18:00:25 #startmeeting 18:00:25 Meeting started Fri Oct 18 18:00:25 2013 UTC. The chair is jwb. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:25 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:00:25 #meetingname fedora-kernel 18:00:25 #meetingtopic Fedora Kernel 18:00:25 #topic init 18:00:25 The meeting name has been set to 'fedora-kernel' 18:00:38 #chair jforbes davej jwb 18:00:38 Current chairs: davej jforbes jwb 18:01:00 we'll do a quick meeting today. who all is around? 18:01:06 * nirik is lurking 18:01:18 yay! our token non-kernel team guy ;) 18:01:25 * nirik bows 18:01:37 * jforbes is here 18:01:43 * brunowolff is here 18:02:08 ok, let's get rolling 18:02:16 #topic f18/f19/f20 18:02:26 jforbes, wanna cover these? 18:03:07 Sure. F18 is in a weird state because there have been problems with a push, but the 3.11.4 rebase should be going out to stable as soon as a push happens, and 3.11.5 should be in updates-testing 18:03:29 fingers crossed the 18/19 push goes out in a bit. 18:04:00 F19 has 3.11.5 waiting for a push as well, hopefully it makes it before 3.11.6 18:04:14 F20 beta should also be on 3.11.5 18:04:25 #info f18 3.11.4 rebase coming soon 18:04:32 #info f19/f20 both on 3.11.5 18:04:39 there's one more keys patch i need to add for f20 18:04:52 fixes a memory/quota leak that prevents the krb5 feature thing from working 18:04:58 i'll do that after the meeting 18:05:05 jwb: can we get that in? 18:05:14 will that need a freeze break? 18:05:16 or after beta? 18:05:22 freeze break most likely 18:05:40 the original bug is still open. i'm sure we'll discuss it next blocker meeting 18:05:45 unless there are no more blocker meetings? 18:05:56 there are. oh there are. ;) 18:06:00 figured 18:06:09 Once the rebase is out on F18, I will be doing a mass update asking for people to test with 3.11 as there are a lot of stale bugs in f18 18:06:21 That's pretty much it 18:06:22 #action jwb to add fix for big_keys quota issue to F20 18:06:51 #action jforbes will do a mass bug update for f18 asking people to test with 3.11 18:06:52 #info Mass bug update for f18 coming, please retest your issues on 3.11 kernels 18:06:56 doh 18:06:56 heh 18:07:00 #undo 18:07:00 Removing item from minutes: 18:07:19 jforbes, maybe when you do that, tell them to reassign the bug to f19 if it's still there? 18:08:06 jwb: Possible that it is not visible in f19 though if the trigger changed. I know that is a small number of bugs 18:08:24 yeah, it doesn't matter hugely either way 18:08:55 ok, anything else on the releases? 18:09:01 nothing at the moment 18:09:06 #topic rawhide 18:09:16 so we're at 3.12-rc5-gitBLAH right now 18:09:27 from what i can tell, 3.12 is pretty stable at this point 18:09:41 * nirik can now test secureboot/efi junk... 18:09:47 working ok here for the most part. 18:09:48 there's nothing really major feature-wise, but i'll take boring over broken any day 18:10:27 I occasionally get stalls when doing lots of disk I/O (yum updates), but that goes back a ways and doesn't happen that often. 18:10:33 did run into backlight woes and problems with ideapad_laptop module on the new laptop... should I file those? or all just well known? 18:10:45 i think i'm going to look at adding the win8 backlight issues patches queued for 3.13 soon. a number of people are hitting that, and getting extra testing before 3.13 isn't bad 18:11:01 nirik, some known. not all. probably not "well" either ;) 18:11:23 ok. I can file the ideapad one (or ask on kernel list first) 18:11:33 either is fine for me 18:12:03 aside from that, i did some janitorial cleanup this week. shouldn't really impact anything at runtime as it was mostly dead patches, etc 18:12:18 #info rawhide at 3.12-rc5-gitX 18:12:35 #action jwb to look at adding win8 backlight issue patches from 3.13 queue 18:12:53 anything else on rawhide? 18:12:55 whats the status on beagleboneblack? 18:13:05 (if any) 18:13:06 working in f20 afaik. not in rawhide 18:13:19 mostly because of time and an underlying moving target in rawhide 18:13:25 ok 18:13:34 Are kernels appropriate for OLPC devices coming soon? 18:13:42 so tc* beta images should work on them? 18:13:49 brunowolff, what? 18:14:16 nirik, i _think_ so. might want to ask pbrobinson and/or kylem 18:14:22 or in #fedora-arm 18:14:34 can do. 18:14:35 Right now olpc runs mostly Fedora, but use special kernels. I was wondering if Fedora ARM kernels will be available on OLPC soon. 18:14:48 Maybe that's a question for the ARM people? 18:15:06 probably. i haven't paid attention to OLPC in ages. i didn't even realize they were using ARM 18:15:20 sounds like a great step up from geode or whatever they were using before 18:15:52 I thought my 1.75 was using ARm, but maybe I am wrong. 18:16:47 ok, let's open floor it 18:16:50 #topic open floor 18:17:31 Any advice in trying to narrow down kernels bugs that look like they may have been triggered by build system changes as opposed to just kernel code? 18:18:29 i saw your question on the list and didn't have any great ideas. if it's really a toolchain issue, you can always use mock to init the root to a certain state, copy in different toolchain rpms, and chroot in update/rebuild 18:18:36 aside from that, not really 18:18:54 I'm still working on tracking down a bug that came in with 2.6.29.1. But it seems like it might really have been something that changed with the build system. 18:18:55 brunowolff: wonder if that reproducable builds thing that halfie was working on could help? 18:19:07 then you would have the exact same buildroot and could tweak just one thing at a time? 18:19:55 I am not sure if I can still get all of the rpms. It was pre-f11 release when the problem showed up, 18:19:59 https://github.com/kholia/ReproducibleBuilds 18:20:12 yeah, dunno. 18:20:22 brunowolff: it came in with 2.6.29.1, but is it still a problem? 18:20:38 Yes, the problem is current. 18:20:40 jwb: BBone Black question? 18:20:56 However the machine seems to hang in a way that makes it very hard to get crash dumps. 18:21:05 pbrobinson: just wondering if it works with f20beta-tc* images? 18:21:07 pbrobinson, nirik was curious if the f20 tcs for beta are working on BBB (and also what rawhide status was) 18:21:21 I haven't been able to get one. 18:21:38 Otherwise I'd be attacking it from that angle. 18:22:00 nirik jwb: yes, they are from tc4 and later but you need a serial cable. I've got 3 half written blog posts on the BBBlack which I need to complete tomorrow 18:23:00 pbrobinson: cool. no worries. 18:24:49 I actually haven't tested in on a current kernel for a month or so. I did want to retest after the gcc fix that was recently highlighted, but I don't think that is likely to affect this. 18:25:39 When I tried last night I found the pulse audio config tool was broken so I couldn't switch sound over to the speakers instead of the headset on rawhide. 18:27:11 And my F10 testing has been complicated by some other audio problem where I am also having trouble switching audio over to the speakers. But I just started playing with that late last night. 18:28:22 All of my tests on kernels built pn f11 after it was released have had the problem. So I wanted to go back to f10 to see how builds on that worked. 18:30:12 if we have any ideas on that, we'll reply on-list 18:30:27 anything else for open floor? 18:31:00 * nirik has nothing 18:31:27 ok, let's close it out then. for the next meeting we'll get into some updates on the test stuff, etc 18:31:34 trying to do that monthly 18:31:43 thanks for coming everyone 18:31:45 #endmeeting