[OE-core] Sanity Failures - Segfaults in qemu images

[Richard Purdie]

On Sun, 2013-04-07 at 15:32 -0700, Khem Raj wrote:
> On Sun, Apr 7, 2013 at 1:23 AM, Richard Purdie
> <richard.purdie at linuxfoundation.org> wrote:
>         http://autobuilder.yoctoproject.org:8011/builders/nightly-x86-64-lsb/builds/87/steps/Running%20Sanity%20Tests/logs/stdio
>
>
> what does complete dmesg output looks like ? I have seen ld.so
> segfaults on real x86_64 hardware but havent narrowed it down since
> it does not seem to bother my testing.​

I don't have the full dmesg output however I have put a wealth of info
about this into:

https://bugzilla.yoctoproject.org/show_bug.cgi?id=4216

Basically these do look like "random" segfaults happening any point on
the system and having a variety of effects. The double fault is
particularly worrying as it suggests the problem is kernel or qemu. I've
put a script which tends to reproduce the issue into the bugzilla.

When you say you've seen ld.so faults on real hardware, have you any
more details? Are they happening often? Just ld.so?

I need to know if this is a qemu issue or on real hardware too since the
latter is a release blocker and extremely serious.

Cheers,

Richard