=== jjohansen is now known as jj-afk === yofel_ is now known as yofel [03:48] what path in git should I use to be able to pull the latiest at all time? like git clone git://kernel.ubuntu.com/ubuntu/? [03:55] sacarlson: "latest" isn't well defined. do you want stable or dev? if dev, then mainline or some experimantal branch. if main, then ubuntu or linus's? I think I can go on... so what do you want? [03:55] I guess dev [03:55] I want the top master [03:56] or at least a path to it I can alway go back [03:57] I want to compare what I now run in a webcam driver ibmcam with all that ubuntu has or will have. I already have the master from kernel.org I would like to compare ubuntu to that also [03:57] git clone git://kernel.ubuntu.com/ubuntu/ubuntu-intrepid.git ubuntu-intrepid [03:58] that's what I have from when I was doing something... sounds like what you want.. maybe. [03:58] I thought we already have like maverick or something and beyond that [03:58] "...from when I was doing something" [03:58] and if new comes is I pull will that move from intrepid to maverik? [03:59] no. but maverick is stable. dev is natty [04:00] but I thought releases were in alpha order like abcd intrepid is like I J K L M maybe maverick dev would be above that? [04:01] what comes after M? :) [04:01] N? [04:01] yes. [04:01] and what does *N*atty begin with... [04:02] n [04:02] or are you still not understaning my git line? [04:02] I was trying to debug a kernel crash back in intrepid time. so that was what I was pulling back then. [04:03] I don't see the N in your git line [04:03] dev is natty [04:03] today. [04:03] clone git://kernel.ubuntu.com/ubuntu/dev ? [04:04] git://kernel.ubuntu.com/ubuntu/ubuntu-natty.git [04:04] ^ sacarlson [04:04] ok good enuf [04:04] jk-: thanks. [04:14] jk-: CarlFK: that was perfect just what I was looking for thanks you guys [04:15] cool. sorry my help was confusing === ericm|ubuntu is now known as ericm-lunch [04:43] bjf[afk]: hmm, 2.6.35.9, presumably? [04:44] ^ RE: [Applied] Maverick update to 2.6.35.0 upstream stable release [04:46] crimsun, heh, yup === ericm-lunch is now known as ericm === ericm is now known as ericm|ubuntu === _LibertyZero is now known as LibertyZero === jj-afk is now known as jjohansen === smb` is now known as smb [08:49] smb: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/684875 [08:49] Launchpad bug 684875 in linux (Ubuntu) (and 1 other project) "Patch to Natty 2.6.37-virtual breaks non-EC2 users (affects: 1) (heat: 14)" [High,Confirmed] [10:05] Is there any guide, wiki page for kernel bug triage? https://wiki.ubuntu.com/Kernel/BugTriage/BugDay [10:15] hi apw , do you know when 37rc5 will be available in mainline [12:06] morning [12:07] can kernel maintainers give me opinion on bug 682681? [12:07] Launchpad bug 682681 in linux (Ubuntu) "allow to build linux-source package only (affects: 1) (heat: 10)" [Undecided,In progress] https://launchpad.net/bugs/682681 [12:09] hrw: which arm system do you use? [12:10] Kano: now only pandaboard, have also beagleboard b7/c3 and several Kano: why you ask? [12:12] well tegra 2 would be interesting [12:13] there is work going on in the #ac100 channel for porting it to at least .36 [12:14] and google has a branch too for chromeos on tegra2 [12:14] someone would have to make either of these upstreamable ;) [12:14] i am more interested in xbmc on a small system [12:15] then i'd rather suggest a pandaboard [12:15] but of course a webbrowser + flash would be cool [12:15] it has all codecs and 3D drivers available [12:15] panda also has unstable memory and slow usb bus [12:16] you get 1GB chip where <700MB works [12:16] tegra 2 is a pain because nvidia pulled all important functions into a binary daemon [12:16] and those do not work? [12:17] and the 3D driver only works with certain kernels (same goes for the X server, only for certain X releases) [12:17] i thought it had full openmax accelleration and there must be a xbmc branch for that [12:17] panda comes with full oüpenmax support by default in ubuntu [12:17] and is way way cheaper [12:18] did you try the openmax xbmc branch? [12:18] beyond that tegra2 doesnt have any NEON support [12:18] not yet, but we have a spec for a mediacenter edition on the panda for natty [12:18] so it will be tested [12:20] nice, boxee, a xbmc clone will not use tegra 2 as they wanted first, they now use atom === diwic is now known as diwic_afk [12:20] yep [12:20] wonder what changed their mind? [12:20] which is silly ... they should have taken omap4 from the beginning ;) [12:21] cking, i would guess missing NEON and flash issues [12:21] shame [12:21] ogra_ac: atom+ion gives you codecs and ability to run any OS [12:21] hrw, yeah, and eats about 10x the power [12:21] and needs a fan etc etc [12:22] who want's a noisy box to watch video? [12:22] atom users ;) [12:22] it's 'cos they are used to a whirring sound back in the good old VHS and Betamax days [12:22] hehe [12:22] Atom CE4100 is no ion [12:23] tomboy arghhhhh... looks like I will have to run internal wiki or something... [12:23] * ogra_ac recommends tiddlywiki === jjohansen is now known as jj-afk [12:35] ogra_ac: save does not work in chromium [12:39] ah, bad [12:40] * hrw -> strace tomboy === diwic_afk is now known as diwic [13:54] * cking wishes jimmy wales would stop nagging me every time I use wikipedia [14:11] or his cronies [14:11] their pleas are deafening in their silence [14:11] yep [14:24] JFo, where is your Ubuntu Kernel Hot Bugs by Team script located [14:25] pgraner, here: http://reports.qa.ubuntu.com/reports/jfo/kernel-buglist-by-team.html [14:25] oh the script [14:25] one sec [14:26] here: https://code.edge.launchpad.net/canonical-qa-tracking [14:26] pgraner, ^ [14:26] JFo, bzr? [14:26] yessir [14:26] it is in the misc-scripts directory [14:26] * pgraner felt a little piece of him die with that statement [14:26] but I plan to move all of that over to our kteam-tools git branch [14:27] heh [14:28] JFo, ok so you get manually generate the bug file list and feed it to the script? [14:28] yessir [14:28] currently [14:30] JFo, ok [14:36] JFo: hi! so I just upgraded to the latest natty kernel and things are slightly strange. for one, music playback over the network (daap) is choppy and imap in evolution is slow. have you seen any network performance issues with 2.6.37-8.21? [14:37] JFo: in terms of bugs [14:37] jdstrand, I have not noticed any, but that doesn't mean there isn't any [14:37] let me have a look at the bugs [14:37] JFo: let me move one that I just filed with evolution over [14:38] JFo: rebooting into the maverick kernel 'fixed' it [14:38] excellent! thank you :) [14:38] JFo: bug #686584 [14:38] Launchpad bug 686584 in linux (Ubuntu) "IMAP message list regeneration is slow after deleting an email (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/686584 [14:39] JFo: now, I had initially ubuntu-bug'd evolution, so it doesn't have the kernel bits attached. shall I do that? [14:39] jdstrand, yes, please :) [14:39] and thank you [14:47] JFo: ok, I apport-collected the bug and adjust the title and description to be more 'kernely' [14:47] s/adjust/adjusted/ [14:48] jdstrand, thank you very much [14:49] sure thing === diwic is now known as diwic_afk [15:24] JFo: just so you don't waste time-- I invalidated that bug. I don't know what went wrong and need to investigate more [15:24] JFo: sorry for the noise [15:24] jdstrand, no worries [15:24] thanks for letting me know [15:24] * jdstrand nods [15:25] I'd be interested still in what you find === bjf[afk] is now known as bjf === xfaf is now known as zul [16:06] bjf, wanna do the 1 hour kt meeting warning? [16:06] ## [16:06] ## Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting [16:06] ## agenda: https://wiki.ubuntu.com/KernelTeam/Meeting [16:06] ## [16:16] anyone have any recommendations on which commits to start reverting from https://lists.ubuntu.com/archives/natty-changes/2010-November/001282.html that might make grub stop loading after a reboot (but works on a cold boot)? 2.6.37-2.10 and earlier work fine, everything later hangs at the grub loading screen after rebooting on my i386 netbook. mainline kernels all work fine [16:20] Sarvatt, its been busted since -2.10 ? thats a lot of kernels. [16:21] the i386 specific patches seem like a good candidate for revert [16:21] yeah, haven't been able to warm boot that thing in a long time and finally got around to looking at it [16:22] * Sarvatt nods [16:22] likely anything with nx in the commit log [16:22] fortunately, it looks like -2.10 to -3.11 wasn't an upstream rebase [16:25] bjf, BTW, I think I fixed that fwts hang, it's in the PPA for testing. [16:25] I manually edited the efi setup variables on this aspire one to enable NX since it didn't have the option in bios so might be my fault :) I just saw people also mentioning grub hanging early like I'm getting on https://bugs.launchpad.net/ubuntu/+source/linux/+bug/683775 [16:25] cking, cool [16:25] Launchpad bug 683775 in linux (Ubuntu) (and 1 other project) "Natty Alpha 1, i915 has blank screen after boot (affects: 12) (dups: 5) (heat: 92)" [High,Fix released] [16:26] Sarvatt, you could try bisecting. there aren't that many patches between Ubuntu-2.6.37-2.10..Ubuntu-2.6.37-3.11 [16:36] doing that now, will file a bug on it [16:36] thanks for the help [16:39] ## [16:39] ## Kernel team meeting in 20 minutes [16:39] ## === BenC__ is now known as BenC === jj-afk is now known as jjohansen [16:53] ## [16:53] ## Kernel team meeting in 5 minutes [16:53] ## === hrw is now known as hrw|gone [17:06] apw, about? [17:10] bjf, am indeed now [17:15] * JFo goes for some lunch [17:48] this is really interesting, probably required reading for any network geek https://gettys.wordpress.com/2010/12/06/whose-house-is-of-glasse-must-not-throw-stones-at-another/ === sconklin is now known as sconklin-lunch [18:17] Someone please tell me an automated script misfired and I don't have to retest Bug #673504 yet again. [18:17] Launchpad bug 673504 in linux-ti-omap4 (Ubuntu Maverick) (and 1 other project) "Pandaboard chooses a new IP address on each boot (affects: 2) (heat: 20)" [High,Fix committed] https://launchpad.net/bugs/673504 [18:19] This fix has been in proposed since Nov 17. I tested it on Nov 23. It should have been released to maverick-updates. [18:21] tgardner: yep you were right, ed6f363a412661f45a1db9c3456db2f9d5057612 i386: NX emulation is the commit that broke warm boots on this netbook [18:22] Sarvatt, hmm, kees might be interested in this [18:23] updating the machine now and i'll file a bug on it [18:23] * kees scratches his head [18:23] GrueMaster, it fired because there is an open task (the main one which is normally meant for development) [18:24] if it isn't needed in Natty, I'd say you are ok [18:24] it hangs on "GRUB loading..." unless I cold boot, kernel selection menu doesn't even pop up [18:24] This is a separate kernel. [18:24] Sarvatt: please subscribe me; NX emu hasn't changed in a few releases. [18:24] I have already tested this in Maverick (and yes, it is needed in Natty). [18:25] looks like pitti wants verification for the one he built [18:25] per the last comment [18:25] Natty is currently running 2.6.35-903.17 because 2.6.35-903.19 (with the fix) has been rotting in proposed. [18:25] hmmm [18:25] IT HAS BEEN TESTED! [18:25] I'd defer to pitti on the what, where and why [18:25] as I am not in on his thought process [18:26] :) [18:26] JFo: Not trying to yell at you directly. Just pissed at the overall inefficiency of the release process. [18:26] Sarvatt: which kernel are you talking about for ed6f363a412661f45a1db9c3456db2f9d5057612 ? [18:27] I had to retest a couple of networking fixes in Lucid on other platforms 5 times before they were released. Each retest was because they sat in -proposed until a security patch clobbered them before they were mainlined. [18:27] kees: 2.6.37-2.10 is fine, everything after 2.6.37-3.11 hangs on a warm boot and it doesn't happen in the mainline kernels, I bisected between those two kernels and that was the bad commit [18:28] Sarvatt: and this netbook can warmboot maverick's kernel? [18:28] Sarvatt: also, I see that as a5f82c0456fa4f0bf55779c055006dfc18314615 not ed6f363a412661f45a1db9c3456db2f9d5057612 [18:28] yeah maverick's kernel is fine [18:28] GrueMaster, I understand :) [18:29] http://paste.ubuntu.com/540721/ is the bisect log [18:29] so strange. that whole patch just pokes at userspace limits. [18:29] <-errand bbiab [18:32] Sarvatt: ah, and you started with Ubuntu-2.6.37-3.11 [18:32] and due to rebasing the commit SHAs are different. got it. [18:32] GrueMaster, I've fixed the tag on that bug and added a comment, that's all that needed to be done. [18:33] Thanks, bjf. [18:35] Home: https://wiki.ubuntu.com/Kernel/ || Maverick Kernel Version: 2.6.35 || Ubuntu Kernel Team Meeting - December-14 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! === bjf changed the topic of #ubuntu-kernel to: Home: https://wiki.ubuntu.com/Kernel/ || Maverick Kernel Version: 2.6.35 || Ubuntu Kernel Team Meeting - December-14 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! [18:35] bjf, methinks pitti's update scripts have run amok [18:36] Great. All we need is more mok's in the system. :P [18:37] tgardner, thanks for the heads up, we'll get it straightened out [18:37] While I understand that these scripts are designed to improve the kernel team's productivity, they can greatly increase the workload on others. Maybe a review is in order? [18:37] sconklin-lunch, ^ [18:38] GrueMaster, there are many scripts being run against LP not just kernel scripts [18:38] GrueMaster, these in particular belong to pitti and the SRU team [18:39] ok. Didn't know what team he was on. [18:41] But I have been hit by other kernel team scripts in the past. [18:41] bjf: does the linux (not the ti-omap4) changelog reference 673504 for some reason? And why does the ti-omap4 changelog reference 673509? [18:41] GrueMaster, noted [18:42] (the reason I ask is because the changelog's what the SRU team's scripts look at) [18:45] the reason I ask is because, when pitti peruses http://people.canonical.com/~ubuntu-archive/pending-sru.html to see what's been verified and is safe to move from -proposed to -updates, unless 673509 gets verified, he'll hold off on promoting the ti-omap4 kernel. [18:45] sbeattie, i understand [18:46] when it seems that the only issue that affects 2.6.35-903.19 has been verified already, and it's sat for longer than 20 days, and so should probably be approved. [18:46] err, s/20 days/7 days/ [18:51] sbeattie, it referenced those two bugs because the patch against the ti-omap4 branch referenced them as BugLinks [18:52] bjf: anyway, I'd approach pitti and point it out to him that the ti-omap4 kernel in -proposed can be promoted. === sconklin-lunch is now known as sconklin [18:55] Sarvatt: what hardware is the warmboot failure happening on? (or is that already in the bug report?) [18:58] kees: it's an acer aspire one AOA150, just finished upgrading the 300something packages to file the bug [18:59] it's a little special because the NX enable/disable options aren't exposed in the bios but I enabled the option manually via efi setup vars, can't disable it again without reflashing [19:00] but I see people reporting a similar problem that might be related on other bugs (hangs at the GRUB loading screen in natty) so might not be isolated to this one [19:01] Sarvatt: but you isolated this to the NX-emu code. It hasn't changed. :( [19:03] Sarvatt: if you're booting 32bit -3.11, you shouldn't have my forced NX on fixes, so we should only be dealing with the nx-emu patch [19:03] Sarvatt: then the variables are the NX bit itself, really. [19:04] Sarvatt: are you booting -generic or -generic-pae ? [19:12] * tgardner --> lunch [19:14] kees: -generic only, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/686705 [19:14] Launchpad bug 686705 in linux (Ubuntu) "System hangs at GRUB loading screen every warm boot since 2.6.37-3.11 (affects: 1) (heat: 8)" [Undecided,New] [19:17] entirely wrong menu option selected :) [19:19] JFo: Pie wasn't on the menu. :P [19:19] unfortunately no :-( [19:20] and I looked everywhere [19:20] :) [19:42] kees: would seeing if it happens on generic-pae be a useful data point? [19:43] Sarvatt: for -3.11, it would, yes. [19:55] kees: 2.6.37-3.11 generic-pae works [19:55] so does 2.6.37-8-generic-pae [19:58] Sarvatt: okay, well, that supports the idea that the NX-emu patch is bad. I still don't understand how it's different from maverick yet. [19:59] Sarvatt: I added another comment to the bug; can you verify the "nx-test" behavior for -3.11 -generic and -generic-pae ? [19:59] sure thing, on it now [19:59] Sarvatt: I'm suspecting the fastsyscall use, but again, that didn't change from maverick, and I don't see how that would stop a warmboot. *scratch head* [20:18] how do i compile just a single module (i915) for my current kernel from a patched ubuntu kernel tree? [20:18] I've checked out the tag corresponding to my currently running kernel, copied /boot/config-$(uname -r) to .config and run make LOCALVERSION= EXTRAVERSION=-23-generic drivers/gpu/drm/i915/i915.ko [20:19] then I copied that to /lib/modules/$(uname -r)/... and did update-initramfs, but on reboot it refuses to load, complaining "i915: no symbol version for module_layout" [20:51] tgardner: so, I'm at a loss on bug 686705. the only other thing I can think of is to build a kernel with natty's nx-emu reverted and put in maverick's nx-emu back in. the only thing that changed was the use of disable_nx, but I don't see anything wrong with it on visual inspecition between the two patches. [20:51] Launchpad bug 686705 in linux (Ubuntu) "System hangs at GRUB loading screen every warm boot since 2.6.37-3.11 seemingly due to nx-emu patch (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/686705 [20:52] kees, I'll look in a bit. [20:55] * ilmari resorts to building the entire kernel [20:56] eek: temp1: +96.0°C (crit = +100.0°C) [20:56] ah, at 97°C the fan sped up and promptly got it down to 90°C [20:59] also, the CPU got throttled from "turbo mode" to 1.2GHz (it's nominally 2GHz) === bjf is now known as bjf[afk] [21:31] * jjohansen -> lunch [21:36] kees, doesn't seem like there is anything in that NX patch that would be affected by a warm boot. [21:37] tgardner: my thoughts exactly. :( [21:41] kees, its likely some BIOS issue. Sarvatt did say he'd been hacking on it to get NX enabled. I wonder if we could get anything info by booting with earlyprintk=vga ? [21:42] anything-->anymore [21:45] tgardner: ironically, he didn't need to do anything to get hw NX enabled since my NX unmasking patch is in the later natty kernels. [21:45] (and other people with similar ASUSes reported success with it) [21:46] kees, huh. === tgardner is now known as tgardner-afk [21:48] I was going to live with that but ran across comments like this one in another bug making me think it might be more widespread so I put what I found out there https://bugs.launchpad.net/ubuntu/+source/linux/+bug/683775/comments/21 (this person has the opposite experience that I do though) [21:48] Launchpad bug 683775 in linux (Ubuntu) (and 1 other project) "Natty Alpha 1, i915 has blank screen after boot (affects: 12) (dups: 5) (heat: 92)" [High,Fix released] [22:02] kees, do we 'undo' the nx-emu protection? restore the segments before calling reboot? perhaps they are surviving the warmboot on this system ... we do use a variety of reboot methods on different h/w for warmboot [22:03] apw: there's nothing really to undo, iiuc -- it's just the cs limit. but no, there's nothing that undoes it before reboot. [22:03] kees, and the grub behaviour is different on natty as it is trying to go into graphics mode so we may need to call an option rom we do not on maverick [22:04] oh... well that's interesting. [22:04] would cs limit affect that? [22:04] kees, so if the warmboot on this machine doesn't restore them it may leave us with much of memory not executable in ring0 or somethign right? [22:05] apw: well, dunno about ring0, I thought the cs limit would just trigger in ring3. [22:05] apw: where is the common "rebooting now" function? we could add something there maybe... [22:07] kees, may be worth finding out if Sarvatt is fixed by using gfxmode=text (or whatever it is) as that reverts us more to the maverick boot sequence [22:07] (thats a grub option) [22:07] I will add that to the bug, excellent. [22:08] hm, not sure it's gfxmode=text, still looking [22:09] kees, its set gfxpayload=text in grub [22:09] oh, not set linux_gfx_mode=text [22:09] that variable ends up in the otehr onem, so i think either [22:10] okay, will add [22:10] kees, he'll prolly need to do that in the /etc/grub/default somehow [22:11] kees, likely he will know how, its a common debug technique right now [22:11] * apw wanders off to watch some DS9 [22:12] this is with gfxpayload=text [22:13] Sarvatt: ah, where do you use that? all I could find was /etc/default/grub GRUB_TERMINAL=console [22:14] GRUB_GFXPAYLOAD_LINUX=text [22:14] it's in my grub.cfg with that in /etc/default/grub [22:15] okay. well... dang. that rules that out. [22:36] if anyone can point me in the right direction regarding how to build individual patched drivers for a specific ubuntu kernel I'd love to write up a wiki page about it [22:37] even just rm debian/stamps/stamp-build-generic && fakeroot debian/rules binary-generic takes far too long when only one driver is changed [22:37] ignoring the initial build which takes yonks and gigabytes