[compiler-rt] r302786 - [MSAN] test failed randomly on ARM when XFAILED for MIPS

[Simon Dardis via llvm-commits]

> -----Original Message-----
> From: llvm-commits [mailto:llvm-commits-bounces at lists.llvm.org] On Behalf
> Of Renato Golin via llvm-commits
> Sent: 11 May 2017 14:42
> To: Alexander Potapenko
> Cc: LLVM Commits
> Subject: Re: [compiler-rt] r302786 - [MSAN] test failed randomly on ARM
> when XFAILED for MIPS
> 
> On 11 May 2017 at 14:35, Alexander Potapenko <glider at google.com> wrote:
> > Any idea what's wrong with that test? It's failing on PPC64 now:
> > http://lab.llvm.org:8011/builders/clang-ppc64le-linux-multistage/build
> > s/2278/steps/ninja%20check%201/logs/FAIL%3A%20MemorySanitizer-
> powerpc6
> > 4le%3A%3A%20wcsncpy.cc
> 
> I have no idea.
> 
> Disabling on MIPS made it fail on ARM. Disabling on ARM made it fail on
> PowerPC. I say this test is busted.
> 
> This looks like one of those tests that are more dependent on the
> environment (libc version, etc) than the architecture, so I'm not hopeful it
> will every be stable on the number of different bots we have.

I think it might be environment / runtime sensitive. I've been trying to clear out
some test failures for MIPS64 and I'm seeing some odd behaviour in other tests.

Thanks,
Simon

> 
> Evgeniy, any idea why it fails randomly on all arches? Is this an x86-only kind
> of test? Or requires a specific version of glibc?
> 
> cheers,
> --renato
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits