18:01:18 #startmeeting 18:01:18 Meeting started Fri Aug 17 18:01:18 2012 UTC. The chair is jwb. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:18 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:01:24 #meetingname fedora-kernel 18:01:24 The meeting name has been set to 'fedora-kernel' 18:01:28 it's that time again. 18:01:33 word 18:01:39 #meetingtopic Fedora Kernel Meeting 18:01:45 #topic role call 18:01:53 ok, so the maintainers are here 18:01:57 anyone else lurking? 18:02:22 Someone's always lurking. 18:02:42 right. moving on then 18:02:53 #topic F16 18:03:52 jforbes, you want to cover this one? 18:03:56 So F16 is still on the 3.4 series right now. 3.4.9 is in updates-testing (test and karma please) 18:04:47 We are holding off on 3.5 until we can get the drm issue resolved. To that end, I am summarizing the bugs and getting an email out to lkml to see if other eyes help. We are seeing this a lot in Fedora, but it doesn't seem like anyone else is seeing it really 18:05:16 * jsmith is lurking 18:05:39 * jsmith catches up on the scrollback 18:06:02 Otherwise, there are a number of bugs that should be fixed by the current updates, and either have been abandoned by the reporter, or just haven't been closed for one reason or another, so I am getting the bug list down a bit 18:06:02 Sounds very reasonable to me 18:06:06 #info 3.4.9 in updates-testing. Test and karma please 18:06:48 ok. anything else? 18:06:52 That's all from me 18:07:00 #topic F17 18:07:04 davej? 18:07:17 the aforementioned drm problem is the biggest pain point right now. 18:07:51 there's a considerable number of people hitting a direct-io bug too 18:08:17 once those two are dealt with, things should be pretty quiet again 18:08:22 to summarize, the "drm problem" is basically a boot hang that some people are seeing with a message about a conflict between fb and the native driver. nomodeset seems to help some, but not really a solution 18:08:58 davej, i'm starting to wonder if that direct-io bug is getting ABRT hash collisions 18:09:15 in some cases yeah, a wireless bug seems to have ended up with the same hash 18:09:27 thanks abrt, you're awesome. 18:09:30 should we clear the hash? 18:09:53 we could, though it would mean a lot of manual duping would need to happen 18:10:23 I'll ping jmoyer later to see what the progress is on the direct-io bug. 18:10:31 ok 18:10:52 #info drm and direct-io issues two biggest problems. currently on 3.5.x 18:11:00 anything else on F17? 18:11:26 k 18:11:31 #topic F18/rawhide 18:11:37 i'll cover this on 18:11:38 e 18:11:56 f18 is branched now, but we're keeping it an rawhide in sync 18:12:16 i built 3.6-rc2 last night, but because of alpha freeze we're still at 3.6-rc1-git6 18:12:41 there have been a few bugs reported on the 3.6 devel kernels, but most of them are fixed or have fixes on-going upstream 18:12:55 otherwise it has been very quiet as usual. likely due to lack of testers 18:13:25 most of the Secure Boot bits are in, with only one main patch missing. it's being worked on 18:13:40 I think the fact we aren't getting swamped in those drm/fb or direct-io reports from rawhide is telling. (I don't think that bug is fixed upstream) 18:13:54 F18 Alpha will ship 3.6-rcX with debugging enabled 18:13:57 davej, yeah :\ 18:14:14 the target release kernel for F18 is still 3.6 18:14:17 davej: it is not, I think one person noticed it in a kvm guest 18:14:38 #info F18/Rawhide on 3.6-rcX. Alpha will ship with debugging enabled 18:14:51 i don't have much else for rawhide 18:14:53 you could propose the signed one as a NTH for alpha... dunno if it's that important 18:15:20 nirik, yeah, i might do that. i'll discuss with pjones and mjg59 18:15:36 there's nothing that explicitly _needs_ it yet afaik 18:15:45 that stuff is really going to need a lot of testing, so shipping it early would be in everyone's interests imo 18:16:11 davej, that's the problem. unless you have a machine that supports SB in the firmware, you can't really test it 18:16:22 and those machines are few and far between at the moment 18:16:29 jwb: I'm more concerned about it not breaking the !sb case tbh 18:16:44 ah, sure 18:17:14 anything else on f18/rawhide? 18:17:31 #topic Open Floor 18:17:35 hrm, probably ought to try and get this morning's grub2 build in as well (same reasons) 18:17:49 * rbergeron wonders how the virtual test day went? 18:17:53 err 18:17:54 fad 18:17:58 pjones, submit them all as a single update? 18:18:05 rbergeron: kind of a non-event 18:18:07 rbergeron, the vFAD was a non-event 18:18:11 jwb: hrm? 18:18:23 pjones, kernel+grub2+pesign as a single update 18:18:28 jwb: why? 18:18:31 not that this is a fedora kernel issue, but anyone happen to know why alsa's bug tracker is still down? Or if it's expected back? or where I should report my alsa bug. ;) 18:18:49 pjones, to make sure they all land in alpha 18:19:13 jwb: okay, go ahead and do that then with whichever kernel is the right one 18:19:14 rbergeron: there was not really any response at all on it, Though I am wondering if timing so close to alpha freeze was the reason 18:19:31 nirik, i do not. you could maybe use alsa-devel@alsa-project.org? 18:19:39 yeah, mailing list is probably best bet. 18:19:46 pjones, ok 18:19:51 sure, can try it... just seems... poor that their bug tracker has been down for like 2 weeks. 18:20:25 I wish our bug tracker would go down for two weeks ;) 18:20:25 poor, or genius? 18:21:03 heh 18:21:06 jforbes: did you guys hit up the test-announce list as well? it's hard to say if timing close to alpha freeze was an impact... from here out it's constant deadlines or fixing periods or newspin periods 18:21:18 oh, you did. that's the list i'm looking at 18:21:58 we've not had good luck at all getting test days or participation in the past either 18:22:17 this is an on-going issue. it's either scary or boring or we aren't yelling loudly enough 18:22:22 Virt wasn't much different. Guess our options are not sexy enough 18:22:36 maybe we could try recruiting from the kernel-newbies people. is that still a thing ? 18:22:46 could threaten to introduce intentional bugs in random kernels 18:22:52 blackmail works best! 18:23:02 maybe just a "why is the kernel reression test suite important" ... or poeple are seeing kernel and going holy crap, i can't do that 18:23:16 christ, someone call mavis beacon for me 18:23:18 Even when I see a bug I don't always know enough to produce a useful report. 18:23:31 davej: there is still a community there 18:25:24 * rbergeron thinks she will send a mail to the marketing list now that we have a human actually doing marketing as their job (YAY) and see about how we can just generally drum up more interest in test days / virtual fads, even if it's just "set this thing to autotweet every 2 hours" 18:25:57 ruth will be ever so happy to tackle marketing kernel testing. i can totally feel that 18:26:00 or even a "you too can test, it's not as scary as it seems" introduction to just testing. 18:26:22 jwb: dude, she's full of creative awesome, she can turn anything into magic. 18:26:27 i'd settle for "boot the rawhide kernel today. did it work? tell us why or why not!" 18:27:50 heh 18:27:51 If you stuck "takes 5 minutes and you don't have to reinstall" in front of that it probably would help :) 18:28:10 there is no guarantee it won't make them reinstall 18:28:32 tiny text disclaimer needed 18:28:42 LOL 18:29:41 i think it's probably less brainy than a lot of people perceive it to be - i suspect a lot of folks just see the word kernel and assume it's not in their domain, but i could be wrong 18:30:05 yeah, quite possibly 18:30:05 kernel is in everyone's domain! 18:30:47 i'll still ship a mail off in a minute to mktg and see if they have any ideas on pimping it out or how to make it sound more accessible in the cases where it really is accessible :) 18:31:09 Any kernel pimping you can get done is much appreciated 18:31:10 something I'd like to do some time is document a few cases from initial bug filing through to resolution, so people have some kind of idea. 18:31:49 #idea document some bug report cases from filing to resolution as examples 18:33:58 anything else for open floor? 18:34:52 ok, thanks everyone 18:34:55 #endmeeting