Bugzilla account

["Marcin M." <marcin.marcin.m@gmail.com>]

Hi!

I'd like to report a bug against GCC. Can you please create an account
for me?

I attach the bug report in case I should:

<marmistrz> Hi! I have a feeling I've just found a bug in GCC 6.3.0
OpenMP implementation (new OpenMP 4.5 features)
<marmistrz> I'm running this code: https://gitlab.com/marmistrz/pw-brandes
<marmistrz> The parallelization is done by the following OpenMP directive:
<marmistrz> #pragma omp parallel for reduction(+: results[:n])
<marmistrz> Where results is obtained as here:
https://gitlab.com/marmistrz/pw-brandes/blob/master/brandes.cpp#L21
<marmistrz> I get a segmentation fault when using the following dataset:
http://snap.stanford.edu/data/com-LiveJournal.html
<marmistrz> (probably you need to remove the first 4 comment lines from
it to work)
<marmistrz> or rather this dataset (both segfaults, but I tested more
the second): http://snap.stanford.edu/data/web-Stanford.html
<marmistrz> The segmentation fault appears only on an Intel Xeon E5-2697
v3 (Haswell) processor, with 2 or more threads
<marmistrz> I have managed to successfully complete the web-Stanford
testcase on an Intel i5-5200U processor, with 4 threads.
<marmistrz> The backtrace points to gomp_thread_start:
<marmistrz> #0  0x0000000000407006 in brandes () at
/home/marmistrz/brandes/brandes.cpp:21
<marmistrz> #1  0x00002aaaab31fc56 in gomp_thread_start () at
../../../gcc-6.3.0/libgomp/team.c:119
<marmistrz> #2  0x00002aaaab753aa1 in start_thread () from
/lib64/libpthread.so.0
<marmistrz> #3  0x00002aaaaba51bcd in clone () from /lib64/libc.so.6
* linuxmodder_ is now known as linuxmodder
<marmistrz> It appears that the segfault at the very beginning of the
thread start.

Regards