Re: [xen-unstable test] 53884: regressions - FAIL

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

On Mon, 2015-05-11 at 09:12 +0100, Andrew Cooper wrote:
> On 11/05/2015 09:08, Jan Beulich wrote:
> >>>> On 09.05.15 at 20:57, <osstest@xenbits.xen.org> wrote:
> >> flight 53884 xen-unstable real [real]
> >> http://logs.test-lab.xenproject.org/osstest/logs/53884/ 
> >>
> >> Regressions :-(
> >>
> >> Tests which did not succeed and are blocking,
> >> including tests which could not be run:
> >>  build-amd64-xsm               5 xen-build                 fail REGR. vs. 50405
> > This is due to e5786bae04 ("x86/pvh: use a custom IO bitmap for
> > PVH hardware domains"). Having requested th addition during
> > review I of course took it for granted that you build tested your
> > change. As dealing with this imo also means moving the code
> > addition into the already existing (or a new) CONFIG_X86 block
> > (along with moving the declaration to a more suitable header), I
> > think a fixup patch would not be the right way to deal with this,
> > and hence I think I'm going to revert that change for the time
> > being.
> 
> Given this reversion (and confirmation that the build is fixed), would a
> force push be justified?  We are now a full month behind, with other 
> OSSTest issues still outstanding.

Force pushing would miss out testing the functional aspects of whatever
tests would have run on the other commits in the failed build had it
succeeded.

But given that this is an -xsm build we perhaps don't care so much about
that yet (few of the tests pass yet anyway).

Flight 54309 tested a57b1fff48a8 which is the current staging, and it
failed with the build-amd64-xsm and the freebsd migration thing (which
IMHO we should just whitelist for the time being).

For there to be any possibility of a force push the revert would have to
be the first thing added on top of a57b1fff48a8.

Whatever we do I would suggest killing flight 55228 since it is just
going to fail this build too.

Ian.