test suite backslides again...

Zack Weinberg zackw@Stanford.EDU
Mon Apr 2 14:29:00 GMT 2001


On Mon, Apr 02, 2001 at 01:38:16PM -0700, Zack Weinberg wrote:
> 
>                 === libjava Summary ===
> # of expected passes            840
> # of unexpected failures        370
> # of unexpected successes       14
> # of expected failures          500

Further note on the libjava failures.

PASS: Shazam -O compilation from source
The next line should say "shazam"
shazam
FAIL: Shazam -O execution from source compiled test

$ ./Shazam 
The next line should say "shazam"
shazam
Aborted

$ strace -f  .libs/lt-Shazam
... a great deal of output ...
[pid  2282] write(1, "The next line should say \"shazam"..., 34) = 34
[pid  2282] write(1, "shazam\n", 7)     = 7
[pid  2282] kill(2280, SIGRT_0)         = 0
[pid  2280] --- SIGRT_0 (Real-time signal 0) ---
[pid  2280] <... rt_sigsuspend resumed> ) = -1 EINTR (Interrupted system call)
[pid  2280] sigreturn()                 = ? (mask now [])
[pid  2280] rt_sigprocmask(SIG_SETMASK, NULL, [32], 8) = 0
[pid  2280] rt_sigsuspend([] <unfinished ...>
[pid  2282] kill(2280, SIGRT_0)         = 0
[pid  2280] --- SIGRT_0 (Real-time signal 0) ---
[pid  2280] <... rt_sigsuspend resumed> ) = -1 EINTR (Interrupted system call)
[pid  2280] sigreturn()                 = ? (mask now [])
[pid  2280] --- SIGSEGV (Segmentation fault) ---
[pid  2280] rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0
[pid  2280] getpid()                    = 2280
[pid  2280] kill(2280, SIGABRT)         = 0
[pid  2280] --- SIGABRT (Aborted) ---
[pid  2280] +++ killed by SIGABRT +++

Seems like some sort of crash on exit...  It doesn't happen when
running under the debugger, bleah.

zw



More information about the Gcc mailing list