On 3 January 2018 at 13:14, Dominik Holler <dholler@redhat.com> wrote:
On Wed, 3 Jan 2018 11:41:05 +0200 Yedidyah Bar David <didi@redhat.com> wrote:
On Wed, Jan 3, 2018 at 11:36 AM, Dominik Holler <dholler@redhat.com> wrote:
If a code change adds a warning to ovirt-engine build process, shouldn't CI mark the change as unstable?
Can you give an example (e.g. a link to a jenkins build with such a warning)?
I created the change https://gerrit.ovirt.org/#/c/85925/ to produce different kinds of warnings. The CI result page http://jenkins.ovirt.org/job/ovirt-engine_master_check-patch-el7-x86_64/3530... says: "Plug-in Result:Unstable - 6 warnings exceed the threshold of 0 by 6 (Reference build: #35300)" Looks like CI checks for added findbugs warnings, but not for javac warnings.
I do not know if we currently emit any warnings. If we do, and you want to ignore only the existing ones and fail new ones, I'd say this is a bit hard and non-maintainable.
The findbugs warnings seems to use something smart like this.
yes because Jenkins has built-in Findbugs capabilities, not sure there is something generic we can use for just any warning. Maybe by emitting XUNIT reports, but I thing those only support errors.
If you want to treat all warnings as fatal errors, I guess this can be done.
I expect there are many warnings, if they are currently not managed.
well, we're after a release so this maybe be the best time for turning them into errors and then doing a fix marathon. Then again stuff is probably accumulating for 4.2.1 quickly, so maybe not.... -- Barak Korren RHV DevOps team , RHCE, RHCi Red Hat EMEA redhat.com | TRIED. TESTED. TRUSTED. | redhat.com/trusted