18:00:29 #startmeeting 18:00:29 Meeting started Fri May 25 18:00:29 2012 UTC. The chair is jwb. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:29 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:00:40 #meetingname fedora-kernel 18:00:40 The meeting name has been set to 'fedora-kernel' 18:00:51 #meetingtopic Fedora Kernel 18:01:02 #topic Release overview 18:01:13 woop woop 18:01:20 fun 18:01:34 start with f15? 18:01:40 yeah 18:01:47 #Release overview - F15 18:01:53 #topic Release overview - F15 18:01:59 a month until eol, and it might as well be dead already tbh. 18:02:11 yeah 18:02:15 we've had 20 bugs opened against it in the last month (compared to the 191 for f16) 18:02:22 i believe we've decided to stick with 3.3.x 18:02:23 3.3.7 is sitting in updates-testing 18:02:56 the last couple of weeks of its life will be just security fixes, so hopefully not too many updates next month. 18:03:13 i plan on doing whatever gregkh releases for 3.3.y, and that's about it 18:03:19 307 f15 bugs still open. I suspect a bunch of those will become 16 (or newer) bugs next month 18:03:48 it's been stuck around the 300 mark for a few months now, we've mostly closed as many as come in for that release. 18:04:24 anything else on this one? 18:04:29 I don't think it's worth mass-migrating them all to f16 bugs though, rather let the users do it if necessary. 18:04:41 I see it as an opportunity to clean out some deadwood bugs 18:04:43 yeah, good point 18:05:01 that's all I have for 15 18:05:24 #info F15 to stay on 3.3.y until EOL 18:05:45 #info bugs will be looked over and triaged one final time 18:05:50 #topic Release overview - F16 18:06:12 who was doing f16? 18:06:12 16 is a different story. still lots of bug activity. 18:06:21 so F16 (and F17) also have 3.3.7 in updates testing at the moment 18:07:13 At the same time we have 3.4 tracking and ready to commit once 3.3.7 has made it's way out, and once the initial mass of stable queue has been pushed 18:07:32 I expect the 3.4 push to happen next week 18:07:33 #topic Release overview - F16/F17 18:07:34 524 open bugs. 191 opened this month, 217 closed! lots of old bugs got closed out. (fallout from the i915 corruptor for eg) 18:07:52 #info F16/F17 moving to 3.4 once 3.3.7 moves to stable updates 18:08:15 And we do have a small number of bugs which I know are addressed in 3.4 and we can close out once it ships. not enough mass to make it worth backporting when we knew release was this close 18:08:56 * nirik notes from the peanut gallery that 3.4 here made iwlwifi unuseable again... expect bugs. ;) 18:09:23 #info iwlwifi seems to have more issues. again. for like the 3rd kernel release 18:09:28 might be worth asking linville about that before we push a 3.4 update, to see if there's anything we need to backport 18:09:34 (or revert) 18:10:25 I will, nirik do you have any details anywhere? 18:10:38 perhaps someone could get the developers some hardware to test with. ;) 18:10:45 #action jforbes to discuss 3.4 iwlwifi issues with linville before rebase 18:10:50 jforbes: no, but if you want I can retry it and gather info... 18:11:00 it would just stop working after a few minutes. 18:11:10 nirik: I would appreciate it 18:11:12 reloading the module made it work again for a few minutes. 18:11:20 nirik, 802.11{a,b,g,n}? 18:12:04 I wonder if it's yet another case of intel moving to the latest shiny chipsets and breaking older ones they no longer test. 18:12:11 i think it's somewhat known that N has issues. i have iwlwifi here and it's worked fine on a G only network 18:12:20 n 18:12:47 but it works fine with 3.3.7 here (what I am on now) 18:13:02 could koji bisect it ;) 18:13:29 so a regression specifically in n. I put it on my wife's laptop last night, but we have a G only network here as well 18:14:06 I have N here, not noticed anything unusual, but I've not been using it as much lately 18:14:22 I can try and gather more info on it. 18:14:32 nirik: Thanks 18:14:53 anything else on F16/F17? i kinda lumped them together 18:14:59 F17 releases tuesday. woot 18:15:20 I suspect we'll get a bunch of shiny new bugs when that gets into peoples grubby hands 18:15:34 #info F17 Gold. Yay. 0-day kernel updates will be pending for all 18:15:47 81 open f17 bugs right now. quite a few I think are fixed in 3.4 18:16:32 Yup, 3.4 will let us close a good number of bugs, let's hope it does't open just as many. It has been stable here for a while though 18:17:08 #info 3.4 rebase (iwlwifi issues aside) should close a number of f17 bugs 18:17:21 anything else? 18:17:35 rawhide ? 18:17:41 #topic Release overview - rawhide 18:17:44 so 18:17:52 the 3.4 final release in rawhide was a non-event 18:17:55 rawhide feels kinda virtually untested these days. 18:17:59 exactly 18:18:10 i've rebased it to 3.5 merge window kernels already 18:18:11 we had just 14 bugs opened this month against rawhide 18:18:15 lower than f15! 18:18:21 so far, the tty lockdep issues are the biggest item found 18:18:40 hit another issue in udev where it would make USB devices act funny. that needs a udev patch 18:18:52 mjg59 provided that already this morning 18:19:04 aside from those two things, it's been fairly boring on my hardware 18:19:21 I think things have been this way since fedora started doing branched releases. I suspect once 17 is out, we'll get a bunch of rawhide users, up until we branch again. 18:19:38 here's hoping 18:20:04 oh 18:20:08 Well, it's good timeing, we should get more rawhide users as 3.5 gets closer to release 18:20:12 uprobes is merged and enabled now 18:20:21 #info rawhide rebased to 3.5 merge window 18:20:26 #info uprobes merged and enabled 18:20:28 I am using rawhide on my home desktop with the nodebug builds and things have been working pretty smoothly. 18:21:02 I fall a little behind because I only use one (the nodebug) build per RC. 18:21:37 brunowolff, cool. good to know :) 18:21:48 ok, anything else on rawhide? 18:22:13 #topic Open Floor 18:22:40 I s 3.6 likely to make it into F18 at release? 18:22:40 just a note that the regression test framework is on fedorahosted http://git.fedorahosted.org/git/?p=kernel-tests.git 18:23:05 #info kernel-test regression test framework on fedorahosted now 18:23:11 brunowolff: probably 18:23:15 #link http://git.fedorahosted.org/git/?p=kernel-tests.git 18:23:22 This is just the framework, I am cleaning up some tests which will be pushed probably over the next couple of days. 18:23:45 Also, the docs have been updated on the wiki to reflect what is in git. https://fedoraproject.org/wiki/KernelRegressionTestGuidelines 18:23:59 cool 18:24:31 I will get a more detailed email out to the kernel list shortly, and once we get a good number tests in the tree I will get emails out to devel-list and see if we can get people actively running it 18:24:57 Also, patches welcome, encouraged, appreciated... kernel list would be the best place to send them 18:26:26 nice 18:26:40 anyone have an other topics or questions? 18:27:31 jforbes: probably want to add an empty logs/ dir in the topdir (or mkdir it in the script if it doesn't exist) 18:29:01 davej: Doh. I have one, but I guess git didn't add the empty dir. Will fix 18:29:29 i think we're starting to lag 18:29:37 meeting seems done :) 18:29:45 yeah, ending early wfm 18:29:56 ok, 30 seconds... 18:30:11 #endmeeting