tests: parameterize --valgrind option
[git/git.git] / t / README
CommitLineData
986aa7f1
JH
1Core GIT Tests
2==============
3
4This directory holds many test scripts for core GIT tools. The
5first part of this short document describes how to run the tests
6and read their output.
7
8When fixing the tools or adding enhancements, you are strongly
9encouraged to add tests in this directory to cover what you are
10trying to fix or enhance. The later part of this short document
11describes how your test scripts should be organized.
12
13
14Running Tests
15-------------
16
17The easiest way to run tests is to say "make". This runs all
18the tests.
19
20 *** t0000-basic.sh ***
5099b99d
ÆAB
21 ok 1 - .git/objects should be empty after git init in an empty repo.
22 ok 2 - .git/objects should have 3 subdirectories.
23 ok 3 - success is reported like this
986aa7f1 24 ...
5099b99d
ÆAB
25 ok 43 - very long name in the index handled sanely
26 # fixed 1 known breakage(s)
27 # still have 1 known breakage(s)
28 # passed all remaining 42 test(s)
29 1..43
30 *** t0001-init.sh ***
31 ok 1 - plain
32 ok 2 - plain with GIT_WORK_TREE
33 ok 3 - plain bare
34
35Since the tests all output TAP (see http://testanything.org) they can
85b0b34e 36be run with any TAP harness. Here's an example of parallel testing
5099b99d
ÆAB
37powered by a recent version of prove(1):
38
39 $ prove --timer --jobs 15 ./t[0-9]*.sh
40 [19:17:33] ./t0005-signals.sh ................................... ok 36 ms
41 [19:17:33] ./t0022-crlf-rename.sh ............................... ok 69 ms
42 [19:17:33] ./t0024-crlf-archive.sh .............................. ok 154 ms
43 [19:17:33] ./t0004-unwritable.sh ................................ ok 289 ms
44 [19:17:33] ./t0002-gitfile.sh ................................... ok 480 ms
45 ===( 102;0 25/? 6/? 5/? 16/? 1/? 4/? 2/? 1/? 3/? 1... )===
46
47prove and other harnesses come with a lot of useful options. The
48--state option in particular is very useful:
49
50 # Repeat until no more failures
51 $ prove -j 15 --state=failed,save ./t[0-9]*.sh
986aa7f1 52
28d836c8
MG
53You can give DEFAULT_TEST_TARGET=prove on the make command (or define it
54in config.mak) to cause "make test" to run tests under prove.
55GIT_PROVE_OPTS can be used to pass additional options, e.g.
56
57 $ make DEFAULT_TEST_TARGET=prove GIT_PROVE_OPTS='--timer --jobs 16' test
58
5099b99d 59You can also run each test individually from command line, like this:
986aa7f1 60
5099b99d
ÆAB
61 $ sh ./t3010-ls-files-killed-modified.sh
62 ok 1 - git update-index --add to add various paths.
63 ok 2 - git ls-files -k to show killed files.
64 ok 3 - validate git ls-files -k output.
65 ok 4 - git ls-files -m to show modified files.
66 ok 5 - validate git ls-files -m output.
67 # passed all 5 test(s)
68 1..5
986aa7f1
JH
69
70You can pass --verbose (or -v), --debug (or -d), and --immediate
4e1be63c
JS
71(or -i) command line argument to the test, or by setting GIT_TEST_OPTS
72appropriately before running "make".
986aa7f1
JH
73
74--verbose::
75 This makes the test more verbose. Specifically, the
76 command being run and their output if any are also
77 output.
78
79--debug::
80 This may help the person who is developing a new test.
81 It causes the command defined with test_debug to run.
0986de94
PK
82 The "trash" directory (used to store all temporary data
83 during testing) is not deleted even if there are no
84 failed tests so that you can inspect its contents after
85 the test finished.
986aa7f1
JH
86
87--immediate::
88 This causes the test to immediately exit upon the first
89 failed test.
90
5e2c08c6
LW
91--long-tests::
92 This causes additional long-running tests to be run (where
93 available), for more exhaustive testing.
94
952af351
TR
95--valgrind=<tool>::
96 Execute all Git binaries under valgrind tool <tool> and exit
97 with status 126 on errors (just like regular tests, this will
98 only stop the test script when running under -i).
986aa7f1 99
3da93652
JS
100 Since it makes no sense to run the tests with --valgrind and
101 not see any output, this option implies --verbose. For
102 convenience, it also implies --tee.
103
952af351
TR
104 <tool> defaults to 'memcheck', just like valgrind itself.
105 Other particularly useful choices include 'helgrind' and
106 'drd', but you may use any tool recognized by your valgrind
107 installation.
108
109 Note that memcheck is run with the option --leak-check=no,
9aec68d3
CMN
110 as the git process is short-lived and some errors are not
111 interesting. In order to run a single command under the same
112 conditions manually, you should set GIT_VALGRIND to point to
113 the 't/valgrind/' directory and use the commands under
114 't/valgrind/bin/'.
115
44138559
JS
116--tee::
117 In addition to printing the test output to the terminal,
118 write it to files named 't/test-results/$TEST_NAME.out'.
119 As the names depend on the tests' file names, it is safe to
120 run the tests with this option in parallel.
121
e4597aae
MO
122--with-dashes::
123 By default tests are run without dashed forms of
124 commands (like git-commit) in the PATH (it only uses
125 wrappers from ../bin-wrappers). Use this option to include
126 the build directory (..) in the PATH, which contains all
127 the dashed forms of commands. This option is currently
128 implied by other options like --valgrind and
129 GIT_TEST_INSTALLED.
130
0d4dbcd3
TR
131--root=<directory>::
132 Create "trash" directories used to store all temporary data during
133 testing under <directory>, instead of the t/ directory.
134 Using this option with a RAM-based filesystem (such as tmpfs)
135 can massively speed up the test suite.
136
e160da7f
MO
137You can also set the GIT_TEST_INSTALLED environment variable to
138the bindir of an existing git installation to test that installation.
139You still need to have built this git sandbox, from which various
140test-* support programs, templates, and perl libraries are used.
141If your installed git is incomplete, it will silently test parts of
142your built version instead.
143
144When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to
145override the location of the dashed-form subcommands (what
146GIT_EXEC_PATH would be used for during normal operation).
147GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`.
148
149
fbd458a3
JN
150Skipping Tests
151--------------
152
153In some environments, certain tests have no way of succeeding
154due to platform limitation, such as lack of 'unzip' program, or
155filesystem that do not allow arbitrary sequence of non-NUL bytes
156as pathnames.
157
158You should be able to say something like
159
160 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh
161
162and even:
163
164 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make
165
166to omit such tests. The value of the environment variable is a
167SP separated list of patterns that tells which tests to skip,
168and either can match the "t[0-9]{4}" part to skip the whole
169test, or t[0-9]{4} followed by ".$number" to say which
170particular test to skip.
171
172Note that some tests in the existing test suite rely on previous
173test item, so you cannot arbitrarily disable one and expect the
174remainder of test to check what the test originally was intended
175to check.
176
177
f50c9f76
PB
178Naming Tests
179------------
180
181The test files are named as:
182
183 tNNNN-commandname-details.sh
184
185where N is a decimal digit.
186
187First digit tells the family:
188
189 0 - the absolute basics and global stuff
190 1 - the basic commands concerning database
191 2 - the basic commands concerning the working tree
192 3 - the other basic commands (e.g. ls-files)
193 4 - the diff commands
194 5 - the pull and exporting commands
195 6 - the revision tree commands (even e.g. merge-base)
8f4a9b62 196 7 - the porcelainish commands concerning the working tree
8757749e
JN
197 8 - the porcelainish commands concerning forensics
198 9 - the git tools
f50c9f76
PB
199
200Second digit tells the particular command we are testing.
201
202Third digit (optionally) tells the particular switch or group of switches
203we are testing.
204
77656600
JH
205If you create files under t/ directory (i.e. here) that is not
206the top-level test script, never name the file to match the above
207pattern. The Makefile here considers all such files as the
63d32945 208top-level test script and tries to run all of them. Care is
77656600
JH
209especially needed if you are creating a common test library
210file, similar to test-lib.sh, because such a library file may
211not be suitable for standalone execution.
212
f50c9f76 213
986aa7f1
JH
214Writing Tests
215-------------
216
217The test script is written as a shell script. It should start
218with the standard "#!/bin/sh" with copyright notices, and an
219assignment to variable 'test_description', like this:
220
221 #!/bin/sh
222 #
223 # Copyright (c) 2005 Junio C Hamano
224 #
225
14cd1ff3 226 test_description='xxx test (option --frotz)
986aa7f1
JH
227
228 This test registers the following structure in the cache
229 and tries to run git-ls-files with option --frotz.'
230
f50c9f76 231
986aa7f1
JH
232Source 'test-lib.sh'
233--------------------
234
235After assigning test_description, the test script should source
236test-lib.sh like this:
237
238 . ./test-lib.sh
239
240This test harness library does the following things:
241
242 - If the script is invoked with command line argument --help
243 (or -h), it shows the test_description and exits.
244
e1ca1c9d
ÆAB
245 - Creates an empty test directory with an empty .git/objects database
246 and chdir(2) into it. This directory is 't/trash
247 directory.$test_name_without_dotsh', with t/ subject to change by
248 the --root option documented above.
986aa7f1
JH
249
250 - Defines standard test helper functions for your scripts to
251 use. These functions are designed to make all scripts behave
252 consistently when command line arguments --verbose (or -v),
253 --debug (or -d), and --immediate (or -i) is given.
254
20873f45
ÆAB
255Do's, don'ts & things to keep in mind
256-------------------------------------
257
6fd45295 258Here are a few examples of things you probably should and shouldn't do
20873f45
ÆAB
259when writing tests.
260
261Do:
262
6fd45295 263 - Put all code inside test_expect_success and other assertions.
20873f45
ÆAB
264
265 Even code that isn't a test per se, but merely some setup code
6fd45295 266 should be inside a test assertion.
20873f45
ÆAB
267
268 - Chain your test assertions
269
270 Write test code like this:
271
272 git merge foo &&
273 git push bar &&
274 test ...
275
276 Instead of:
277
278 git merge hla
279 git push gh
280 test ...
281
282 That way all of the commands in your tests will succeed or fail. If
00648ba0
EN
283 you must ignore the return value of something, consider using a
284 helper function (e.g. use sane_unset instead of unset, in order
285 to avoid unportable return value for unsetting a variable that was
286 already unset), or prepending the command with test_might_fail or
287 test_must_fail.
20873f45 288
0c357544
ÆAB
289 - Check the test coverage for your tests. See the "Test coverage"
290 below.
291
63d32945
MW
292 Don't blindly follow test coverage metrics; if a new function you added
293 doesn't have any coverage, then you're probably doing something wrong,
e8b55f5c
ÆAB
294 but having 100% coverage doesn't necessarily mean that you tested
295 everything.
296
297 Tests that are likely to smoke out future regressions are better
298 than tests that just inflate the coverage metrics.
299
95b104c8
JS
300 - When a test checks for an absolute path that a git command generated,
301 construct the expected value using $(pwd) rather than $PWD,
302 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on
303 Windows, where the shell (MSYS bash) mangles absolute path names.
304 For details, see the commit message of 4114156ae9.
305
20873f45
ÆAB
306Don't:
307
308 - exit() within a <script> part.
309
310 The harness will catch this as a programming error of the test.
311 Use test_done instead if you need to stop the tests early (see
312 "Skipping tests" below).
313
ad78585e
JH
314 - use '! git cmd' when you want to make sure the git command exits
315 with failure in a controlled way by calling "die()". Instead,
316 use 'test_must_fail git cmd'. This will signal a failure if git
317 dies in an unexpected way (e.g. segfault).
318
319 - use perl without spelling it as "$PERL_PATH". This is to help our
320 friends on Windows where the platform Perl often adds CR before
321 the end of line, and they bundle Git with a version of Perl that
322 does not do so, whose path is specified with $PERL_PATH.
323
324 - use sh without spelling it as "$SHELL_PATH", when the script can
325 be misinterpreted by broken platform shell (e.g. Solaris).
326
327 - chdir around in tests. It is not sufficient to chdir to
328 somewhere and then chdir back to the original location later in
329 the test, as any intermediate step can fail and abort the test,
330 causing the next test to start in an unexpected directory. Do so
331 inside a subshell if necessary.
332
20873f45
ÆAB
333 - Break the TAP output
334
6fd45295
JH
335 The raw output from your test may be interpreted by a TAP harness. TAP
336 harnesses will ignore everything they don't know about, but don't step
337 on their toes in these areas:
20873f45
ÆAB
338
339 - Don't print lines like "$x..$y" where $x and $y are integers.
340
341 - Don't print lines that begin with "ok" or "not ok".
342
6fd45295 343 TAP harnesses expect a line that begins with either "ok" and "not
20873f45
ÆAB
344 ok" to signal a test passed or failed (and our harness already
345 produces such lines), so your script shouldn't emit such lines to
346 their output.
347
348 You can glean some further possible issues from the TAP grammar
349 (see http://search.cpan.org/perldoc?TAP::Parser::Grammar#TAP_Grammar)
350 but the best indication is to just run the tests with prove(1),
351 it'll complain if anything is amiss.
352
353Keep in mind:
354
6fd45295 355 - Inside <script> part, the standard output and standard error
20873f45
ÆAB
356 streams are discarded, and the test harness only reports "ok" or
357 "not ok" to the end user running the tests. Under --verbose, they
358 are shown to help debugging the tests.
359
360
b5500d16
ÆAB
361Skipping tests
362--------------
363
681186ae 364If you need to skip tests you should do so by using the three-arg form
99d9050d
ÆAB
365of the test_* functions (see the "Test harness library" section
366below), e.g.:
367
ad78585e
JH
368 test_expect_success PERL 'I need Perl' '
369 "$PERL_PATH" -e "hlagh() if unf_unf()"
370 '
99d9050d
ÆAB
371
372The advantage of skipping tests like this is that platforms that don't
373have the PERL and other optional dependencies get an indication of how
374many tests they're missing.
375
376If the test code is too hairy for that (i.e. does a lot of setup work
377outside test assertions) you can also skip all remaining tests by
378setting skip_all and immediately call test_done:
b5500d16
ÆAB
379
380 if ! test_have_prereq PERL
381 then
382 skip_all='skipping perl interface tests, perl not available'
383 test_done
384 fi
14cd1ff3 385
99d9050d
ÆAB
386The string you give to skip_all will be used as an explanation for why
387the test was skipped.
388
986aa7f1
JH
389End with test_done
390------------------
391
392Your script will be a sequence of tests, using helper functions
393from the test harness library. At the end of the script, call
394'test_done'.
395
396
397Test harness library
398--------------------
399
400There are a handful helper functions defined in the test harness
401library for your script to use.
402
9a897893 403 - test_expect_success [<prereq>] <message> <script>
986aa7f1 404
72942a61 405 Usually takes two strings as parameters, and evaluates the
986aa7f1
JH
406 <script>. If it yields success, test is considered
407 successful. <message> should state what it is testing.
408
409 Example:
410
411 test_expect_success \
412 'git-write-tree should be able to write an empty tree.' \
413 'tree=$(git-write-tree)'
414
9a897893 415 If you supply three parameters the first will be taken to be a
72942a61 416 prerequisite; see the test_set_prereq and test_have_prereq
9a897893
ÆAB
417 documentation below:
418
419 test_expect_success TTY 'git --paginate rev-list uses a pager' \
420 ' ... '
421
93a57246
ÆAB
422 You can also supply a comma-separated list of prerequisites, in the
423 rare case where your test depends on more than one:
424
425 test_expect_success PERL,PYTHON 'yo dawg' \
426 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
427
9a897893 428 - test_expect_failure [<prereq>] <message> <script>
986aa7f1 429
41ac414e
JH
430 This is NOT the opposite of test_expect_success, but is used
431 to mark a test that demonstrates a known breakage. Unlike
432 the usual test_expect_success tests, which say "ok" on
433 success and "FAIL" on failure, this will say "FIXED" on
434 success and "still broken" on failure. Failures from these
435 tests won't cause -i (immediate) to stop.
986aa7f1 436
9a897893
ÆAB
437 Like test_expect_success this function can optionally use a three
438 argument invocation with a prerequisite as the first argument.
439
986aa7f1
JH
440 - test_debug <script>
441
442 This takes a single argument, <script>, and evaluates it only
443 when the test script is started with --debug command line
444 argument. This is primarily meant for use during the
445 development of a new test script.
446
447 - test_done
448
449 Your test script must have test_done at the end. Its purpose
450 is to summarize successes and failures in the test script and
451 exit with an appropriate error code.
452
00884968
JS
453 - test_tick
454
455 Make commit and tag names consistent by setting the author and
63d32945 456 committer times to defined state. Subsequent calls will
00884968
JS
457 advance the times by a fixed amount.
458
459 - test_commit <message> [<filename> [<contents>]]
460
461 Creates a commit with the given message, committing the given
462 file with the given contents (default for both is to reuse the
463 message string), and adds a tag (again reusing the message
464 string as name). Calls test_tick to make the SHA-1s
465 reproducible.
466
467 - test_merge <message> <commit-or-tag>
468
469 Merges the given rev using the given message. Like test_commit,
470 creates a tag and calls test_tick before committing.
986aa7f1 471
72942a61 472 - test_set_prereq <prereq>
9a897893
ÆAB
473
474 Set a test prerequisite to be used later with test_have_prereq. The
be53deef
ÆAB
475 test-lib will set some prerequisites for you, see the
476 "Prerequisites" section below for a full list of these.
477
478 Others you can set yourself and use later with either
479 test_have_prereq directly, or the three argument invocation of
480 test_expect_success and test_expect_failure.
9a897893 481
72942a61 482 - test_have_prereq <prereq>
9a897893
ÆAB
483
484 Check if we have a prerequisite previously set with
485 test_set_prereq. The most common use of this directly is to skip
486 all the tests if we don't have some essential prerequisite:
487
488 if ! test_have_prereq PERL
489 then
490 skip_all='skipping perl interface tests, perl not available'
491 test_done
492 fi
493
2fac6a4b
ÆAB
494 - test_external [<prereq>] <message> <external> <script>
495
496 Execute a <script> with an <external> interpreter (like perl). This
497 was added for tests like t9700-perl-git.sh which do most of their
498 work in an external test script.
499
500 test_external \
501 'GitwebCache::*FileCache*' \
502 "$PERL_PATH" "$TEST_DIRECTORY"/t9503/test_cache_interface.pl
503
504 If the test is outputting its own TAP you should set the
505 test_external_has_tap variable somewhere before calling the first
506 test_external* function. See t9700-perl-git.sh for an example.
507
508 # The external test will outputs its own plan
509 test_external_has_tap=1
510
511 - test_external_without_stderr [<prereq>] <message> <external> <script>
512
513 Like test_external but fail if there's any output on stderr,
514 instead of checking the exit code.
515
516 test_external_without_stderr \
517 'Perl API' \
518 "$PERL_PATH" "$TEST_DIRECTORY"/t9700/test.pl
519
892e6f7e
ÆAB
520 - test_expect_code <exit-code> <command>
521
522 Run a command and ensure that it exits with the given exit code.
523 For example:
524
525 test_expect_success 'Merge with d/f conflicts' '
526 test_expect_code 1 git merge "merge msg" B master
527 '
528
c9667456
JN
529 - test_must_fail <git-command>
530
531 Run a git command and ensure it fails in a controlled way. Use
971ecbd1
BC
532 this instead of "! <git-command>". When git-command dies due to a
533 segfault, test_must_fail diagnoses it as an error; "! <git-command>"
534 treats it as just another expected failure, which would let such a
535 bug go unnoticed.
c9667456
JN
536
537 - test_might_fail <git-command>
538
539 Similar to test_must_fail, but tolerate success, too. Use this
540 instead of "<git-command> || :" to catch failures due to segv.
541
542 - test_cmp <expected> <actual>
543
544 Check whether the content of the <actual> file matches the
545 <expected> file. This behaves like "cmp" but produces more
546 helpful output when the test is run with "-v" option.
547
fb3340a6
JN
548 - test_line_count (= | -lt | -ge | ...) <length> <file>
549
550 Check whether a file has the length it is expected to.
551
72942a61
ML
552 - test_path_is_file <path> [<diagnosis>]
553 test_path_is_dir <path> [<diagnosis>]
2caf20c5
MM
554 test_path_is_missing <path> [<diagnosis>]
555
72942a61
ML
556 Check if the named path is a file, if the named path is a
557 directory, or if the named path does not exist, respectively,
558 and fail otherwise, showing the <diagnosis> text.
2caf20c5 559
c9667456
JN
560 - test_when_finished <script>
561
562 Prepend <script> to a list of commands to run to clean up
563 at the end of the current test. If some clean-up command
564 fails, the test will not pass.
565
566 Example:
567
568 test_expect_success 'branch pointing to non-commit' '
569 git rev-parse HEAD^{tree} >.git/refs/heads/invalid &&
570 test_when_finished "git update-ref -d refs/heads/invalid" &&
571 ...
572 '
573
c4d2539a
JL
574 - test_pause
575
576 This command is useful for writing and debugging tests and must be
577 removed before submitting. It halts the execution of the test and
578 spawns a shell in the trash directory. Exit the shell to continue
579 the test. Example:
580
581 test_expect_success 'test' '
582 git do-something >actual &&
583 test_pause &&
584 test_cmp expected actual
585 '
586
be53deef
ÆAB
587Prerequisites
588-------------
589
590These are the prerequisites that the test library predefines with
591test_have_prereq.
592
593See the prereq argument to the test_* functions in the "Test harness
594library" section above and the "test_have_prereq" function for how to
595use these, and "test_set_prereq" for how to define your own.
596
597 - PERL & PYTHON
598
599 Git wasn't compiled with NO_PERL=YesPlease or
600 NO_PYTHON=YesPlease. Wrap any tests that need Perl or Python in
601 these.
602
603 - POSIXPERM
604
605 The filesystem supports POSIX style permission bits.
606
607 - BSLASHPSPEC
608
609 Backslashes in pathspec are not directory separators. This is not
610 set on Windows. See 6fd1106a for details.
611
612 - EXECKEEPSPID
613
614 The process retains the same pid across exec(2). See fb9a2bea for
615 details.
616
617 - SYMLINKS
618
619 The filesystem we're on supports symbolic links. E.g. a FAT
620 filesystem doesn't support these. See 704a3143 for details.
2fac6a4b 621
c91cfd19
ÆAB
622 - SANITY
623
624 Test is not run by root user, and an attempt to write to an
625 unwritable file is expected to fail correctly.
2fac6a4b 626
8f852ce6
MK
627 - LIBPCRE
628
629 Git was compiled with USE_LIBPCRE=YesPlease. Wrap any tests
630 that use git-grep --perl-regexp or git-grep -P in these.
631
ac39aa61
MG
632 - CASE_INSENSITIVE_FS
633
634 Test is run on a case insensitive file system.
635
5b0b5dd4
MG
636 - UTF8_NFD_TO_NFC
637
638 Test is run on a filesystem which converts decomposed utf-8 (nfd)
639 to precomposed utf-8 (nfc).
640
986aa7f1
JH
641Tips for Writing Tests
642----------------------
643
644As with any programming projects, existing programs are the best
645source of the information. However, do _not_ emulate
646t0000-basic.sh when writing your tests. The test is special in
647that it tries to validate the very core of GIT. For example, it
648knows that there will be 256 subdirectories under .git/objects/,
649and it knows that the object ID of an empty tree is a certain
65040-byte string. This is deliberately done so in t0000-basic.sh
651because the things the very basic core test tries to achieve is
652to serve as a basis for people who are changing the GIT internal
653drastically. For these people, after making certain changes,
654not seeing failures from the basic test _is_ a failure. And
655such drastic changes to the core GIT that even changes these
656otherwise supposedly stable object IDs should be accompanied by
657an update to t0000-basic.sh.
658
659However, other tests that simply rely on basic parts of the core
660GIT working properly should not have that level of intimate
661knowledge of the core GIT internals. If all the test scripts
662hardcoded the object IDs like t0000-basic.sh does, that defeats
663the purpose of t0000-basic.sh, which is to isolate that level of
664validation in one place. Your test also ends up needing
665updating when such a change to the internal happens, so do _not_
666do it and leave the low level of validation to t0000-basic.sh.
d15e9ebc 667
0c357544
ÆAB
668Test coverage
669-------------
670
671You can use the coverage tests to find code paths that are not being
672used or properly exercised yet.
673
674To do that, run the coverage target at the top-level (not in the t/
675directory):
676
677 make coverage
678
679That'll compile Git with GCC's coverage arguments, and generate a test
680report with gcov after the tests finish. Running the coverage tests
681can take a while, since running the tests in parallel is incompatible
682with GCC's coverage mode.
683
684After the tests have run you can generate a list of untested
685functions:
686
687 make coverage-untested-functions
688
689You can also generate a detailed per-file HTML report using the
690Devel::Cover module. To install it do:
691
692 # On Debian or Ubuntu:
693 sudo aptitude install libdevel-cover-perl
694
695 # From the CPAN with cpanminus
696 curl -L http://cpanmin.us | perl - --sudo --self-upgrade
697 cpanm --sudo Devel::Cover
698
699Then, at the top-level:
700
701 make cover_db_html
702
703That'll generate a detailed cover report in the "cover_db_html"
704directory, which you can then copy to a webserver, or inspect locally
705in a browser.