Validation tests not yet run against Final

[Adam Williamson]

On Wed, 2013-06-26 at 07:38 -0400, Konrad Rzeszutek Wilk wrote:

> >HOWEVER, I saw an issue when trying to boot an already installed x86_64
> >PV guest. Basically, if I use a disk image with F19 already installed
> >and try to boot it by passing the kernel and the initrd directly,
> >here's
> >what I seem to get:
> >
> > - using {kernel,initramfs}-3.9.6-200.fc18 it boots fine.
> >
> > - using {kernel,initramfs}-3.9.6-301.fc19 it hangs.
> >
> >And, in the latter case, it hangs because he fails to detect the
> >virtual
> >hard disk (I tried both LVM volume and an image file), and hence to
> >mount the rootfs from there.
> >
> >That seems a bit weird, as it's the same kernel version, but I'm quite
> >able to reproduce this.
> >
> >I'm giving this one more try right now and I'll let you know, as, if
> >that is verified, I think it can be seen as a violation of the DomU
> >criterion.
> >
> >Konrad, have you seen anything like this? Since I saw you mostly tested
> >i386 configs, could it be some 32-vs-64 thing?
> >
> >Regards,
> >Dario
> 
> Hey Dario and Adam
> 
> I did not see the issue you mentioned. I had done 32 and 64bit to
> installs (but forgot to update the test matrix) using a 64 bit dom0.
> 
> My feeling is that it is a green go from DomU perspective.  Thought I
> am unsure what the issue Dariof is seeing (perhaps initrd did not
> bundle all modules?) The toolstack ones are not part of the release so
> we have some time to backport patches. 

Thanks again to both of you! Dario, you could add an entry to the matrix
if you liked; if I were you I'd file any bugs you've found that Konrad
has not yet filed, make your result a 'warn' and reference those bugs,
but in this situation (there are bugs related to the test but it's not
really 'failing') whether you use 'warn' or 'fail' is a pretty
subjective call.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net