18:00:34 #startmeeting 18:00:34 Meeting started Fri Feb 3 18:00:34 2012 UTC. The chair is jwb. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:34 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:00:43 #meetingname fedora-kernel 18:00:43 The meeting name has been set to 'fedora-kernel' 18:00:57 #meetingtopic Fedora Kernel meeting 18:01:23 everybody ready? 18:01:27 let's get this party started. 18:01:39 #topic Release overview 18:01:57 so, F15 18:02:08 you want to cover f15/f16 davej? 18:02:10 sure 18:02:23 I'll get the 15 update built and pushed out today. 18:02:26 rebasing to 3.2.3 18:02:41 and a similar update to 3.2.3 for 16 at the same time 18:02:57 if all looks well, I'll push 15 live sometime next week 18:03:26 bug situation is still awful (shocker). 18:03:35 nothing else is new I don't think ? 18:03:38 #info F15/F16 will be updated to 3.2.3 today 18:03:50 #action davej to push 3.2.3 updates 18:04:02 no, nothing new that i can think of either 18:04:09 F15 F16 Rawhide 18:04:10 open: 18:04:10 389 451 144 (984) 18:04:16 creeping awfully close to 1000 again :( 18:04:28 hopefully the 15 rebase will fix a bunch 18:04:43 i don't think anyone has really looked at the f15 pile in a while 18:04:47 most of the focus has been on f16 18:04:54 yeah 18:04:59 even so, 16 is a mess 18:05:42 are there specific areas you need help, or just general bug zapping? 18:05:54 the numbers are high, but we seem to hit a lot of individual issues 18:05:57 just general help would be appreciated. 18:05:59 (aside from wireless) 18:06:03 especially things like looking for dupes 18:06:25 oh, and hibernate is still horrible 18:06:33 yeah, still not sure wtf to do there. 18:06:48 I'm curious if that has been showing up on other distros 18:06:56 we have that bug open to disable it. perhaps we should push for that in rawhide 18:06:59 might have a look through ubuntu's bugs later 18:07:16 anything else on f15/f16? 18:07:39 ok, moving on 18:07:43 f17/rawhide 18:07:54 * Southern_Gentlem waits for the updates to push new livecds 18:07:56 so we rebased to 3.3-rc1 right after FUDCon 18:08:09 things, overall, have gone pretty smoothly 18:08:11 however 18:08:23 i'm kind of convinced that is because nobody is using rawhide kernels 18:08:37 that might pick up after alpha ? 18:08:48 I am running 3.3 and it's working well for me. 18:08:48 hope so... there were a large number of compat-wireless bugs that popped up in f16 that we never even got reported on rawhide 18:09:13 even though it was the same stack 18:09:19 I don't use wireless on that machine though. 18:09:25 I think we're always going to get more people on the release than rawhide, so it's not totally surprising I guess. 18:09:26 yeah, me either on my box 18:09:52 at any rate, i haven't seen anything that is majorly broken so it seems to be the status quo there 18:10:01 #info Rawhide/F17 at 3.3-rc2-gitx 18:10:18 one thing coming up that slightly impacts us is the UsrMove feature 18:10:36 harald has a patch to have the kernel start installing modules to /usr/lib/ 18:11:07 if that lands (which seems like it would soon), then there might be some minor fallout from it 18:11:31 i believe the "kernel update after conversion results in broken initramfs" bug has been fixed, so that is good 18:12:13 on the feature-ish front, about the only thing I can think of is that the gma500 (poulsbou) drivers moved from staging to REAL DRIVERS now 18:12:16 so we're building those 18:12:39 some bits of it are still in staging though right ? 18:12:46 or was that something else 18:12:56 yeah, for some intel embedded stuff we don't really care about 18:12:59 ok 18:13:00 and maybe a fb thing 18:13:25 the original requester for DRM_PSB filed a bug to get us to switch and reported the moved drivers are working well, so i think we're good 18:14:00 i spent some time looking at merge_config.sh, which seems to be a wash 18:14:18 the details are on the kernel list. follow up is to see if we can get some other tool for verification 18:14:41 and unless i've missed something, i think that covers rawhide... 18:15:03 upstream seems to have slowed down a bit as we get deeper into -rc 18:15:27 yeah, which is good 18:15:39 I think Linus has started getting more conservative about that the last release or two 18:15:45 even refusing to pull some peoples trees 18:15:45 If you really want people testing rawhide kernels, make sure they can install them with F16 userspace. 18:15:56 tibbs: right now, that should work 18:16:02 if we do the usrmove thing though.. 18:16:08 Yeah, who knows. 18:16:18 oh, right. so i wanted to ask kay about that combination 18:16:31 I'm thinking that if you know it works, tell people it works and how to do it and you'll get more testing. 18:16:32 but he's at FOSDEM, so it will have to be next week 18:16:49 Note to self: Kay is a he. 18:17:03 tibbs, 'yum update --enablerepo=rawhide kernel' 18:17:11 ignoring usrmove, that should work just fine 18:17:20 Well, sure _I_ know how to do it. 18:17:27 But if you want general testing.... 18:17:50 i'm missing what your suggesion is 18:17:55 we might try that more often when diagnosing f16 bugs (ask the user to try rawhide) 18:18:01 Tell the masses how to do that if you want more testers. 18:18:19 davej, maybe with a caveat of "ignore the slowness, it's the debug options" 18:18:23 heh 18:18:47 on usrmove, i'm thinking if we don't apply harald's patch then everythign will still work backwards and forwards 18:19:02 because modules will still go to /lib and it's a symlink in rawhide 18:19:17 * nirik notes usrmove just tagged into rawhide. ;) 18:19:28 Oh, much fun. 18:19:41 #action jwb to talk to Kay and Harald about usrmove and installing on older releases 18:19:43 jwb: yeah leaving as is should work just fine 18:19:58 nirik, that's fine for now. the patch hasn't gone in 18:20:08 well, "fine" for the kernel anyway 18:21:25 anything else on rawhide? 18:21:51 #topic questions 18:21:58 any questions? 18:22:06 anyone bring patches so they can yell at us? 18:22:26 * tomspur wonders if these 'power limit notification' are really safe to ignore: https://bugzilla.kernel.org/show_bug.cgi?id=36182 18:22:35 the patch would be to revert the bad commit ;) 18:23:08 that sounds like something mjg59 was poking at a while ago 18:23:16 where we were reading things in the wrong order or something 18:23:46 can I help with testing there somehow? 18:23:54 bug report has it with a 3.2 kernel, so i don't think the issue is fixed 18:24:50 tomspur, not sure. you might as in the kernel.org buzilla 18:25:22 I haven't seen any change on yum allowing multiple versions of kernel-modules-extra to be installed at the same time by default. 18:25:52 that's a yum bug, right? 18:26:05 or a feature request against yum i should say 18:26:16 Yes, but you guys may have more influence than me in getting it changed. 18:26:36 i agreed to it in the bug after initially being stupid 18:26:45 i guess i can bug james and seth a bit more 18:26:58 bug 770444 18:27:09 #action jwb to pester skvidal about bug 770444 for kernel-modules-extra 18:27:50 Thanks 18:27:55 sure, thanks for the reminder 18:27:58 anything else from anyone? 18:29:04 you are all very chatty today 18:29:13 ok, i think that does it for the meeting 18:29:24 yeah, justin is on vacation I think 18:29:31 he is. 18:29:33 he was going to give an update on the testing framework thing 18:29:38 next time. 18:29:39 The problems I have are only very rarely kernel-related. 18:30:10 also, I had an action from the last meeting to look into bundling up some tests etc. since I got back from vacation, I've not had chance to look at that at all. hopefully by next time. 18:30:10 #info jforbes will enthrall us with a dictation on kernel-related autoqa testing next time. 18:30:12 At least from my perspective, things have been great recently. 18:30:54 #action davej will bundle up thoughs on what tests to throw into the autotest framework (for real this time!) 18:31:13 ok, end early ? 18:31:17 indeed 18:31:25 thank you all for coming 18:31:33 #endmeeting