18:01:57 #startmeeting Fedora Kernel Meeting 18:01:57 Meeting started Fri Aug 3 18:01:57 2012 UTC. The chair is jforbes. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:57 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:02:10 #chair jwb davej 18:02:10 Current chairs: davej jforbes jwb 18:02:37 Hello folks. Let's start with who's here 18:03:11 #meetingname Fedora Kernel meeting 18:03:11 The meeting name has been set to 'fedora_kernel_meeting' 18:03:41 i'm here. davej is 1/2 here 18:03:55 Okay, guess it is a pretty small list at this point. Let's start with the regular run down of the releases 18:03:57 * nirik is lurking around 18:04:16 Then we will get into the Virtual FAD we have planned for the kernel regression tests suite 18:04:27 #topic Rawhide 18:04:36 jwb: care to take this one? 18:04:45 sure! 18:05:21 so Justin started the 3.6 merge window kernel rebases. we've rolled through that and tomorrow's rawhide should contain 3.6-rc1 (with debug options off) 18:05:44 aside from just the normal rebase stuff, we've also included the first set of secure boot patches 18:06:10 for those really wanting to be daring, there is a kernel command line parameter that will allow you to essentially put your box in SB mode 18:06:23 without having to actually have UEFI, or keys, etc 18:06:53 secureboot_enable=1 will cause the kernel to fake out the permissions bits 18:07:05 i wouldn't recommend setting it yet though. X likely will not start 18:07:35 Anything else on rawhide? 18:07:39 I thought ajax already built X with that fix? 18:07:51 pjones, ajax has patches. they were submitted upstream 18:08:36 ah, okay 18:08:36 ok, ajax said he thinks they're in rawhide 18:08:42 I mean, he already gave me a package 18:09:01 yeah, i have one too. so if someone wants to play and X doesn't start, let us know 18:09:27 Excellent, anything else on rawhide? 18:09:27 jforbes, i don't have anything else on rawhide 18:09:35 #topic F17 18:09:43 ok, ajax said the xorg patches aren't in rawhide 18:09:51 14:09 <@ajax> ugh, no, not done in rawhide. thought i had, have the patches in my git checkout, but clearly never checked it in 18:09:54 anyway 18:09:55 F17 :) 18:10:06 davej: I know you are on a crippled connection, so I will summarize, and you can add anything 18:10:23 thx 18:10:55 The 3.5 update has made it through updates-testing and pushed to stable. Most of the major patches already there for obvious 3.5 problems 18:11:35 There is still the power issue with i915 that doesn't have a fix yet, though it is not exactly a major issue either. Upstream is looking at it, and I think there will be a fix soon 18:12:15 SDGathman: heh, you had the same idea 18:12:17 I would guess 3.5.1 will be tagged soon now that the merge window is closed, and we will get that into F17 as soon as possible 18:12:37 davej: anything to add? 18:13:53 no. thats about it 18:13:53 anyone else on F17? 18:14:59 Okay... moving on 18:14:59 #topic F16 18:14:59 yes, F17 18:14:59 cpuobsessed: okay, go ahead 18:15:00 i have a laptop with X2300 Radeon, I updated to the 3.5.0-2 kernel and now the video doesn't work; i filed a bug 18:15:17 https://bugzilla.redhat.com/show_bug.cgi?id=845631 18:15:17 F17 here also. Dell D610 laptop won't boot with 3.5.0-2 18:15:32 https://bugzilla.redhat.com/show_bug.cgi?id=845388 18:16:01 mine would freeze during boot: fb: conflicting fb hw usage radeondrmfb vs VESA VGA - removing generic driver 18:16:21 is there a possible workaround? or another kernel update? 18:16:45 if nobody has commented in the bug yet, it means we haven't had time to look into it or haven't found anything yet 18:17:11 Okay, so we have bugs filled on both. It looks like a couple of people have reported the same thing. There is no update to address it right now, though I think I have seem some upstream discussion on the issue 18:17:46 if someone is capable of doing a git bisect between 3.4 and 3.5 to find out which commit caused it, that would be very helpful 18:18:00 I will see what I can find on it, though I have not seen that behavior here 18:18:28 i think i pointed airlied at one of the initial bugs. he hadn't seen it either 18:18:56 But at this time, I don't know a quick workaround, and we don't have an update to address it, so the work around is to keep using 3.4.6 until a fix is available 18:19:16 thanks, i may have another comment to add to my bug 18:19:48 And please respond if someone requests more information on your bug. It would be nice to get this fixed, but without being able to reproduce, we need your feedback 18:20:03 Anything else on F17? 18:21:00 Okay, moving on to F16 18:21:45 Right now F16 has 3.4.7 sitting in updates-testing. It seems we are falling into the same pattern that F15 was in before. We have no one really testing and providing feedback 18:30:28 Once 3.4.7 pushes, I will start pushing 3.5 for F16 18:30:28 There are a number of bugs filed against F16 which we know are fixed with 3.5, so hopefully that will reduce the bug count a bit more 18:30:28 Anyone else have anything on F16? 18:30:28 Okay, moving on to the Virtual FAD 18:30:29 #topic Regression Test Suite Virtual FAD 18:30:29 A message went out to the lists today as well, but we will be hosting a Virtual Fedora Activity Day on August 14th to try and get things moving faster on the kernel regression test suite 18:30:29 more details are available at https://fedoraproject.org/wiki/KernelRegressionTesting_Virtual_FAD_20120814 18:30:29 #link https://fedoraproject.org/wiki/KernelRegressionTesting_Virtual_FAD_20120814 18:30:29 What we are looking for most is people to write new tests, flesh out existing tests, or make improvements to the overall test harness. 18:30:29 The code is available on git://git.fedorahosted.org/kernel-tests.git and there is a good deal of information on the wiki with what is required, what is in progress, and what still needs work 18:30:46 Anyone have any questions, comments, ideas on how we can make this a successful event? 18:30:46 If not, we will move on, though we would love to see you participate in the actual event 18:30:46 #topic rawhide linux-next 18:30:57 We have had some discussions regarding moving rawhide to the linux-next tree 18:30:58 jwb: want to take this one? 18:52:18 nothing earth shattering, but definitely stuff that should have been caught by the linux-next process 18:52:18 * nirik welcomes zodbot back 18:52:18 except hardly anybody runs linux-next 18:52:32 that leads to some trouble getting last minute fixes in upstream, etc 18:52:55 2) rawhide kernel usage is pretty small. aside from the kernel maintainers, we get very few reports on things in rawhide 18:53:24 I think the debugging scares a lot of people away 18:53:25 now, 2 is kinda unrelated to 1 but i was thinking since the audience is small and more technically oriented, we might try and build linux-next in rawhide 18:53:32 nirik, yeah 18:53:55 running linux-next wouldn't really remove that factor, but maybe we'd catch more issues sooner 18:54:22 there are some windows we'd want to watch out for if we did this 18:54:28 basically around release branch time 18:55:01 it isn't set in stone at all, but i wanted to bring the possibility up 18:55:25 do people think it's crazy? is it something we should do as a side effort for a bit first, etc etc 18:55:41 I like the idea, though like you said, we have times when we couldn't do it. Like between now and the branch would be a bad idea 18:56:39 right 18:56:44 how would we handle versioning ? 18:57:13 pre-release style i guess 18:57:44 I'm just curious because if we're doing linux-next pre branching, and then we flip back to following linus' tree, we don't want version #'s to go backwards 18:58:06 3.6.0-rc0.0.git0.1 would be linux-next during the 3.5 stuff. then if switch to actual 3.6 merge window kernels, we'd use 3.6.0-rc0.1.git0.1 18:58:24 i have not thought about that in detail since i was kinda assuming people would tell me i'm crazy 18:59:00 I don't think it is crazy at all, just might require some more thought 18:59:26 maybe start a wiki page to flesh out all the different cases we'd have to handle 18:59:38 yeah, sounds like a good idea 18:59:45 That's probably a good idea, and we can take it to list for further feedback 18:59:55 Does anyone have anything else before the clouds roll in? 19:00:32 Okay, going to endmeeting and let them have it then 19:00:37 #endmeeting