|
From: Eric C. <Eri...@gi...> - 2016-10-19 13:48:22
Attachments:
valgrind_out.carres_jumeaux_04.n0.txt
|
Hi, I recently upgraded to valgrind 3.11.0 (was 3.8.1 before). I also upgraded Intel MKL to the 2015 version. We use valgrind since 2011-10-27 to test our code nightly with about 2200 tests. Since this upgrade, some tests (not always the same) have the same trouble (see full log in attachment): ================================ ... blockSane: fail -- redzone-hi valgrind: m_mallocfree.c:2042 (vgPlain_arena_free): Assertion 'blockSane(a, b)' failed. host stacktrace: ... ================================ I would like some help to debug this. It is quite annoying since for the same test, the problem appear or disappear from one day over another... Today we got the problem on 6 tests (on 2200), yesterday, only 1 test failed, before yesterday, none failed.... Thanks for any insights or clues! Eric |
|
From: Mark W. <mj...@re...> - 2016-10-19 14:33:52
|
On Wed, 2016-10-19 at 09:29 -0400, Eric Chamberland wrote: > I recently upgraded to valgrind 3.11.0 (was 3.8.1 before). > I also upgraded Intel MKL to the 2015 version. > > We use valgrind since 2011-10-27 to test our code nightly with about > 2200 tests. > > Since this upgrade, some tests (not always the same) have the same > trouble (see full log in attachment): > > ================================ > ... > blockSane: fail -- redzone-hi > > valgrind: m_mallocfree.c:2042 (vgPlain_arena_free): Assertion > 'blockSane(a, b)' failed. > > host stacktrace: > ... > > ================================ > > I would like some help to debug this. > > It is quite annoying since for the same test, the problem appear or > disappear from one day over another... > > Today we got the problem on 6 tests (on 2200), yesterday, only 1 test > failed, before yesterday, none failed.... > > Thanks for any insights or clues! I believe this is https://bugs.kde.org/show_bug.cgi?id=356457 Sadly it is not well understood. If you could add any information to that bug that could help us understand better when this issue triggers that would be very helpful. There are some hints in that bug report how to get more information if possible. Thanks, Mark |
|
From: Julian S. <js...@ac...> - 2016-10-19 15:01:12
|
Eric, >> Thanks for any insights or clues! > > I believe this is https://bugs.kde.org/show_bug.cgi?id=356457 > Sadly it is not well understood. If you could add any information to > that bug that could help us understand better when this issue triggers > that would be very helpful. There are some hints in that bug report how > to get more information if possible. To second Mark's comments -- yes -- please do add your observations (all the information in your original email) to https://bugs.kde.org/show_bug.cgi?id=356457 This is a difficult bug which we've been aware of for some time but have no way to reproduce, hence no way to fix. If you add your comments to 356457, then it may be that, if enough people do the same, we can see some pattern or similarities between the crashes, and so make some progress. That would be very helpful. J |