#title #ubuntu-meeting Meeting Meeting started by NCommander at 15:00:14 UTC. The full logs are available at http://ubottu.com/meetingology/logs/ubuntu-meeting/2011/ubuntu-meeting.2011-09-01-15.00.log.html . == Meeting summary == ''LINK:'' https://wiki.ubuntu.com/ARM/Meeting/2011/20110901 (NCommander, 15:00:35) *Action Items *NCommander to duplicate 806751 to the right masterbug *Standing Items ''LINK:'' http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel.html (NCommander, 15:03:47) ''LINK:'' http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel-ubuntu-11.10-beta-1.html (NCommander, 15:04:02) ''LINK:'' https://wiki.ubuntu.com/ARM/Meeting/2011/20110825 - the links are the same! (NCommander, 15:05:39) ''LINK:'' http://people.canonical.com/~platform/workitems/oneiric/ubuntu-arm-ubuntu-11.10-beta-1.html (GrueMaster, 15:06:12) *ARM Server Status (NCommander, Daviey) *Kernel Status (cooloney, ppisati) *ARM Image Status (ogra, NCommander) *ARM Porting/FTBFS status (NCommander, janimo) *QA Status (GrueMaster, mahmoh) *AOB ''ACTION:'' NCommander to add linaro topic during meeting before QA (NCommander, 15:58:43) Meeting ended at 16:02:48 UTC. == Votes == == Action items == * NCommander to add linaro topic during meeting before QA == Action items, by person == * NCommander ** NCommander to add linaro topic during meeting before QA == People present (lines said) == * ogra_ (124) * NCommander (46) * ppisati (46) * rsalveti (41) * GrueMaster (40) * infinity (23) * meetingology (4) * cmagina (4) * ahs3 (2) * ubottu (1) * skaet (1) * davidm (1) * Daviey (1) == Full Log == 15:00:14 #startmeeting 15:00:14 Meeting started Thu Sep 1 15:00:14 2011 UTC. The chair is NCommander. Information about MeetBot at http://wiki.ubuntu.com/AlanBell/mootbot. 15:00:14 15:00:14 Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired 15:00:35 [link] https://wiki.ubuntu.com/ARM/Meeting/2011/20110901 15:01:51 #topic Action Items 15:02:02 [topic] NCommander to duplicate 806751 to the right masterbug 15:02:37 i think GrueMaster did that meanwhile 15:02:42 I couldn't find th eoriginal guided partitioning bug which is a dupe of 806751. Its posible it got closed by someone else, but I couldn't flush it out with Google 15:03:25 So keep this one open until fixed. 15:03:31 ++ 15:03:31 [topic] Standing Items 15:03:35 GrueMaster: I intend to 15:03:47 [link] http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel.html 15:03:55 would it help if we just filled the netboot images with zeros ? 15:04:02 [link] http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel-ubuntu-11.10-beta-1.html 15:04:08 * ogra_ needs to move WIs .... 15:04:10 * NCommander beats ogra with a stick 15:04:34 * ogra_ beats NCommander to use the right links 15:04:59 ogra_: you can update the wiki with it 15:05:04 Yes please. These workitems show nothing being done by me, which is highly inaccurate. 15:05:04 can you update them in the wiki ? (i did that for last week, why didnt you carry that over= 15:05:11 NCommander, i did ! 15:05:17 ogra_: i copied the page from last week and edited it 15:05:39 https://wiki.ubuntu.com/ARM/Meeting/2011/20110825 - the links are the same! 15:05:50 well, i updated the one i used when i was running the meeting rercentlY 15:06:08 which was the 25th :-P, did you hit save? 15:06:10 I DID IT DURING TEH MEETING (AND FILED THEM TWOCE ACTUALLY) 15:06:12 [link] http://people.canonical.com/~platform/workitems/oneiric/ubuntu-arm-ubuntu-11.10-beta-1.html 15:06:14 ARGH 15:06:19 MY CAPS IS BROKEN 15:06:29 ah, better, sorry 15:06:31 ogra_: hold shift? :-) 15:07:27 what's the correct overall link so I can save it into the page? 15:07:45 see above :) 15:07:51 GrueMaster, pasted the right one 15:08:03 I mean what's the replacement for http://people.canonical.com/~platform/workitems/oneiric/ubuntu-armel.html 15:08:04 its just s/armel/arm/ 15:08:08 oh 15:08:11 status.ubuntu.com 15:08:30 [topic] ARM Server Status (NCommander, Daviey) 15:08:35 links fixed 15:08:40 :) 15:09:14 preinstaleld pool misbehaves on preinstalled server ... 15:09:21 *installed 15:09:23 OMAP3 images are currently suffering from some ethernet issues. It looks like the root cause of the SMP ARM issue has been found (L2 caching issues), with a patch being tested 15:09:29 but infinity is on it i guess 15:10:07 NCommander, we are using the patch in ac100 already ... USB disk access went from 7M/s to 19M/s for me 15:10:13 \o/ 15:10:25 so it should help panda the same way 15:10:30 _\o/_ 15:10:34 or other SMP arms 15:10:46 are these your legs ?!? 15:10:57 * NCommander shrugs 15:11:07 I claim many talents 15:11:17 ... 15:11:21 * ogra_ cant imagine NCommander doing the splits really 15:11:32 This meeting just got weird. 15:11:36 lol 15:11:39 move then :) 15:11:39 :) 15:11:43 It usually happens when the chair is sleep deprieved 15:11:55 [topic] Kernel Status (cooloney, ppisati) 15:12:23 as already said, Ming posted a patch that fixes the disk i/o problem 15:12:38 i'm rolling a new omap4 kernel with a new BSP 15:12:42 and that patch too 15:12:55 should be out available on zinc today or tomorrow 15:13:02 * davidm waives 15:13:20 on the other hand, another round of new kernels for every release/flavour is being pushed out 15:13:32 sound seems to get critical 15:13:39 there are no devices at all 15:13:39 i know 15:13:52 it's on my list 15:14:00 Hopefully the new bsp fixes it. 15:14:03 it has been disabled on purpose 15:14:05 and we surely will need some time for adjusting userspace 15:14:20 actually the new kernel introduced new options 15:14:27 ppisati, thats fine, but we need it back soon, else the alsa changes and pulkse fixes come to late 15:14:28 and these options (as per defaultconfig) 15:14:39 break omapfb 15:14:45 and i would like that we after three releases finally have one with working sound 15:14:50 ogra_: i can turn it on now 15:15:01 ogra_: but then pulseaudio will go nuts 15:15:05 do that and lest see how userspace explodes :) 15:15:11 ooooooook :) 15:15:13 so 15:15:16 i push out the new kernel 15:15:23 and then i revert the "no sound" config 15:15:24 well, because fo kernel bugs or pulse bugs ? 15:15:29 pulse 15:15:41 if its pulse side we actually want it to break 15:15:48 ok, i'll do that then 15:15:49 so we can see if its fixed at some point ;) 15:16:01 and we have a pulse guy in #ubuntu-arm atm 15:16:04 We can't break the apps if there is no driver support. 15:16:09 yeah 15:16:14 yep 15:16:20 +1 15:16:36 but i din't want to push out a kernrl that would break everyone who updated 15:16:41 i hoped for a fix 15:16:51 but anyway, let's turn it on again 15:16:54 and have fun :) 15:16:58 ppisati: sometimes the best way to get that is to break the world 15:16:59 Better to do it now than after release. 15:17:14 on a sidenote ac100 kernel is pending a new git checkout, i will do that after beta (some time next week) 15:17:17 anything else? 15:17:24 nothing else here 15:17:32 As long as it still boots, it can be debugged and fixed. 15:18:42 [topic] ARM Image Status (ogra, NCommander) 15:19:10 well, apart from omap3 netboot everything we built for beta1 is releasable 15:19:24 mx5 waits for the milestone freeze to fall 15:19:26 ogra_: omap3 has a documented workaround, and the ethernet partially works 15:19:28 same for ac100 15:19:42 Yes 15:19:51 NCommander, oh, telll that in -release please, i marked it as unreleasable 15:20:07 ogra_: I poked skaet last night but no ping reply 15:20:07 omap3 is workable. Painfull but workable. 15:20:17 all i read was "beagle netboot is unbootable after install" 15:20:20 Little late, we just released. 15:20:21 I told her as much yesterday. 15:20:31 sigh. 15:20:37 oh well 15:20:38 NCommander, i updated the pad this morning after checking the images and isotracker 15:20:51 so its marked as NO on the pad atm 15:21:22 btw, could we pretty please stick to do *all* image testing conversation in #ubuntu-arm 15:21:24 And you didn't ask me what the situation was? And didn't read the bug info? 15:21:35 so its in logs and retrievable later 15:22:07 GrueMaster, all i saw was a non passed test with two blocker bugs and there was no conversation about testing in any public channel 15:22:20 so i could only judge by the data i had 15:22:47 We can still pull images if we need to. 15:22:47 Who did you want me to talk to? Myself? 15:23:13 Meh. It's not the end of the world. 15:23:16 And there was some minor discussion on #ubuntu-release. 15:23:22 GrueMaster, you mean me ? 15:23:29 We can push that image out with beta, or we can not care. Both are valid options. 15:23:46 GrueMaster, well, i see about 3h of conversation in #arm from last night 15:24:46 iso tracker currently indicates that netboot-omap is rebuilding??? 15:24:55 heh 15:25:02 probably a d-i uplaod 15:25:17 cant avoid it if there is a rebuild for another arch 15:25:25 anything else w.r.t to images? 15:25:35 nope 15:25:35 omap4 isn't rebuilding though. 15:25:50 tracker issue i would say then 15:26:46 sorry, late, ogra_ enabling sound at omap 4 will make pulseaudio to consume 100% of the cpu 15:27:01 the image will be unusable 15:27:04 rsalveti, fine, so there is a reason to fix it 15:27:30 there's always the reason to fix it, just don't know if enabling it now would be the best choice 15:27:32 rsalveti, we have a pulse maintainer in #ubuntu-arm, how would he fix it without being able to reproduce 15:27:39 as it'll probably block desktop testing 15:27:50 We need a chance to fix it sometime. 15:28:00 for that ppisati can produce new packages and make it available at a ppa/people.canonical 15:28:05 just my 2 cents 15:28:07 And if we can't, disabling/crippling Pulse on ARM is saner than a kernel with missing drivers. 15:28:12 well, actually he could roll his own kernel 15:28:15 tobin knows how to chmod -x a binary on the image before testing i bet :) 15:28:16 with audio enabled 15:28:16 If the userspace is broken, fix or disable the userspace. 15:28:27 fix is the attempt 15:28:46 s/on ARM/on OMAP4/ 15:28:50 rsalveti, the prob is that alsaucm isnt done either yet 15:28:55 yup 15:29:05 the linaro multimedia working group is on it for this month 15:29:18 do they fix it in the archive ? 15:29:19 don't expect them to fix it, but they should know how to get it done for oneiric in some sort 15:29:37 * ogra_ knows how to get it done ... i know that every release ... 15:29:44 ... about a week before release sady 15:29:47 *sadly 15:29:53 :P 15:29:57 :-) 15:30:08 anyway, we have a bug for that already, let me find it 15:30:10 i want to have that sorted eralier this time 15:30:30 preferably minimally working sound for b2 15:30:48 bug 816638 15:30:50 Launchpad bug 816638 in Linaro-Ubuntu "Pulseaudio consumes 100% of the cpu when trying to play a sound with natty's linaro LEB and 3.0.0-1402-linaro-lt-omap " [High,Confirmed] https://launchpad.net/bugs/816638 15:30:51 rsalveti: while here 15:31:04 rsalveti: the new tilt introduced a DRM_OMAP option 15:31:16 rsalveti: and updateconfigs put it =m 15:31:24 rsalveti: but doing this we loose omapfb 15:31:37 ppisati: put it as =y and replace omapfb 15:31:39 rsalveti: do you know what's the correct config in this case? 15:31:49 rsalveti: DRM_OMAP? 15:31:56 ppisati: yes 15:32:08 ppisati: but test it first, I wasn't able to test the new drm driver still 15:32:17 but it's fine to just replace omapfb 15:32:17 rsalveti: well 15:32:26 X doesn't start in my case 15:32:35 insist in trying to use omapfb 15:32:43 and my fb console is powered on 15:32:44 ppisati: let's follow this off-line 15:32:48 but nothing is printed 15:32:50 ok 15:32:54 I want to check it with the leb images today too 15:33:32 if you want i can push it as is to zinc 15:33:41 you can take a look at the config 15:33:46 ppisati: please at least make it available at a git tree somewhere 15:33:48 and tell me what do i have to tweak 15:33:49 and paste me the link :-) 15:34:02 anything else? 15:34:03 that was the plan, ok 15:35:03 NCommander: move 15:35:04 * Daviey says hello? 15:35:16 [topic] ARM Porting/FTBFS status (NCommander, janimo) 15:35:43 libx86 is still ftbts !!!! 15:35:49 :) 15:36:05 oh, why does the kernel fail 15:36:13 They're on it. 15:36:19 ah, k 15:36:20 Yea, we need that to help debug libdrm-intel. 15:36:22 They accidentally enforced an x86-only config. 15:36:26 looks good so far 15:36:32 rsalveti: git://kernel.ubuntu.com/ppisati/ubuntu-oneiric.git ti-omap4-next 15:36:45 ppisati: thanks 15:37:04 apart from likewise, NCommander Daviey, is that ciritical in any respect ? 15:37:09 should finally be able to post the ftbfs bugs for the rebuild late today 15:37:12 NCommander: Did I hear talk earlier in the week (or last week) that the likewise-open failure was a repeat of one you'd fixed earlier? 15:37:13 for server 15:37:24 infinity, twice iirc 15:37:28 infinity: several times :-/ 15:37:34 So, easy to upload a fix? 15:37:38 no 15:37:44 do we even still use it in server ? 15:38:08 ogra_: It's in main. "Use it in server" doesn't mean much, since we install nothing by default. :P 15:38:40 it used to be in the server ship seed ... which kept it in main 15:38:56 i'm just wondering if its still needed 15:39:00 or wanted 15:39:00 Since it works with active directory, it should be fixed and tested (why did I say that?) 15:39:12 a volunteer ! 15:39:23 damn. 15:39:38 heh 15:40:10 Well, the upshot is that he can physically assault NCommander to find his old patches. 15:40:13 Anyhow. Moving on? 15:40:20 yeah 15:40:40 NCommander, ? 15:41:33 ........ 15:43:27 [topic] QA Status (GrueMaster, mahmoh) 15:43:45 * NCommander was being yelled at by the local powers that be 15:44:03 Not much testing has been going on with the preinstalled images due to the workload on server testing. 15:44:03 yell back ! 15:44:26 * ogra_ noticed that as well 15:44:48 in the past we all were jack of all trades somehow in the team ... 15:45:05 over time we all have our special focus and somehow nobody but you tests regulary 15:45:12 that used to be different in the past 15:45:16 On the server side, a lot of tests have had to be created from scratch as there hasn't been tests in those areas. 15:45:35 i was wondering if we should have a rotating schedule for non QA people to have regular test tasks 15:46:07 iSCSI somehow breaks the kernel to lock up the system. Appears to be initiator side only. 15:46:15 We should all be testing more as we near the home stretch, I suspect. 15:46:22 i know that if i personally test i also fix the bugs as i catch them sometimes ... 15:46:25 Well, unless we're to busy fixing RC bugs. :/ 15:46:33 That would help. Especially if someone just popped in the desktop image once a week. 15:46:39 so it helps to have devs do reguzlar testing :) 15:47:13 GrueMaster, lets work ouot a plan that doesnt put to much extra work on each individual but makes sure we get regular dev testing too 15:47:19 (offline) 15:47:22 In previous cycles, I would do deep testing between releases, filing bugs and taking notes. Then at release time, I would list all the bugs still open. 15:47:32 ok 15:47:59 that was all from me 15:48:17 On the server side, I would still like it if someone could take a look at tahi.org and port the ipv6 test suite to Linux (from BSD). 15:48:51 Or at least give me some sort of tangible test plan. Currently there is none. 15:49:32 bah, why dont they have one big tarball 15:49:55 each test seems to have its own 15:50:03 GrueMaster: tahi is looking to be quite a bit of work to port at least for one who is unfamiliar with the code. 15:50:31 Tell me about it. I had started doing it myself, and I am not much of a programmer. 15:50:33 GrueMaster: the suse package contains one giant patch to port it 15:50:35 whats there to port actually ? 15:50:55 it operates in userspace on top of a defined protocol 15:51:05 oh, its bsd only ? 15:51:09 yeah 15:51:10 * ogra_ didnt get that 15:51:12 sorry 15:51:38 it's got bsd-ism's all over the place...nasty to port... 15:51:45 I did hours of searching for a linux test suite. The suse port was all I could find, and it is old. 15:51:49 i started to poke around to see if there are any alternatives, haven't seen one yet 15:51:54 and there is no equivalent tool onm linux ? 15:52:03 hrm 15:52:19 While I agree that there's some value in a decent IPv6 test suite, we clearly don't have one on amd64/i386 either... 15:52:27 So, I'm not sure this task is ours to worry about. :P 15:52:30 * ogra_ doubts thats really 11.10 material then 15:53:00 no, no real equivalent. and TAHI is the only org willing to give out any sort of certification for ipv6 15:53:21 I'll look to see if there is at least some documentation for enabling ipv6 dns/dhcp and do some rudamentary testing from there. 15:53:56 Our server testing is very lacking overall (not just on arm). 15:54:42 The few we have is highly dependent on libvirt/kvm, which doesn't help armel testing. 15:55:04 cant libvirt deal with actual qemu VMs too ? 15:55:23 would be slow as hell indeed 15:55:53 It can. 15:55:56 Not sure, but I don't trust qemu to do a lot of these tests. I don't trust kvm for the same reason. Too much can break on raw silicon. 15:55:59 i thought so 15:56:27 well, kvm us just qemu with kernel side support 15:57:19 anyway, we're running out of time 15:57:24 NCommander, ? 15:57:33 [topic] AOB 15:57:40 * rsalveti waves 15:57:50 as I said in the private channel 15:57:59 I'd like to have a topic just related with linaro 15:58:05 anyone coming to plumbers? 15:58:06 so I can explain what we're planning to deliver 15:58:11 ++ 15:58:13 and improve the communication with the ubuntu arm team 15:58:16 NCommander, can you add that 15:58:22 k 15:58:29 ppisati: I'm at Plumbers. 15:58:34 before QA. 15:58:43 [action] NCommander to add linaro topic during meeting before QA 15:58:43 * meetingology NCommander to add linaro topic during meeting before QA 15:58:55 :D 15:59:03 great 15:59:12 for this month we're just about to finish the 11.09 planning 15:59:13 now GrueMaster can sleep longer :) 15:59:24 Pfft. hardly. 15:59:30 mostly related with CI work, so nothing that will touch ubuntu atm 15:59:41 infinity: so, see you there :) 15:59:42 CI ? 15:59:50 and for u-boot we may have more PXE working platforms and USB working at SPL as omap4boot 15:59:53 continuous integration 15:59:58 ah 16:00:05 i thought common interface :) 16:00:08 the idea is to build daily packages for kernel/u-boot and feed that at lava 16:00:14 its the only context i ever herad CI in 16:00:16 to get properly tested before landing at ubuntu and such 16:00:27 * GrueMaster had guessed Crazy Ideas. 16:00:31 but for u-boot it may be quite late for the cycle 16:00:32 hehe 16:00:57 the only thing that we know for sure that will hit ubuntu is the linux-linaro kernel packages update 16:01:03 rsalveti: Will we have usb/pxe on beagleXM? 16:01:05 that should reflect the ones released at 11.08 16:01:25 GrueMaster: idea is to have it working at beagle xM, igepv2, vexpress and imx53 16:01:37 cool 16:01:48 vexpress ftw ... that could go with a libvirt setup :) 16:02:24 We may be out of time. 16:02:26 but that's all from my side 16:02:35 I'm done ;-) 16:02:40 infinity, may be ? we are definitely ... :) 16:02:43 anything else or can I clsoe it out? 16:02:47 go 16:02:48 #endmeeting Generated by MeetBot 0.1.5 (http://wiki.ubuntu.com/AlanBell/mootbot)