summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorLukas Slebodnik <lslebodn@redhat.com>2014-09-02 15:26:11 +0200
committerJakub Hrozek <jhrozek@redhat.com>2014-09-05 17:25:12 +0200
commit6b104b0f1cf3f74b56a59458f40922919e3eae0c (patch)
tree56d0a809ef2415f88721ab40630788bf6d3d9c6a
parent7ec88ea204c0d1db40a93502c70c7bc0491ec262 (diff)
downloadsssd-6b104b0f1cf3f74b56a59458f40922919e3eae0c.tar.gz
sssd-6b104b0f1cf3f74b56a59458f40922919e3eae0c.tar.xz
sssd-6b104b0f1cf3f74b56a59458f40922919e3eae0c.zip
refcount-tests: Do not force to run test in CK_FORK mode
There was an access to uninitialised data in old version of libcheck when tests were executed in fork mode. ==2842== Syscall param timer_create(evp) points to uninitialised byte(s) ==2842== at 0x977FE72: timer_create@@GLIBC_2.3.3 (timer_create.c:82) ==2842== by 0x4E3647B: srunner_run (in /usr/lib64/libcheck.so.0.0.0) ==2842== by 0x401616: main (refcount-tests.c:233) ==2842== Address 0xffefff460 is on thread 1's stack This bug is already fixed in check-0.9.14, but we want to execute tests with valgrind on older platforms. The refcount test had hardcoded CK_FORK in source code and therefore it could not be overwritten with environment variable, which we use in CI script. Reviewed-by: Nikolai Kondrashov <Nikolai.Kondrashov@redhat.com>
-rw-r--r--src/tests/refcount-tests.c1
1 files changed, 0 insertions, 1 deletions
diff --git a/src/tests/refcount-tests.c b/src/tests/refcount-tests.c
index 7fb398e7a..6d569994e 100644
--- a/src/tests/refcount-tests.c
+++ b/src/tests/refcount-tests.c
@@ -228,7 +228,6 @@ int main(int argc, const char *argv[])
suite = create_suite();
sr = srunner_create(suite);
- srunner_set_fork_status(sr, CK_FORK);
/* If CK_VERBOSITY is set, use that, otherwise it defaults to CK_NORMAL */
srunner_run_all(sr, CK_ENV);
failure_count = srunner_ntests_failed(sr);