|
From: Florian K. <br...@ac...> - 2011-10-27 16:21:30
|
This testcase is failing on x86-linux due to an additional reported problem:
+Thread #x's call to pthread_cond_wait failed
+ with error code 22 (EINVAL: Invalid argument)
+ at 0x........: pthread_cond_wait_WRK (hg_intercepts.c:...)
+ by 0x........: pthread_cond_wait@* (hg_intercepts.c:...)
+ by 0x........: main (tc23_bogus_condwait.c:72)
+
That seems reasonable to me given that on line 69 we do
r= pthread_cond_wait(&cv, (pthread_mutex_t*)(4 + (char*)&mx[0]) );
which may modify cv in some bogus way. The same cv is used on line 72
causing the EINVAL. But I'm no expert on pthreads. Not sure whether my
analysis is correct.
Assuming the above is valid, can the testcase be rewritten to avoid the
additional complaint? Perhaps by using a separate condition variable for
the bogus mutex test on line 69?
I would prefer to not have an additional exp file if it can be avoided.
Florian
|