fedora-meeting
LOGS
19:00:20 <jforbes> #startmeeting Fedora-kernel
19:00:20 <zodbot> Meeting started Fri Nov 15 19:00:20 2013 UTC.  The chair is jforbes. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:20 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
19:00:33 <jforbes> #topic Who's here?
19:00:42 <jforbes> Anyone around for the kernel meeting?
19:00:46 * nirik is hanging around
19:01:28 <jforbes> Okay, this should be fairly quick
19:01:36 <jforbes> #Topic releases
19:02:04 <davej> yo
19:02:11 <jforbes> The 3.11.8 updates are pusing to F20 now, and pending updates testing in F18 and F19
19:02:36 <jforbes> They will essentially be following 3.11.x stable until the F20 release where we can push 3.12
19:03:32 <jforbes> For people wishing to run 3.12 now, I highly recommend the Rawhide no-debug repository. At this point it is not following rawhide, but staying on 3.12 stable releases until the F20 release
19:03:40 <jforbes> We will give notice before it changes back to follow rawhide
19:03:56 * nirik nods.
19:04:43 <jforbes> #info Users wishing to run 3.12 on stable releases should use the rawhide-nodebug repository until Fedora 20 ships
19:04:54 <jforbes> #link http://jforbes.livejournal.com/14128.html
19:05:22 <jforbes> Nothing else major from the release point of view at this time. Do you have anything to add davej ?
19:05:29 <davej> not really
19:05:37 <jforbes> #topic rawhide
19:06:03 <jforbes> Rawhide is in the merge window now, and being built fairly frequently.
19:06:31 <jforbes> Nothing really out of the ordinary in that regard, and there is still some time left in the merge window so we don't know everything that will land yet
19:06:42 <jforbes> Anyone have anything else on rawhide?
19:06:46 <nirik> seems to be working ok here.
19:06:50 <nirik> so far
19:07:01 <davej> merge window has been pretty stable for the most part
19:07:21 <jforbes> yeah, I haven't run into anything major
19:07:32 <davej> though we've still got another week to go unless Linus' backtracks on his earlier decision
19:07:50 <jforbes> #Topic Trinity
19:08:05 <jforbes> davej: want to give a brief trinity update? How's that BSD port going?
19:08:08 <davej> lol
19:08:32 <davej> so I've been fixing up some bugs in trinity trying to get another point release done soon.
19:09:03 <davej> at the beginnig of the merge window it wasn't finding many bugs, but that seems to have changed this last week
19:09:28 <davej> some of those bugs have been there for a while, and are being shaken out by me fixing trinity bugs
19:09:44 <davej> most of the stuff it's finding right now seems to be locking bugs
19:09:55 <davej> hopefully that'll calm down a bit post rc1
19:10:09 <davej> that's about it I think
19:10:13 <jforbes> thanks davej
19:10:25 <jforbes> #Topic Test/QA
19:10:51 <jforbes> Splitting my time right now on the testing bits.  I have two main priorities
19:11:49 <jforbes> The first has been getting the client to send results to a central website/database/etc. This is really required for badge integration, but has the added advantage of letting us see what tests are failing, what percentage, etc.
19:13:36 <jforbes> So it will be opt in, but the client for regression tests will send results to the central database, nothing is stored except for kver, and which tests failed. Additionaly you can provide FAS details and have it track for badges. If you choose this, only the fas username is stored with the kver/pass/fail
19:14:45 <jforbes> The second thing, which has taken a bit more precedence this week with beta shipping, is getting arm testing going.  While there are a lot of hardware aspects to ARM that just wont be tested, we should be testing basics in qemu
19:15:07 <davej> if you get an oops for eg, will that be on the website ?
19:15:35 <jforbes> davej: no, we aren't tracking that much.  Just specific test names that fail
19:15:45 <jforbes> davej: I suppose I could add that
19:17:04 <davej> alternatively, rely on something like kerneloops to do it for you, and just log the link.
19:17:34 <jforbes> davej: that might be a better idea. I was trying to keep the database small.
19:17:56 <jforbes> Anything else on test?
19:18:33 <jforbes> #topic open floor
19:18:40 <jforbes> Anything else on anything else?
19:19:18 <davej> coverity stuff has been going well
19:19:40 <davej> for the most part during the merge window, there have been more bugs being fixed/going away than new bugs being added
19:19:48 <davej> the last couple days, that trend has reversed slightly
19:20:04 <davej> but as we get to rc1, hopefully things will continue to improve
19:20:23 <jforbes> People care more after your kernelsummit mentions?
19:20:24 <davej> but this merge window is going way better than the previous window in that regard
19:20:37 <davej> yeah, quite a few maintainers signing up (as well as the usual slew of janitor types)
19:20:59 <davej> something that was suggested was setting up a mailing list for a first-pass review of patches before sending them on to maintainers
19:21:10 <davej> I might see about getting that set up soon
19:21:23 <jforbes> Cool
19:21:29 <davej> might alleviate the problems that some maintainers are getitng fed up with crap patches from coverity
19:22:03 <davej> still having conversations with the coverity guys about improving some of their checks too
19:22:20 <davej> because there are quite a few bugs that would just go away if they fixed up some of those checks
19:22:43 <davej> still hovering at just over 5000 outstanding reports
19:23:15 <davej> I'll look into making some pretty graphs at some point
19:23:29 <davej> think that's about all.
19:23:42 <jforbes> Excellent? Anything else from anyone?
19:25:09 <jforbes> Okay, closing out, thanks for coming
19:25:13 <jforbes> #endmeeting