Bug#673596: libc6: FTBFS on wheezy/sid amd64 (test suite failures)
On Tue, May 22, 2012 at 02:01:44AM +0100, Philip Ashmore wrote:
> On 20/05/12 20:22, Aurelien Jarno wrote:
> >tag 673596 + unreproducible
> >thanks
> >
> >On Sun, May 20, 2012 at 04:01:10AM +0100, Philip Ashmore wrote:
> >>On 20/05/12 03:02, Jonathan Nieder wrote:
> >>>Hi Philip,
> >>>
> >>>Philip Ashmore wrote:
> >>>
> >>>>annexc.out, Error 1 (ignored)
> >>>>tst-eintr1.out, Error 1
> >>>>tst-writev.out, Error 1
> >>>>***************
> >>>>Encountered regressions that don't match expected failures:
> >>>>tst-eintr1.out, Error 1
> >>>
> >>>What does the tst-eintr1.out file say?
> >>>
> >>>Curious,
> >>>Jonathan
> >>>
> >>.tf1: pthread_create failed: Resource temporarily unavailable
> >>tf1: pthread_create failed: Resource temporarily unavailable
> >>
> >>That dot before "tf1" is repeated 3734 times.
> >>
> >
> >I am unable to reproduce this bug on my machines, and it really looks
> >like an environment specific bug. I am therefore not planning to work on
> >this bug. If someone is able to reproduce it, please work on it.
> Hi there.
> I created VMware virtual machines based on the following ISOs.
> debian-6.0.5-amd64-netinst.iso
> debian-wheezy-DI-a1-amd64-netinst.iso
>
> I wanted to use qemu, but I couldn't get kvm to work so it was too slow.
>
> I ran through the eglibc build on both of them.
> The 6.0.5 installation built eglibc-2.11.3 successfully.
Which is kind of expected, given we don't check for testsuite results on
stable version of the libc. This is for example to avoid breaking the
stable and security updates in case an incompatible new version of the
kernel in installed two years after.
> The wheezy installation failed to build eglibc-2.13-32 with the following errors<<EOF
>
> make[2]: Target `check' not remade because of errors.
> make[2]: Leaving directory `/home/contact/eglibc-2.13'
> make[1]: *** [check] Error 2
> make[1]: Leaving directory `/home/contact/eglibc-2.13/build-tree/amd64-libc'
> #
> # Testsuite failures, someone should be working towards
> # fixing these! They are listed here for the purpose of
> # regression testing during builds.
> # Format:<Failed test>, Error<Make error code> [(ignored)]
> #
> annexc.out, Error 1 (ignored)
> tst-cpuclock2.out, Error 1
> tst-eintr1.out, Error 1
> tst-writev.out, Error 1
> ***************
> Encountered regressions that don't match expected failures:
> tst-eintr1.out, Error 1
> make: *** [/home/contact/eglibc-2.13/stamp-dir/check_libc] Error 1
> dpkg-buildpackage: error: debian/rules build gave error exit status 2
> debuild: fatal error at line 1350:
> dpkg-buildpackage -rfakeroot -D -us -uc -b -j4 failed
> EOF
>
> I suspended the VM part way through which may account for the cpuclock2 error.
Anyway it's not the cause of the failure here, we have disabled it
because the test fails often on machines with variable CPU frequency.
> Can you tell me what your kernel version is, hardware etc?
I have not been able to reproduce the issue on:
- A Core i5 M560, running a 3.2.0-2-amd64 kernel
- A core i5 2500, running a 3.2.0-2-amd64 kernel
- An 4 x AMD Opteron 6134, running a 2.6.32-5-amd64 kernel
- A KVM instance on a Core i7 2600k, running a 2.6.32-5-amd64 kernel
> Do you have KDE4 installed?
On some of the machines yes. Have you experienced it makes a difference?
I really doubt.
> For squeeze, the default root fs is ext3, for wheezy it's ext4.
I really doubt this makes a difference.
The problem is clearly that at some point the system is not able to
create new threads anymore due to resources limits. In your case it's
likely due to the fact you are machine is a virtual VMWare machine.
Have you tried to run the build on a real hardware?
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
Reply to: