Bug#693895: #693895: eglibc: FTBFS on powerpcspe

[Roland Stigge <stigge@antcom.de>]

Hi Aurelien,

you wrote:
>> Encountered regressions that don't match expected failures:
>> tst-cancel4.out, Error 1
>> tst-cancel5.out, Error 1
>> tst-cancelx4.out, Error 1
>> tst-cancelx5.out, Error 1
>> tst-key1.out, Error 1
>> tst-key4.out, Error 1
>> make: *** [/«PKGBUILDDIR»/stamp-dir/check_libc] Error 1
>> dpkg-buildpackage: error: debian/rules build-arch gave error exit
status 2
>> [...]
>>
>> Attaching a patch that fixes this.
>>
>
> More than fixes the issues, it just ignores them. Have you investigated
>why these tests are failing? Ignoring the tst-key1 and tst-key4 tests
>seems a bit worrying to me.

Right!

Upon tracing this back, I noticed that when building this package (with
sbuild) on a btrfs snapshot, the above issues surface, while they are
gone when manually building on an ext4.

I guess this is related to the maximum number of threads being tested in
tst-key1. Maybe btrfs somehow reduces the number of available threads
somehow?

How about reassigning this to some btrfs package?

Using Linux 3.6.0.

Thanks,

Roland