerce_offset0(i32 %V, i32* %P) {
219 store i32 %V, i32* %P
220 %P2 = bitcast i32* %P to i8*
222 %P3 = getelementptr i8* %P2, i32 2
223 %A = load i8* %P3
225 ret i8 %A
226 ; CHECK: @coerce_offset0
227 ; CHECK-NOT: load
228 ; CHECK: ret i8
229 }
230 The CHECK-DAG: directive
232 If it's necessary to match strings that don't occur in a strictly
233 sequential order, "CHECK-DAG:" could be used to verify them between two
234 matches (or before the first match, or after the last match). For exam‐
235 ple, clang emits vtable globals in reverse order. Using CHECK-DAG:, we
236 can keep the checks in the natural order:
237 // RUN: %clang_cc1 %s -emit-llvm -o - | FileCheck %s
239 struct Foo { virtual void method(); };
241 Foo f; // emit vtable
242 // CHECK-DAG: @_ZTV3Foo =
243 struct Bar { virtual void method(); };
245 Bar b;
246 // CHECK-DAG: @_ZTV3Bar =
247 CHECK-NOT: directives could be mixed with CHECK-DAG: directives to
249 exclude strings between the surrounding CHECK-DAG: directives. As a
250 result, the surrounding CHECK-DAG: directives cannot be reordered, i.e.
251 all occurrences matching CHECK-DAG: before CHECK-NOT: must not fall
252 behind occurrences matching CHECK-DAG: after CHECK-NOT:. For example,
253 ; CHECK-DAG: BEFORE
255 ; CHECK-NOT: NOT
256 ; CHECK-DAG: AFTER
257 This case will reject input strings where BEFORE occurs after AFTER.
259 With captured variables, CHECK-DAG: is able to match valid topological
261 orderings of a DAG with edges from the definition of a variable to its
262 use. It's useful, e.g., when your test cases need to match different
263 output sequences from the instruction scheduler. For example,
264 ; CHECK-DAG: add [[REG1:r[0-9]+]], r1, r2
266 ; CHECK-DAG: add [[REG2:r[0-9]+]], r3, r4
267 ; CHECK: mul r5, [[REG1]], [[REG2]]
268 In this case, any order of that two add instructions will be allowed.
270 If you are defining and using variables in the same CHECK-DAG: block,
272 be aware that the definition rule can match after its use.
273 So, for instance, the code below will pass:
275 ; CHECK-DAG: vmov.32 [[REG2:d[0-9]+]][0]
277 ; CHECK-DAG: vmov.32 [[REG2]][1]
278 vmov.32 d0[1]
279 vmov.32 d0[0]
280 While this other code, will not:
282 ; CHECK-DAG: vmov.32 [[REG2:d[0-9]+]][0]
284 ; CHECK-DAG: vmov.32 [[REG2]][1]
285 vmov.32 d1[1]
286 vmov.32 d0[0]
287 While this can be very useful, it's also dangerous, because in the case
289 of register sequence, you must have a strong order (read before write,
290 copy before use, etc). If the definition your test is looking for
291 doesn't match (because of a bug in the compiler), it may match further
292 away from the use, and mask real bugs away.
293 In those cases, to enforce the order, use a non-DAG directive between
295 DAG-blocks.
296 The CHECK-LABEL: directive
298 Sometimes in a file containing multiple tests divided into logical
299 blocks, one or more CHECK: directives may inadvertently succeed by
300 matching lines in a later block. While an error will usually eventually
301 be generated, the check flagged as causing the error may not actually
302 bear any relationship to the actual source of the problem.
303 In order to produce better error messages in these cases, the
305 "CHECK-LABEL:" directive can be used. It is treated identically to a
306 normal CHECK directive except that FileCheck makes an additional
307 assumption that a line matched by the directive cannot also be matched
308 by any other check present in match-filename; this is intended to be
309 used for lines containing labels or other unique identifiers. Conceptu‐
310 ally, the presence of CHECK-LABEL divides the input stream into sepa‐
311 rate blocks, each of which is processed independently, preventing a
312 CHECK: directive in one block matching a line in another block. If
313 --enable-var-scope is in effect, all local variables are cleared at the
314 beginning of the block.
315 For example,
317 define %struct.C* @C_ctor_base(%struct.C* %this, i32 %x) {
319 entry:
320 ; CHECK-LABEL: C_ctor_base:
321 ; CHECK: mov [[SAVETHIS:r[0-9]+]], r0
322 ; CHECK: bl A_ctor_base
323 ; CHECK: mov r0, [[SAVETHIS]]
324 %0 = bitcast %struct.C* %this to %struct.A*
325 %call = tail call %struct.A* @A_ctor_base(%struct.A* %0)
326 %1 = bitcast %struct.C* %this to %struct.B*
327 %call2 = tail call %struct.B* @B_ctor_base(%struct.B* %1, i32 %x)
328 ret %struct.C* %this
329 }
330 define %struct.D* @D_ctor_base(%struct.D* %this, i32 %x) {
332 entry:
333 ; CHECK-LABEL: D_ctor_base:
334 The use of CHECK-LABEL: directives in this case ensures that the three
336 CHECK: directives only accept lines corresponding to the body of the
337 @C_ctor_base function, even if the patterns match lines found later in
338 the file. Furthermore, if one of these three CHECK: directives fail,
339 FileCheck will recover by continuing to the next block, allowing multi‐
340 ple test failures to be detected in a single invocation.
341 There is no requirement that CHECK-LABEL: directives contain strings
343 that correspond to actual syntactic labels in a source or output lan‐
344 guage: they must simply uniquely match a single line in the file being
345 verified.
346 CHECK-LABEL: directives cannot contain variable definitions or uses.
348 FileCheck Pattern Matching Syntax
350 All FileCheck directives take a pattern to match. For most uses of
351 FileCheck, fixed string matching is perfectly sufficient. For some
352 things, a more flexible form of matching is desired. To support this,
353 FileCheck allows you to specify regular expressions in matching
354 strings, surrounded by double braces: {{yourregex}}. Because we want
355 to use fixed string matching for a majority of what we do, FileCheck
356 has been designed to support mixing and matching fixed string matching
357 with regular expressions. This allows you to write things like this:
358 ; CHECK: movhpd {{[0-9]+}}(%esp), {{%xmm[0-7]}}
360 In this case, any offset from the ESP register will be allowed, and any
362 xmm register will be allowed.
363 Because regular expressions are enclosed with double braces, they are
365 visually distinct, and you don't need to use escape characters within
366 the double braces like you would in C. In the rare case that you want
367 to match double braces explicitly from the input, you can use something
368 ugly like {{[{][{]}} as your pattern.
369 FileCheck Variables
371 It is often useful to match a pattern and then verify that it occurs
372 again later in the file. For codegen tests, this can be useful to
373 allow any register, but verify that that register is used consistently
374 later. To do this, FileCheck allows named variables to be defined and
375 substituted into patterns. Here is a simple example:
376 ; CHECK: test5:
378 ; CHECK: notw [[REGISTER:%[a-z]+]]
379 ; CHECK: andw {{.*}}[[REGISTER]]
380 The first check line matches a regex %[a-z]+ and captures it into the
382 variable REGISTER. The second line verifies that whatever is in REGIS‐
383 TER occurs later in the file after an "andw". FileCheck variable ref‐
384 erences are always contained in [[ ]] pairs, and their names can be
385 formed with the regex [a-zA-Z][a-zA-Z0-9]*. If a colon follows the
386 name, then it is a definition of the variable; otherwise, it is a use.
387 FileCheck variables can be defined multiple times, and uses always get
389 the latest value. Variables can also be used later on the same line
390 they were defined on. For example:
391 ; CHECK: op [[REG:r[0-9]+]], [[REG]]
393 Can be useful if you want the operands of op to be the same register,
395 and don't care exactly which register it is.
396 If --enable-var-scope is in effect, variables with names that start
398 with $ are considered to be global. All others variables are local.
399 All local variables get undefined at the beginning of each CHECK-LABEL
400 block. Global variables are not affected by CHECK-LABEL. This makes it
401 easier to ensure that individual tests are not affected by variables
402 set in preceding tests.
403 FileCheck Expressions
405 Sometimes there's a need to verify output which refers line numbers of
406 the match file, e.g. when testing compiler diagnostics. This intro‐
407 duces a certain fragility of the match file structure, as "CHECK:"
408 lines contain absolute line numbers in the same file, which have to be
409 updated whenever line numbers change due to text addition or deletion.
410 To support this case, FileCheck allows using [[@LINE]], [[@LINE+<off‐
412 set>]], [[@LINE-<offset>]] expressions in patterns. These expressions
413 expand to a number of the line where a pattern is located (with an
414 optional integer offset).
415 This way match patterns can be put near the relevant test lines and
417 include relative line number references, for example:
418 // CHECK: test.cpp:[[@LINE+4]]:6: error: expected ';' after top level declarator
420 // CHECK-NEXT: {{^int a}}
421 // CHECK-NEXT: {{^ \^}}
422 // CHECK-NEXT: {{^ ;}}
423 int a
424 Matching Newline Characters
426 To match newline characters in regular expressions the character class
427 [[:space:]] can be used. For example, the following pattern:
428 // CHECK: DW_AT_location [DW_FORM_sec_offset] ([[DLOC:0x[0-9a-f]+]]){{[[:space:]].*}}"intd"
430 matches output of the form (from llvm-dwarfdump):
432 DW_AT_location [DW_FORM_sec_offset] (0x00000233)
434 DW_AT_name [DW_FORM_strp] ( .debug_str[0x000000c9] = "intd")
435 letting us set the FileCheck variable DLOC to the desired value
437 0x00000233, extracted from the line immediately preceding "intd".
438
440 Maintained by The LLVM Team (http://llvm.org/).
441
443 2003-2019, LLVM Project
4445 2019-02-01 FILECHECK(1)