Re: [xen-4.6-testing test] 62015: regressions - FAIL

[Ian Campbell <ian.campbell@citrix.com>]

On Thu, 2015-09-17 at 13:40 +0100, Wei Liu wrote:
> On Thu, Sep 17, 2015 at 06:44:05AM +0000, osstest service owner wrote:
> > flight 62015 xen-4.6-testing real [real]
> > http://logs.test-lab.xenproject.org/osstest/logs/62015/
> > 
> > Regressions :-(
> > 
> > Tests which did not succeed and are blocking,
> > including tests which could not be run:
> >  test-amd64-i386-xl-qemut-stubdom-debianhvm-amd64-xsm 9 debian-hvm
> > -install fail REGR. vs. 61745
> > 
> 
> I think this test case is unreliable.  For example, it passed in 61745
> (xen-4.6-staging) but failed in 62012 (linux-3.14) while both had the
> same version of Xen.

This has failed twice in a row on xen-4.6-testing, in 61839 and here. The
bisector already had a go at the initial failure, but concluded
unreproducible as shown by the yellow box in

http://logs.test-lab.xenproject.org/osstest/results/bisect/xen-4.6-testing/test-amd64-i386-xl-qemut-stubdom-debianhvm-amd64-xsm.debian-hvm-install.html

> The history of this test case shows more failures than successes. It's
> failing in all branches.

Looking at each branch individually under http://logs.test-lab.xenproject.o
rg/osstest/results/history/test-amd64-i386-xl-qemut-stubdom-debianhvm-amd64
-xsm/ it looks to me like it might be specific to a class of machines or
something like that.

> I think a force push is justified.

I think I agree.

Ian.