[00:37] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/793796 i filed a bug about it, if anyone has appropriate tags to add; or people to subscribe [00:37] Ubuntu bug 793796 in linux "2.6.38-10 panic after ejecting drive" [Undecided,New] [00:53] woo confirmed, i wonder if someone was waiting for a report :D === michael is now known as Guest65036 === _LibertyZero is now known as LibertyZero === _LibertyZero is now known as LibertyZero === ericm-afk is now known as ericm|ubuntu [08:07] ohsix, the confirmed update was likely a bot [08:07] ohsix, can you get the picture attached to the bug please [08:07] i attached one, did it not make it? [08:09] ohsix, bah you did, stupid interface hiding it from me [08:10] there were only 2 scsi/block/vm patches in the changelog that might have something to do with it but none that i can see how [08:12] you can probably reproduce it easily enough; it's on x86_64 but i don't know if that matters [08:12] ohsix, yep i am sure they will, i've seen similar on .39, but neither of those boxes boot at the moment [08:12] i've tagged it up for the stable peep to sort out [08:13] i was trying .39 from the xorg-edgers ppa for other reasons & hit it a week or two ago; didn't report a bug but i did tell one of the guys on top of that stuff [08:21] Morning all [08:23] Ive managed to reproduce https://launchpad.net/bugs/792291 and the machine is still on, is there anyone about that can help me get anything out of the machine to help debug? [08:23] Ubuntu bug 792291 in linux "Machine hung, display frozen" [Undecided,Confirmed] [08:27] jussi, do you know what you did to reproduce it? if so record that [08:27] i seem to remember we didn't know if the machine is still alive on the network or not, you could try and determine that [08:28] apw: unfortunately not, its happened several times, at random intervals. [08:28] apw: let me try [08:29] Hrm, Ive no idea of that machines ip (we have a dhcp system here), but the network lights near the cable are still blinking [08:29] jussi, if you know the machine name, name.local often works as a hostname on the local lan [08:30] yeah, it doesnt appear to be reponsive [08:31] jussi, and you have tried the sysrq options out? [08:32] ahh, reisub rebooted it. [08:33] jussi, so in theory you may have some output in your logs sync'd to disk maybe, cirtainly worth looking now [08:33] jussi, and record that reisub did something too [08:33] apw: ok. which logs in particular are wothe looking at ? [08:34] syslog contains the output of dmesg if we are lucky === smb` is now known as smb [08:34] looknig for sysrq may help [08:34] morning [08:34] smb, moin [08:35] jussi, was this the one with a moving mouse cursor ? [08:35] apw: no, total freeze, different machine [08:36] is it safe to pastebin the syslog? [08:36] jussi, has this problem aways existed on natty? [08:36] (trying to work out where it might have been introduced) [08:36] apw: we only bought/installed the machine a few weeks ago, and it has happened several times since [08:36] jussi, i think pastbin is ok, you could grep for your password and name before [08:37] jussi, ok so we have no idea if its a regression [08:37] jussi, was there anything interesting in syslog ? [08:38] apw: Im looking, but let me share it with you also, because Im not certain what to look for. [08:41] jussi: are you running 2.6.38-10 on that? [08:42] ohsix: no, 2.6.38-8 (x86_64) [08:45] Jun 7 10:32:05 monster kernel: [81028.414275] SysRq : Keyboard mode set to system default [08:45] Jun 7 10:32:07 monster kernel: [81029.756536] SysRq : Terminate All Tasks [08:45] ok so the machine was alive per-se [08:45] as the logging deamons saw the sysrq keys and reported them [08:46] so that is useful information and should be recorded in the bug [08:47] apw: Ill attach that syslog if you think its useful (and doesnt contain too personal stuff) [08:47] jussi, also the fact there isn't any panic before the Sysrq [08:47] right, so its likely not in the kernel then. [08:47] jussi, include like 5 lines before the SysRQ and the first couple of sysrq things [08:47] jussi, its as likely X i'd guess [08:48] so next time jussi could you do the sysrq-R then try to switch VTs [08:48] and also then try Sysrq-w, to see if anything is hung [08:48] then perhaps Sysrq-t [08:48] apw: ok, sure. Thank you for taking the time to help. [08:49] now the latter will take a while likely, you should be able to see it producing disk activity i suspect as the logs are being pulled out [08:49] (maybe leave it for 1min after then continue with the reisub) [08:49] then look for those in your syslog and see hwats there [08:50] apw: excellent. Ill note that down for next time something happens. [08:55] jussi, good luck [08:56] apw: Thanks - I think Ill need it. :D [08:57] jussi, i've recorded that in the bug for posterity [08:59] apw: excellent! :D [09:01] hi, where are the daily builds? also 3.0-rc2 is missing [09:05] Kano, they are missing because some fool changes the version number radically and breaks every damn tool we have [09:06] well that was done with rc1, no time to fix em yet? [09:07] yes we fixed it in part last time, but not comopletely it seems [09:10] haha [09:12] There was some discussion on the release announcement about the tarball being screwed, at least originally. [09:12] Presumably the kernels are built straight from git though? [09:12] Tommeh, luckily we don't use the tarballs [09:13] ah [09:13] but we are getting hammered by the tags not being v2.6.* [09:20] what fun [09:21] i know ... and right now i can't build them with the right version as module-init-tools specifically depmod won't work [09:21] ARRRRG [10:17] * apw bounces to verify a unity bug which someone has FIXED !?! [10:23] lol [10:28] seems to work too ... amazing, shame i have found another new bug now [10:28] hey apw [10:28] jk-, hiya === jibel_ is now known as jibel [14:26] apw, ogasawara, AceLan: tangerine needs a reboot for kernel update. lemme know when is a good time. [14:26] tgardner, don't believe i am using it currently [14:27] tgardner: am in the middle of a quick build, will ping you when it's done [14:27] apw: I'm hopefully zeroing in on the bad commit for the hp mini [14:28] ogasawara, awsome, bisecting ? [14:28] apw: yep. your mainline-build-one script comes in handy [14:28] :) thats something [14:30] I went ahead and pushed the -rc2 rebase to master-next, if anyone wants it [14:38] ogasawara, i assume its no more working than before [14:38] apw: nope, still the same [14:40] tgardner: rsu added [14:40] *SRU [14:40] ppisati, ack [14:43] tgardner: your good to reboot tangerine from my end [14:44] bouncing [14:44] you're [14:46] damn, you know it's bad when apw corrects your misspelling :) [14:46] ogasawara, an embaressment indeed [14:46] hehe [14:47] ogI just figured apw had blurted a sentence fragment [14:47] ogasawara, ^^ [15:02] thunderbid formatting of emails is really annoying [15:02] i can't make a message without stupid formatting... [15:03] tgardner, ogasawara: can one of you accept my natty nomination on bug 770232 ? [15:03] Launchpad bug 770232 in linux-firmware "148f:3072 Ralink RT3072 WLAN not working with rt2800usb - firmware missing" [Undecided,Incomplete] https://launchpad.net/bugs/770232 [15:04] sforshee, done [15:04] tgardner, thanks! === michael is now known as Guest40009 [15:37] ## [15:37] ## Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting [15:37] ## agenda: https://wiki.ubuntu.com/KernelTeam/Meeting [15:37] ## [15:47] * ogasawara back in 20 [16:02] ppisati, remind me, mvl-dove is identicle in lucid and maverick isn't it, and we update lucid [16:05] apw: yep [16:05] apw: i think tomorrow i'll do the rebase [16:05] ppisati, whenever, just interested for a cve i am finishing up [16:23] tgardner, about ? i see we don't have armel cross compilers in our lucid chroots on tangerine, is this something we can get easily from somewhere? [16:24] apw, I'm not sure they exist for Lucid unless we pull the deb from Maverick [16:24] oh hrm, bugger [16:25] herton, please have a look at bug #794096. It loooks like it could be a stable update regression. [16:25] Launchpad bug 794096 in linux "SMTP and posting to a web-form time out (probably due to netfilter changes)" [Undecided,In progress] https://launchpad.net/bugs/794096 [16:25] tgardner: ack [17:08] * herton --> lunch [17:16] smb, about? got a sec to review a debdiff for m-i-t for me ? [17:16] apw, currently taken up in server meeting [17:16] a min [17:16] smb np [17:18] apw, k, they moved on. what where? [17:18] chinstrap:~apw/DEBDIFF [17:33] HRM somethign bad happened when i right clicked on a link [17:33] X exited [17:44] ## [17:44] ## Kernel team meeting in 15 minutes [17:44] ## [17:45] already? where did the day go? [17:45] yep [17:53] bah, it helps if I actually install the test kernel I downloaded before I reboot [17:54] ## [17:54] ## Kernel team meeting in 5 minutes [17:54] ## [18:00] Gah! its time [18:05] smb: a minute later and you would have missed the meeting :) [18:06] Sarvatt, Too true. :) [18:43] apw: cool, I've isolated the regression [18:43] commit 6067aaeadb5b3df26f27ac827256b1ef01e674f5 [18:43] Author: Jesse Barnes [18:43] Date: Thu Apr 28 15:04:31 2011 -0700 [18:43] drm/i915: split clock gating init into per-chipset functions [18:43] [18:43] This helps contain the mess to init_display() instead. [18:43] apw: and it appears there's already a fix coming down the pipe [18:43] apw: https://lkml.org/lkml/2011/6/5/12 [18:43] apw: gonna build us a test kernel to verify [18:48] ogasawara, i just modified the moin paragraph levels that kt-meeting-stats generates, be sure to use the latest version [18:48] bjf: ack === bjf changed the topic of #ubuntu-kernel to: Home: https://wiki.ubuntu.com/Kernel/ || Natty Kernel Version: 2.6.38 || Ubuntu Kernel Team Meeting - June-14 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! [19:03] * tgardner --> lunch [19:16] ogasawara, awsome, let me have some test kernels when you've tested :) [19:27] * apw waits with baited breath for ogasawara's testing results [19:58] apw: \o/, it works. debs are on tangerine in my oneiric-i386 dir [19:59] ogasawara, this is for an HP mini? [19:59] tgardner: yep [19:59] tgardner: am gonna push the fix to master-next [19:59] ogasawara, cool, I'll give it a try as well. [20:01] apw: if you can, would you also test on the other atom kit you said was having issues with -rc1. just curious if it's still there with an -rc2 kernel. [20:07] ogasawara, i've got both 32 and 64 bit suffering, got and 64bit debs ? [20:07] apw: I'll kick off a build right now [20:21] ogasawara, good on my 32 bit machine (not the mini) [20:21] yell when we have amd64s [20:21] apw: should almost be finished up [20:21] apw: it'll appear on tangerine:oneiric-amd64 [20:23] * jjohansen -> lunch [20:23] ogasawara, i'll watch for it [20:26] apw: generic flavour should be there now [20:33] ogasawara, how come you aren't using 3.0.0-0.1 in the oneiric changelog ? There are still 3 version digits in the Makefile [20:34] tgardner, cause it isn't going to be called 3.0.0 but 3.0 when it releases [20:34] tgardner, and if we use 3.0.0 we can't go back down to 3.0 which is < [20:35] he only put the extra .0 in the makefile cause you couldn't build it without it at the time [20:35] apw, but this is a packaging version thats only loosely related to the kernel version. [20:36] tgardner, that is true, but we were trying to continue the current system which is to use the version before stable [20:37] apw, and I suspect the first stable version will be 3.0.1, right? [20:37] tgardner, yep, but we don't expose the stable version in our package version [20:37] 2.6.38.8 is still 2.6.38 in the archive [20:38] apw, right, as will 3.0.0-ABI-N [20:39] it will simplify a lot of our scripting problems if we use 3 digits. [20:39] tgardner, we can do that, its just a change in meaning [20:39] tgardner, though all the scirpting issues we have look to be cause we have Ubuntu-2.6 as searches and the like [20:39] not so much because of the length of the prefix right ? [20:40] apw, well, I was thinking about the regexes that extract ABI and the like [20:40] they should be using the 'before' and 'after' really [20:41] they should be using the 'before' and 'after' the - really [20:41] apw, perhaps, but fixing them in all places will be a PITA [20:42] tgardner, i think that most of it is working now, that isn't using 2.6 as far as i can tell [20:42] I'm thinking about the LTS backports as we go forward. [20:43] All of our Lucid stuff wants 3 digits. (I just don't know where yet) [20:43] the lts backports carries the whole changelog and debian machinary with it no? [20:43] the only thing which needs a little love is depmod and that is easy to fix [20:43] likely, but its easy to avoid the 'noid by just carrying the 3rd digit. I hate inconsistencies and special cases [20:44] well it feels like an inconsistancy to represent 3.0 with an extra digit and not 2.6.x [20:45] but i am happy to be persuaded [20:46] well, lets see how it goes. we've got a little time before things get cast in concrete [20:46] tgardner, as long as we start with 3.0 we can move 'up' to 3.0.0, the inverse is not possible [20:48] ogasawara, ok this looks good for me, i'll get these two uploaded together and tested, then we can upload to the archive in the morning [20:48] apw: ack, sounds good [21:15] hi - just installed natty. would like to try out xen - any pointers? I last tried xen about a year ago, but heard that the 4.1 is much better [21:16] google points me to pages that require manual kernel config/make [21:16] is this the right place to ask? [21:19] Guest24499, i am not sure that the .38 kernel has all the pieces you need to for a dom0 kernel which makes it hard to do [21:19] previous versions of ubuntu has xen-dom0 kernels - natty doesn't have it? [21:21] Guest24499: hardy was the only kernel that had all the pieces for xen-dom0 [21:21] ahh [21:21] they have all had the domU pieces [21:21] I guess I was lucky when I last tried xen I hit hardy :) [21:22] the oneiric kernel (11.10) will have the pieces for dom0 and domU [21:22] should I try 11.10 then? is it even in beta yet or in a position to be used? [21:22] Guest24499: its because since hardy we have tried to follow upstream, where in hardy we had a special xen kernel [21:23] Guest24499: its not in beta yet, and the upstream 3.0 kernel that has all the domo pieces is at rc2 [21:23] and uhm isn't the most stable thing at the moment [21:23] so, would it be a good idea to follow https://help.ubuntu.com/community/Xen and compile my own kernel? [21:23] at least from my experience [21:23] heh. [21:24] Guest24499: you could, or you could try the oneric kernel [21:24] man, linux has changed so much from my slackware days... [21:24] Guest24499: just don't expect all the bugs to be ironed out yet [21:25] so the oneric kernel should be ok for dom0? [21:27] Guest24499: yes it should [21:28] any pointers or URLs for a quick howto to pin a oneric kernel to natty? [21:28] I would just download and install the package [21:28] dpkg -i .deb [21:29] Guest24499: packages can be found on packages.ubuntu.com [21:30] ok, thanks [21:30] just search in oneiric for linux === yofel_ is now known as yofel [21:34] umm, I'm in http://packages.ubuntu.com/oneiric/kernel - don't see any kernels? [21:36] http://packages.ubuntu.com/oneiric/linux [21:36] Guest24499: ^ [21:37] ahh, in admin [21:37] not in kernel [21:37] ogasawara, so, how are you building this oneiric pile ? I'm getting some really weird errors. 'WARNING: Couldn't open directory /home/rtg/oneiric/ubuntu-oneiric/debian/linux-image-3.0-0-generic//lib/modules/2.6.38-10-server: No such file or directory' [21:38] tgardner: using the kteam-tools buildscripts. but a usual fakeroot debian/rules binary-generic should work too [21:39] ogasawara, thats what I'm doing. hmm [21:41] ogasawara, are you keeping a log of your builds? I'm thinking this might be some fallout due to the 2 digit version number. to whit: 'Missing argument in printf at debian/scripts/abi-check line 174.' [21:42] tgardner: yep, look on tangerine:oneiric-amd64/build.log [21:42] tgardner: you have apw's patches to fix up the version? [21:43] ogasawara, no, I just pulled from the repo. [21:43] tgardner, that is the bug with depmod, you need an up to date m-i-t in the chroot [21:44] apw, ah, thats it [21:44] tgardner, tangerine in my home are some fixed ones [21:45] apw, which one ? [21:45] once they are built in my PPA i'll be uploading them to the archive [21:45] 3.13-1ubuntu1~test4 ? [21:45] those sound right yes [21:46] module-init-tools_3.13-1ubuntu1~test4_amd64.deb [21:46] module-init-tools_3.13-1ubuntu1~test4_i386.deb [21:46] those are the right ones [21:53] ogasawara, this kernel is looking nice, am pretty sure my fans are on less [21:54] apw: definitely in much better shape than we were a few days ago [21:56] ogasawara, yeah ... [22:13] ogasawara, LKML: Re: 3.0.0-rc2 fails to boot on Atom appliance (bisected, drm/i915) [22:15] tgardner: thanks will take a look [22:16] ogasawara, looks like you bisected to the same patch [23:35] apw, ogasawara: bug 771445 in hardy (commit 5caf3ae4c4bed98bd6148021e6e934d94b5dea1d) mentions "backport of commit 272b62c1f0f6f742046e45b50b6fec98860208a0" but it is actually a backport of commit b00916b189d13a615ff05c9242201135992fcda3. the resulting hardy kernel is fine, but we have now a tracking glitch. how should we handle this? [23:35] Launchpad bug 771445 in linux-ti-omap4 "CVE-2010-4655" [Undecided,In progress] https://launchpad.net/bugs/771445 [23:37] kees, hrm, thats a bit of a problem, i suspect we need an exceptions table for my match up tool [23:38] kees, to handle that ... let me think about it [23:38] apw: okay, cool. [23:38] i think only my tool needs to handle any errors [23:39] well, it's a bit odd since I'll only discover this when comparing u-c-t against the changelog. in this case I found an extra CVE in the changelog and went looking for it, at which point we can add exception maps, etc [23:41] kees, yeah, but i suspect there will be more than one error in there over time [23:41] and i've already wanted to mark an old commit as being a cherry-pick when it wasn't which would use the same [23:41] mapping table. [23:41] anyhow will let the shower brain think about it [23:42] i fear my tooling will keep unding any corrections otherwise