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 -D<VAR=VALUE>
82 Sets a filecheck variable VAR with value VALUE that can be used
83 in CHECK: lines.
84 -version
86 Show the version number of this program.
87
89 If FileCheck verifies that the file matches the expected contents, it
90 exits with 0. Otherwise, if not, or if an error occurs, it will exit
91 with a non-zero value.
92
94 FileCheck is typically used from LLVM regression tests, being invoked
95 on the RUN line of the test. A simple example of using FileCheck from
96 a RUN line looks like this:
97 ; RUN: llvm-as < %s | llc -march=x86-64 | FileCheck %s
99 This syntax says to pipe the current file ("%s") into llvm-as, pipe
101 that into llc, then pipe the output of llc into FileCheck. This means
102 that FileCheck will be verifying its standard input (the llc output)
103 against the filename argument specified (the original .ll file speci‐
104 fied by "%s"). To see how this works, let's look at the rest of the
105 .ll file (after the RUN line):
106 define void @sub1(i32* %p, i32 %v) {
108 entry:
109 ; CHECK: sub1:
110 ; CHECK: subl
111 %0 = tail call i32 @llvm.atomic.load.sub.i32.p0i32(i32* %p, i32 %v)
112 ret void
113 }
114 define void @inc4(i64* %p) {
116 entry:
117 ; CHECK: inc4:
118 ; CHECK: incq
119 %0 = tail call i64 @llvm.atomic.load.add.i64.p0i64(i64* %p, i64 1)
120 ret void
121 }
122 Here you can see some "CHECK:" lines specified in comments. Now you
124 can see how the file is piped into llvm-as, then llc, and the machine
125 code output is what we are verifying. FileCheck checks the machine
126 code output to verify that it matches what the "CHECK:" lines specify.
127 The syntax of the "CHECK:" lines is very simple: they are fixed strings
129 that must occur in order. FileCheck defaults to ignoring horizontal
130 whitespace differences (e.g. a space is allowed to match a tab) but
131 otherwise, the contents of the "CHECK:" line is required to match some
132 thing in the test file exactly.
133 One nice thing about FileCheck (compared to grep) is that it allows
135 merging test cases together into logical groups. For example, because
136 the test above is checking for the "sub1:" and "inc4:" labels, it will
137 not match unless there is a "subl" in between those labels. If it
138 existed somewhere else in the file, that would not count: "grep subl"
139 matches if "subl" exists anywhere in the file.
140 The FileCheck -check-prefix option
142 The FileCheck -check-prefix option allows multiple test configurations
143 to be driven from one .ll file. This is useful in many circumstances,
144 for example, testing different architectural variants with llc. Here's
145 a simple example:
146 ; RUN: llvm-as < %s | llc -mtriple=i686-apple-darwin9 -mattr=sse41 \
148 ; RUN: | FileCheck %s -check-prefix=X32
149 ; RUN: llvm-as < %s | llc -mtriple=x86_64-apple-darwin9 -mattr=sse41 \
150 ; RUN: | FileCheck %s -check-prefix=X64
151 define <4 x i32> @pinsrd_1(i32 %s, <4 x i32> %tmp) nounwind {
153 %tmp1 = insertelement <4 x i32>; %tmp, i32 %s, i32 1
154 ret <4 x i32> %tmp1
155 ; X32: pinsrd_1:
156 ; X32: pinsrd $1, 4(%esp), %xmm0
157 ; X64: pinsrd_1:
159 ; X64: pinsrd $1, %edi, %xmm0
160 }
161 In this case, we're testing that we get the expected code generation
163 with both 32-bit and 64-bit code generation.
164 The CHECK-NEXT: directive
166 Sometimes you want to match lines and would like to verify that matches
167 happen on exactly consecutive lines with no other lines in between
168 them. In this case, you can use "CHECK:" and "CHECK-NEXT:" directives
169 to specify this. If you specified a custom check prefix, just use
170 "<PREFIX>-NEXT:". For example, something like this works as you'd
171 expect:
172 define void @t2(<2 x double>* %r, <2 x double>* %A, double %B) {
174 %tmp3 = load <2 x double>* %A, align 16
175 %tmp7 = insertelement <2 x double> undef, double %B, i32 0
176 %tmp9 = shufflevector <2 x double> %tmp3,
177 <2 x double> %tmp7,
178 <2 x i32> < i32 0, i32 2 >
179 store <2 x double> %tmp9, <2 x double>* %r, align 16
180 ret void
181 ; CHECK: t2:
183 ; CHECK: movl 8(%esp), %eax
184 ; CHECK-NEXT: movapd (%eax), %xmm0
185 ; CHECK-NEXT: movhpd 12(%esp), %xmm0
186 ; CHECK-NEXT: movl 4(%esp), %eax
187 ; CHECK-NEXT: movapd %xmm0, (%eax)
188 ; CHECK-NEXT: ret
189 }
190 "CHECK-NEXT:" directives reject the input unless there is exactly one
192 newline between it and the previous directive. A "CHECK-NEXT:" cannot
193 be the first directive in a file.
194 The CHECK-SAME: directive
196 Sometimes you want to match lines and would like to verify that matches
197 happen on the same line as the previous match. In this case, you can
198 use "CHECK:" and "CHECK-SAME:" directives to specify this. If you
199 specified a custom check prefix, just use "<PREFIX>-SAME:".
200 "CHECK-SAME:" is particularly powerful in conjunction with "CHECK-NOT:"
202 (described below).
203 For example, the following works like you'd expect:
205 !0 = !DILocation(line: 5, scope: !1, inlinedAt: !2)
207 ; CHECK: !DILocation(line: 5,
209 ; CHECK-NOT: column:
210 ; CHECK-SAME: scope: ![[SCOPE:[0-9]+]]
211 "CHECK-SAME:" directives reject the input if there are any newlines
213 between it and the previous directive. A "CHECK-SAME:" cannot be the
214 first directive in a file.
215 The CHECK-NOT: directive
217 The "CHECK-NOT:" directive is used to verify that a string doesn't
218 occur between two matches (or before the first match, or after the last
219 match). For example, to verify that a load is removed by a transforma‐
220 tion, a test like this can be used:
221 define i8 @coerce_offset0(i32 %V, i32* %P) {
223 store i32 %V, i32* %P
224 %P2 = bitcast i32* %P to i8*
226 %P3 = getelementptr i8* %P2, i32 2
227 %A = load i8* %P3
229 ret i8 %A
230 ; CHECK: @coerce_offset0
231 ; CHECK-NOT: load
232 ; CHECK: ret i8
233 }
234 The CHECK-DAG: directive
236 If it's necessary to match strings that don't occur in a strictly
237 sequential order, "CHECK-DAG:" could be used to verify them between two
238 matches (or before the first match, or after the last match). For exam‐
239 ple, clang emits vtable globals in reverse order. Using CHECK-DAG:, we
240 can keep the checks in the natural order:
241 // RUN: %clang_cc1 %s -emit-llvm -o - | FileCheck %s
243 struct Foo { virtual void method(); };
245 Foo f; // emit vtable
246 // CHECK-DAG: @_ZTV3Foo =
247 struct Bar { virtual void method(); };
249 Bar b;
250 // CHECK-DAG: @_ZTV3Bar =
251 CHECK-NOT: directives could be mixed with CHECK-DAG: directives to
253 exclude strings between the surrounding CHECK-DAG: directives. As a
254 result, the surrounding CHECK-DAG: directives cannot be reordered, i.e.
255 all occurrences matching CHECK-DAG: before CHECK-NOT: must not fall
256 behind occurrences matching CHECK-DAG: after CHECK-NOT:. For example,
257 ; CHECK-DAG: BEFORE
259 ; CHECK-NOT: NOT
260 ; CHECK-DAG: AFTER
261 This case will reject input strings where BEFORE occurs after AFTER.
263 With captured variables, CHECK-DAG: is able to match valid topological
265 orderings of a DAG with edges from the definition of a variable to its
266 use. It's useful, e.g., when your test cases need to match different
267 output sequences from the instruction scheduler. For example,
268 ; CHECK-DAG: add [[REG1:r[0-9]+]], r1, r2
270 ; CHECK-DAG: add [[REG2:r[0-9]+]], r3, r4
271 ; CHECK: mul r5, [[REG1]], [[REG2]]
272 In this case, any order of that two add instructions will be allowed.
274 If you are defining and using variables in the same CHECK-DAG: block,
276 be aware that the definition rule can match after its use.
277 So, for instance, the code below will pass:
279 ; CHECK-DAG: vmov.32 [[REG2:d[0-9]+]][0]
281 ; CHECK-DAG: vmov.32 [[REG2]][1]
282 vmov.32 d0[1]
283 vmov.32 d0[0]
284 While this other code, will not:
286 ; CHECK-DAG: vmov.32 [[REG2:d[0-9]+]][0]
288 ; CHECK-DAG: vmov.32 [[REG2]][1]
289 vmov.32 d1[1]
290 vmov.32 d0[0]
291 While this can be very useful, it's also dangerous, because in the case
293 of register sequence, you must have a strong order (read before write,
294 copy before use, etc). If the definition your test is looking for
295 doesn't match (because of a bug in the compiler), it may match further
296 away from the use, and mask real bugs away.
297 In those cases, to enforce the order, use a non-DAG directive between
299 DAG-blocks.
300 The CHECK-LABEL: directive
302 Sometimes in a file containing multiple tests divided into logical
303 blocks, one or more CHECK: directives may inadvertently succeed by
304 matching lines in a later block. While an error will usually eventually
305 be generated, the check flagged as causing the error may not actually
306 bear any relationship to the actual source of the problem.
307 In order to produce better error messages in these cases, the
309 "CHECK-LABEL:" directive can be used. It is treated identically to a
310 normal CHECK directive except that FileCheck makes an additional
311 assumption that a line matched by the directive cannot also be matched
312 by any other check present in match-filename; this is intended to be
313 used for lines containing labels or other unique identifiers. Conceptu‐
314 ally, the presence of CHECK-LABEL divides the input stream into sepa‐
315 rate blocks, each of which is processed independently, preventing a
316 CHECK: directive in one block matching a line in another block. If
317 --enable-var-scope is in effect, all local variables are cleared at the
318 beginning of the block.
319 For example,
321 define %struct.C* @C_ctor_base(%struct.C* %this, i32 %x) {
323 entry:
324 ; CHECK-LABEL: C_ctor_base:
325 ; CHECK: mov [[SAVETHIS:r[0-9]+]], r0
326 ; CHECK: bl A_ctor_base
327 ; CHECK: mov r0, [[SAVETHIS]]
328 %0 = bitcast %struct.C* %this to %struct.A*
329 %call = tail call %struct.A* @A_ctor_base(%struct.A* %0)
330 %1 = bitcast %struct.C* %this to %struct.B*
331 %call2 = tail call %struct.B* @B_ctor_base(%struct.B* %1, i32 %x)
332 ret %struct.C* %this
333 }
334 define %struct.D* @D_ctor_base(%struct.D* %this, i32 %x) {
336 entry:
337 ; CHECK-LABEL: D_ctor_base:
338 The use of CHECK-LABEL: directives in this case ensures that the three
340 CHECK: directives only accept lines corresponding to the body of the
341 @C_ctor_base function, even if the patterns match lines found later in
342 the file. Furthermore, if one of these three CHECK: directives fail,
343 FileCheck will recover by continuing to the next block, allowing multi‐
344 ple test failures to be detected in a single invocation.
345 There is no requirement that CHECK-LABEL: directives contain strings
347 that correspond to actual syntactic labels in a source or output lan‐
348 guage: they must simply uniquely match a single line in the file being
349 verified.
350 CHECK-LABEL: directives cannot contain variable definitions or uses.
352 FileCheck Pattern Matching Syntax
354 All FileCheck directives take a pattern to match. For most uses of
355 FileCheck, fixed string matching is perfectly sufficient. For some
356 things, a more flexible form of matching is desired. To support this,
357 FileCheck allows you to specify regular expressions in matching
358 strings, surrounded by double braces: {{yourregex}}. FileCheck imple‐
359 ments a POSIX regular expression matcher; it supports Extended POSIX
360 regular expressions (ERE). Because we want to use fixed string matching
361 for a majority of what we do, FileCheck has been designed to support
362 mixing and matching fixed string matching with regular expressions.
363 This allows you to write things like this:
364 ; CHECK: movhpd {{[0-9]+}}(%esp), {{%xmm[0-7]}}
366 In this case, any offset from the ESP register will be allowed, and any
368 xmm register will be allowed.
369 Because regular expressions are enclosed with double braces, they are
371 visually distinct, and you don't need to use escape characters within
372 the double braces like you would in C. In the rare case that you want
373 to match double braces explicitly from the input, you can use something
374 ugly like {{[{][{]}} as your pattern.
375 FileCheck Variables
377 It is often useful to match a pattern and then verify that it occurs
378 again later in the file. For codegen tests, this can be useful to
379 allow any register, but verify that that register is used consistently
380 later. To do this, FileCheck allows named variables to be defined and
381 substituted into patterns. Here is a simple example:
382 ; CHECK: test5:
384 ; CHECK: notw [[REGISTER:%[a-z]+]]
385 ; CHECK: andw {{.*}}[[REGISTER]]
386 The first check line matches a regex %[a-z]+ and captures it into the
388 variable REGISTER. The second line verifies that whatever is in REGIS‐
389 TER occurs later in the file after an "andw". FileCheck variable ref‐
390 erences are always contained in [[ ]] pairs, and their names can be
391 formed with the regex [a-zA-Z_][a-zA-Z0-9_]*. If a colon follows the
392 name, then it is a definition of the variable; otherwise, it is a use.
393 FileCheck variables can be defined multiple times, and uses always get
395 the latest value. Variables can also be used later on the same line
396 they were defined on. For example:
397 ; CHECK: op [[REG:r[0-9]+]], [[REG]]
399 Can be useful if you want the operands of op to be the same register,
401 and don't care exactly which register it is.
402 If --enable-var-scope is in effect, variables with names that start
404 with $ are considered to be global. All others variables are local.
405 All local variables get undefined at the beginning of each CHECK-LABEL
406 block. Global variables are not affected by CHECK-LABEL. This makes it
407 easier to ensure that individual tests are not affected by variables
408 set in preceding tests.
409 FileCheck Expressions
411 Sometimes there's a need to verify output which refers line numbers of
412 the match file, e.g. when testing compiler diagnostics. This intro‐
413 duces a certain fragility of the match file structure, as "CHECK:"
414 lines contain absolute line numbers in the same file, which have to be
415 updated whenever line numbers change due to text addition or deletion.
416 To support this case, FileCheck allows using [[@LINE]], [[@LINE+<off‐
418 set>]], [[@LINE-<offset>]] expressions in patterns. These expressions
419 expand to a number of the line where a pattern is located (with an
420 optional integer offset).
421 This way match patterns can be put near the relevant test lines and
423 include relative line number references, for example:
424 // CHECK: test.cpp:[[@LINE+4]]:6: error: expected ';' after top level declarator
426 // CHECK-NEXT: {{^int a}}
427 // CHECK-NEXT: {{^ \^}}
428 // CHECK-NEXT: {{^ ;}}
429 int a
430 Matching Newline Characters
432 To match newline characters in regular expressions the character class
433 [[:space:]] can be used. For example, the following pattern:
434 // CHECK: DW_AT_location [DW_FORM_sec_offset] ([[DLOC:0x[0-9a-f]+]]){{[[:space:]].*}}"intd"
436 matches output of the form (from llvm-dwarfdump):
438 DW_AT_location [DW_FORM_sec_offset] (0x00000233)
440 DW_AT_name [DW_FORM_strp] ( .debug_str[0x000000c9] = "intd")
441 letting us set the FileCheck variable DLOC to the desired value
443 0x00000233, extracted from the line immediately preceding "intd".
444
446 Maintained by The LLVM Team (http://llvm.org/).
447
449 2003-2019, LLVM Project
4506 2019-07-25 FILECHECK(1)