On Fri, Mar 02, 2012 at 09:49:10AM +0200, Itamar Heim wrote:
> On 03/01/2012 08:18 PM, Ewoud Kohl van Wijngaarden wrote:
> >On Thu, Mar 01, 2012 at 03:18:59PM +0000, Itzik Brown wrote:
> >>I have a problem building vdsm on RHEL 6.2
> >>After running the following:
> >># git clone http://gerrit.ovirt.org/p/vdsm.git
> >># cd vdsm
> >># git fetch http://gerrit.ovirt.org/p/vdsm refs/changes/93/1693/5&& git
> >>checkout FETCH_HEAD
> >># ./autobuild.sh
> cc-ing eyal - lets make sure we have a jenkins job running
> ./autobuild.sh on master (if not already)?
Just make sure *not* to automatically run any code posted to gerrit.
You can run pyflakes *on* it, but blind 'make' before human review invites evil
> >Already worked this out in #vdsm, but for the archives.
> >Using git bisect you can find the commit that broke it. So we know HEAD
> >is bad (because ./autobuild.sh fails) and we assume master is good, thus
> >we start bisecting:
> > git bisect start HEAD master
> >Since we already have a script that can tell us if it's good or bad we
> >can run:
> > git bisect run ./autobuild.sh
> >This will produce output and if you're lucky produce a line similar to:
> > 57dbd4f19a953f07860d74d3c08d450e3e76701b is the first bad commit
> >Next step is to look in gerrit. This leads us to a review
> >(http://gerrit.ovirt.org/#change,1038) which is still in progress. In
> >this case you can make a review to make sure it doesn't get pushed in
> >this state. If it was merged already you've found a regression. Note
> >that we assumed master was good, but that may be wrong so double check.
> >vdsm-devel mailing list
> vdsm-devel mailing list
vdsm-devel mailing list