[00:31] Sarvatt: awesome, thanks [00:32] Sarvatt: ATI will show 'drm'? [00:32] ati/intel have drmfb's nouveau only has KMS [00:32] ah yes - [drm] radeon defaulting to kernel modesetting. [00:32] its looking for nouveaufb or drmfb [00:33] Sarvatt: sorry, confused. again from the beginning. [00:34] /proc/fb contains a list of frame buffer devices, with the frame buffer device number and the driver that controls it [00:34] yeah if nouveaufb or any of the drmfb's are in there it's using KMS [00:35] Sarvatt: and just to be clear those drmfb's are inteldrmfb, nouveaufb, radeonfb [00:35] nouveaufb radeondrmfb or inteldrmfb [00:35] cool, thanks Sarvatt [00:40] no problem, also I forgot but vmwgfx is svgadrmfb as well [00:47] Sarvatt: which are? [01:45] Can someone please mentor me with l-b-m? I want to add an updated wacom module fix problems with new wacom bamboo tablets. I've already discussed this and submitted a proposal on the mailing list, but nobody has responded. See bugs #568064 and #606278 [01:45] Launchpad bug 568064 in linux (Ubuntu) "Wacom new bamboo models are not recognized in Lucid (affects: 19) (heat: 101)" [Undecided,Confirmed] https://launchpad.net/bugs/568064 [01:45] Launchpad bug 606278 in linux-backports-modules-2.6.35 (Ubuntu) "Add linux-backports-modules-input (affects: 6) (dups: 1) (heat: 42)" [Undecided,New] https://launchpad.net/bugs/606278 [01:46] Since I can see any sign of anybody working on this, I wanted to know if somebody can help me write a patch instead. I've taken a look at the l-b-m source, but it's so complicated I can't really make much sense of it. [01:46] *don't see any sign [01:49] It might be nice to have CONFIG_COMPACTION on. [01:51] I'm not some noob with packaging, I was just hoping somebody give me some hints or guidance on how to start adding a new component to l-b-m [02:41] ripps: does l-b-m make sense for something thats not a backport? what's wrong with your dkms package? [02:42] Sarvatt: nobody would accept it. MOTU's said to try and get it into Debian, and Debian said that it was just working around a problem with the kernel. Problem is, the linuxwacom project isn't updating their wacom kernel source because apparently their refactoring it or something. [02:44] After discussing it here, I was advised to propose an addition to linux-backport-modules. But apparently, nobody wants to help me get in there, either. [07:21] any known regressions in the current lucid kernel? I [07:21] hrm [07:22] I'm unable to do a full install (~2500 packages) anymore, since dpkg sees some packages are corrupt (same size, different md5sum as on the archive) [07:23] or when unpacking some config stubs in /tmp are missing etc [07:23] maybe I should try with ext3 to rule it out [07:30] ah, actually.. the installer image is the release one :) [07:30] the one that worked was from March [07:34] though the corruptness happens on an installed system as well, and that has the latest kernel [08:47] /usr/share/doc/debian-policy/upgrading-checklist.txt.gz [08:59] hi ogasawara, is anything being done to implement this? https://bugs.launchpad.net/ubuntu/+source/ndiswrapper/+bug/590090/comments/7 [08:59] Ubuntu bug 590090 in ndiswrapper (Ubuntu) (and 1 other project) "package ndiswrapper-dkms 1.56-1 failed to install/upgrade: ndiswrapper kernel module failed to build (affects: 40) (dups: 3) (heat: 206)" [Medium,Confirmed] [10:32] if ! git remote | grep FOO >/dev/null; then echo NO; fi === amitk is now known as amitk-afk === amitk-afk is now known as amitk === yofel_ is now known as yofel [12:33] can anyone suggest things to try and get useful debug information? Got a user who's USB controller seems to hang ~once a day (leaving him without working keyboard and mouse). Nothing interesting in kernel/X logs, and lsusb hangs on open("/dev/bus/usb/002/005") [12:34] nm that, after a while, starteg tetting task blocked messages in kernel logs === BenC____ is now known as BenC [13:29] hi, is there any chance that https://bugzilla.kernel.org/show_bug.cgi?id=16315 will be backported to 10.04 LTS? [13:29] bugzilla.kernel.org bug 16315 in i386 "icebp (opcode 0xf1) no longer causing a SIGTRAP, breaks Wine" [Normal,New] [13:30] Ejdesgaard_, If the patch goes upstream as a stable patch it will come back as soon as upstream stable picks it up [15:07] lag, the picture is up on my Facebook === jjohansen is now known as jj-afk [15:09] sent to you in e-mail === bjf[afk] is now known as bjf === sconklin-gone is now known as sconklin [15:30] JFo: That's awesome! [15:30] :) [15:40] JFo: http://people.canonical.com/~ljones/lastsupper/ [15:47] nice! [15:48] not a bad approximation [15:55] It's there or there abouts [15:55] Someone is facing completely the wrong way and I'm a few degrees off [15:55] Overall I think we did okay :) [15:56] lol! [15:57] you guys werent kidding around :D [15:57] smb, it was included in .35-rc4 [15:58] ## [15:58] ## Kernel team meeting in two hours [15:58] ## [16:14] era: it seems that bug was recently resolved in the last few hours by the package maintainer [16:18] era: I'm not opposed to going with their plan B in the future and completely dropping ndiswrapper from the kernel but that's a discussion that needs to probably take place on the kernel-team mailing list [16:31] jcrigby: has the binary package names in the linux-linaro upload been OKed by the kernel team? [16:32] Riddell, its been uploaded, but is awaiting acceptance by an archive admin [16:32] Riddell: I don't know. What constitutes approval. [16:32] tgardner: why do you think I'm asking :) [16:33] jcrigby: tgardner's blessing? :) [16:33] Riddell, random curiosity? [16:33] Riddell, slangasek said he had some concerns, but he's at debconf [16:34] why do some of the binary packages have "linaro" in the name and some go for the -1000 suffix? [16:34] Riddell, um, not supposed to. I guess I'd better look closer. I only looked at the control file [16:35] I tried to follow the abstracted debian model. May have messed up. [16:35] well it looks deliberate I presume there's a reason why that's the best way to do it [16:35] tgardner: I removed a bunch of entries from the control file yesterday. It now looks like the ti-omap4 control file. [16:35] e.g. linux-headers-2.6.35-1000 [16:35] Riddell, whats the URL for looking at the generated packagesd? [16:36] tgardner: there are no generated packages, I'm reviewing the source [16:38] I think you will see the same convention in the ubuntu kernel [16:38] without the 1000 [16:38] Riddell, according to debian/control I think the package names are correct. it follows the same model as other kernel topic branches [16:38] yeah that's the ABI number [16:39] but it seems curious to namespace some packages with "linaro" and some with [16:39] Ejdesgaard_, Is that definitely needed for 2.6.32 (10.04 LTS) as the patch seems to have been sent for 2.6.33.y specifically [16:40] Riddell, oh, you mean the 'linaro' part? I guess he's using it to distinguish these kernel binaries from the omap4 packages that are already in the archive. [16:41] tgardner: yeah that makes sense. bumping the ABI to 1000 is what seems strange to me [16:42] jcrigby, so, do you want to respin with new binary package names? [16:42] Riddell, we use the ABI number to distinguish between kernel topic branches [16:42] otherwise we'd have binary package name collision [16:42] tgardner:would be glad to if someone tells me what to change them to [16:42] I'd have thought e.g. linux-linaro-headers-2.6.35-1 would make more sense [16:43] Riddell, its a restriction of the installer, which insists on the linux-header* pattern [16:43] ahh [16:43] similarly for linux-image* [16:43] tgardner: ah, I knew there'd be a perfectly rational explanation [16:43] well, rational is a stretch :) [16:45] on a different matter, I changed the control yesterday and removed some stuff and made Architecture: armel for everything [16:45] so the build system only tries to build on armel [16:46] jcrigby: in that case should I reject the current upload and save the non-arm buildds the hassle of building something they don't want? [16:46] jcrigby, I'm not sure you can do that. Some of the packages have to be constructed on the default arch (i386), e.g., the headers _all.deb [16:47] tgardner:I made it look like the ti-omap4 control [16:47] is that known good? [16:47] Riddell, why don't you go ahead and reject this upload. I [16:47] I'll work with jcrigby to fix the control file issues [16:49] tgardner:thanks [16:49] ok rejected, ping me when you reupload if you want a quick review [16:50] Riddell, later today I think. Thanks for your help. [16:51] tgardner:what is best for you. Do you want to just send me a patch? [16:52] jcrigby, send a pull request with your current bits [16:52] with the ti-omap4 changes? [16:52] style changes that is [16:52] jcrigby, yep [16:53] ok, I notice one thing that your comment above sheds some light on: [16:53] Package: linux-headers-2.6.35-1000 [16:53] Architecture: armel [16:53] Section: devel [16:53] Priority: optional [16:53] Depends: ${misc:Depends}, coreutils | fileutils (>= 4.0) [16:53] #Provides: linux-linaro-headers, linux-linaro-headers-2.6 [16:53] Description: Header files related to Linux kernel version 2.6.35 [16:53] This package provides kernel header files for version 2.6.35, for sites [16:53] that want the latest kernel headers. Please read [16:53] /usr/share/doc/linux-linaro-headers-2.6.35-1000/debian.README.gz for details [16:54] explains the commented out Provides: [16:55] jcrigby, correct [17:07] bjf, is there a meeting today? [17:07] apw, yes [17:07] ahh i see your 2 hour warning now [17:14] sconklin, bjf Seems there is no need to start a new release on Lucid, that was already done [17:15] smb: yes, I'm just applying the patches. I'll start with the things we wanted on the short list [17:15] sconklin, ok [17:17] sconklin, We should sync on the writeback patches. I got the feeling there is a patchset you got and one I have done [17:18] smb let me check what I have and see if they're in my public repo [17:18] bug 643617 [17:18] smb: Error: Bug #643617 not found. [17:18] hm [17:18] bug 543617 [17:18] Launchpad bug 543617 in linux (Fedora) (and 3 other projects) "Unmount of an fs with dirty cache buffers causes pathological slowdown (affects: 11) (dups: 2) (heat: 101)" [Unknown,Unknown] https://launchpad.net/bugs/543617 [17:19] same as bug 585092 [17:19] Launchpad bug 585092 in linux (Ubuntu) "giant IO delays (affects: 1) (heat: 22)" [Medium,In progress] https://launchpad.net/bugs/585092 [17:20] smb: yes, there's already a branch "bug bug543617" in my public repo that I used to build the test kernel for that bug [17:21] oops, branch name is simply bug543617 [17:21] sconklin, Let me check what is there. Cause I have been working on that and probably got a different set [17:25] sconklin, My set is there (http://kernel.ubuntu.com/git?p=smb/linux-2.6.32.y.git;a=shortlog;h=refs/heads/writeback) and looks like being a superset of yours. I am currently trying to get Jens Axboe look at those to get them backported to stable upstream. There is a set for 2.6.34.y as well [17:25] sconklin, Have you sent your patches for review already. I cannot remember having seen that [17:26] as I recall, they were sent for review weeks ago, and I just picked them up. Let me check [17:27] smb: [1/2] writeback: fix WB_SYNC_NONE writeback from umount [17:27] I was holding back because I at least wanted to have some feedback on them before. These series has been taking a lot of time to mature upstream [17:28] [2/2] writeback: ensure that WB_SYNC_NONE writeback with sb pinned is sync [17:28] Ok, maybe my bad not deleting those [17:28] smb: OK, I just grabbed them and built a test while you were out sick because rtg asked about them === jj-afk is now known as jjohansen [17:28] sconklin, Please don't pick them for applying now [17:29] smb: they were still active in patchwork also [17:29] ok, I won't pick them [17:29] Right, I should have marked them as superseeded [17:30] Thanks. Unfortunately there was a time in between when I knew the ones we had were not good but upstream was not fixed yet. And not to forget I did not set them as rejected [17:31] ## [17:31] ## Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting [17:31] ## agenda: https://wiki.ubuntu.com/KernelTeam/Meeting [17:31] ## [17:32] sconklin, The set now seems to be ok, but as you can see in that tree its a huge set. [17:33] sconklin, The only thing we need to remember is to revert the workaround when we apply the writeback set [17:33] smb: I'll apply the graphics fixes for the two issues we have patches for, then the writeback patches, then the stable update, OK? [17:35] sconklin, Leave out the writeback patches completely. I want to get feedback on the backport first and then do another round of sending those to the kt-mailing list [17:35] smb: ok [17:36] sconklin: i assume you need that sru sooner than later for the x201 display bug? [17:36] vanhoof: yes, I'd hate for it to end up being the gating factor [17:36] manjo: ^^ :D [17:37] yes building kernel with patch now [17:37] I will send it to SRU as soon as I am done [17:37] manjo: ok, i can re-test if you'd like [17:38] yep that will be great! [17:38] vanhoof, manjo: also, people have been ignoring the SRU requests or not completing them properly. We've been fixing them up without complaining, but I'm going to start bouncing them until they are correct. [17:39] sconklin, I removed the writeback parts from patchwork for now so there won't be confusion on that part [17:40] smb: ok, but I think I acked them on the mailing list, so you'll want to follow up on that [17:41] sconklin, There is another one to disable CONFIG_LGUEST for hardy lpia which went in as part of the security update to fix the build failure [17:41] sconklin, I probably should post a message there, too [18:29] Is there an issue with tangerine ? [18:29] manjo@tangerine:~$ dchroot -c lucid-amd64 [18:29] W: Group ‘sbuild’ not found [18:29] E: Access not authorised [18:29] I: You do not have permission to access the schroot service. [18:29] I: This failure will be reported. [18:29] manjo@tangerine:~$ [18:29] I was able to dchroot just couple of hrs ago [18:29] <-starving... headed to lunch [18:31] tgardner, ^^ any idea ? [18:31] manjo, I guess you're not special. [18:31] ? [18:32] tgardner, i think it's because he *is* "special" [18:32] manjo, just kidding. lemme check [18:32] bjf, thanks [18:33] apw, smbgot some oops on tangerine. [18:34] * manjo brb [18:34] smb, ^^ need to reboot. [18:34] Hm, just for compiling six kernels in paralell [18:34] ok [18:34] tgardner, stuck in writeback ... smb could that be your writeback issue you had patches for ? [18:34] shall we try the 2.6.35 kernel? [18:34] ogasawara, you want me to leave "delta-review" on the agenda? [18:34] tgardner, apw Let me have a quick look [18:35] bjf: I'd say take it off === bjf changed the topic of #ubuntu-kernel to: Home: https://wiki.ubuntu.com/Kernel/ || Maverick Kernel Version: 2.6.35 || Ubuntu Kernel Team Meeting - August-10 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! [18:36] bjf: the last big item is to review if any of our ubuntu drivers need updating (which I suspect they won't) [18:36] ogasawara, it's gone [18:36] smb: do we put the buglink for the stable update bug in every patch that's part of a stable pile? [18:37] sconklin, yes. maint-modify-patch [18:37] ok, doing that now [18:37] tgardner, apw Seems the process that was stuck first was returning later. So it might well be the slow writeback problem [18:38] apw, you still have two agenda items do you want to keep them? [18:38] smb, lets try the LTS backport kernel for awhile. [18:38] I want to triage bugs for kernel at first. So I'm going to read documents of Ubuntu Kernel Bug triaging .. [18:39] bjf, the misc one is likely redundant [18:39] tgardner, as long as we can switch back to the lts kernel later, ok. Cause that machine is good for running stress io tests [18:40] tgardner, I am off it again, so you can reboot if you like [18:43] * smb is gone [18:45] tgardner, please let me know when tangerine is back [18:50] simar, sounds great for us ... [18:51] simar, talk to JFo he should be able to help you as well [18:52] simar, he leads the bug triaging effort [18:56] tgardner, looks like tangerine is back... can I use it ? [18:57] ah same issue cannot dchroot ... [18:57] * manjo getting lunch will be back soon [18:58] manjo, should be OK now. you'll need to logout first. === bjf is now known as bjf[food] [20:33] * ogasawara lunch [20:33] * jjohansen lunch too === bjf[food] is now known as bjf [22:06] manjo, vanhoof: The graphics patches you wanted are now in the lucid repo [22:06] * manjo sconklin beer++ [22:15] ogasawara, sconklin just pushed some patches to the lucid repo is there anything that will automatically build that (some kind of pre-proposed / stable crack of the day)? [22:15] bjf: hrm, I can't remember if smb's pre-proposed in the kernel-ppa builds automatically [22:16] ogasawara, bjfAFAIK it does [22:16] bjf: it certainly should be easy enough to make it do so automatically as apw does the same for the latest maverick tip [22:16] ogasawara, that's what i thought, was just wondering if we already did it [22:17] tgardner, i guess we'll know in a bit then [22:19] latest maverick iso i386 from cdimage current does not seem to boot ... [22:19] I recall apw posted something in this regard wrt to usb keys .. anyone recall ? [22:21] manjo: bug 608382 ? [22:21] sconklin: awesome! [22:21] Launchpad bug 608382 in usb-creator (Ubuntu) (and 1 other project) "USB images of Maverick CDs fail to boot with -- Error: Unkown keyword in configuration file (affects: 13) (dups: 1) (heat: 70)" [High,New] https://launchpad.net/bugs/608382 [22:21] looking [22:23] ogasawara, thanks a ton [22:24] bjf, kernel-ppa@zinc:~$ crontab -l [22:24] # m h dom mon dow command [22:24] 0 * * * * $HOME/buildscripts/mainline-build/kernel-version-map >$HOME/public_html/info/kernel-version-map.html.new && mv $HOME/public_html/info/kernel-version-map.html.new $HOME/public_html/info/kernel-version-map.html [22:24] 0 09 * * * USER=kernel-ppa $HOME/kteam-tools/mainline-build/mainline-trigger >>$HOME/logs/mainline-trigger 2>&1 [22:24] 5 * * * * USER=kernel-ppa $HOME/kteam-tools/mainline-build/cod-execute >>$HOME/logs/cod-execute 2>&1 [22:24] @hourly /srv/kernel.ubuntu.com/www/scripts/gitfind > /dev/null 2>&1 [22:25] ogasawara, if you are channel operator can you add "10.10 USB boot issues see https://launchpad.net/bugs/608382 for workaround" ? [22:25] Ubuntu bug 608382 in usb-creator (Ubuntu) (and 1 other project) "USB images of Maverick CDs fail to boot with -- Error: Unkown keyword in configuration file (affects: 13) (dups: 1) (heat: 70)" [High,New] === ogasawara changed the topic of #ubuntu-kernel to: Home: https://wiki.ubuntu.com/Kernel/ || Maverick Kernel Version: 2.6.35 || Ubuntu Kernel Team Meeting - August-10 - 17:00 UTC || If you have a question just ask, and do wait around for an answer! || 10.10 USB boot issues see https://launchpad.net/bugs/608382 for workaround === sconklin is now known as sconklin-gone [23:23] ogasawara, bjf, sconklin-gone, the pre-proposed engine builds all of the releases, assuming they have non-ignored commits following a release [23:24] apw, yes, i was looking at the scripts and sort of decided that [23:24] apw, if i understand it, they should kick off 0900 UTC [23:25] bjf yes indeedy [23:25] bjf with the new queuing ability we could reschedule the pre-proposed to a better time independantly if needed [23:26] night [23:26] g'night === bjf is now known as bjf[afk]