[07:37] Hi, how do I could debug a kernel panic, possibly caused by a scheduler [07:38] I have ~daily kernel panics, and I see line of "BUG: Error while scheduling atomic: " sort [07:43] this is 2.6.38-11-generic running on Intel, I suppose this CPU: http://ark.intel.com/products/41498/Intel-Celeron-Processor-900-%281M-Cache-2_20-GHz-800-MHz-FSB%29 [07:53] * apw yawns, bouncy internet this morning, not good [07:54] heh, struggled 2 first hours in the office for constantly disconnecting wlan [07:54] s/for/with/ [07:54] hey apw [07:54] morning [07:54] been keeping the rioters at bay? [07:55] jk-, two nights ago they were burning dowm the shops 5 mins from my house, but its sounding quieter now [07:55] jeeeez :/ [07:56] jk-, first and only time i've packed a bag in case i had to run [07:56] and i lived though the constant bombing by the IRA, somehow this is more dangerous, more random [07:59] but i heard nothing last night so perhaps we are past the worst [08:01] i hope so... [08:13] we have had a pile of police on the streets, thats going to cost a few bob sadly [08:13] "peaceful night across the country" is todays message ... phew [08:16] "those who think they have got away with looting will be supprised today when the doors start to go in" like the sound of that === yofel_ is now known as yofel [13:11] sconklin: bug 824520 during QA for 2.6.38-11.48, generic-pae. Did not reoccur, so I am not marking as failure. Nevertheless you folks should have a look [13:11] Launchpad bug 824520 in linux "WARNING: at /build/buildd/linux-2.6.38/arch/x86/kernel/apic/ipi.c :109 default_send_IPI_mask_logical+0xb1/0xf0()" [Undecided,New] https://launchpad.net/bugs/824520 === smb` is now known as smb [13:52] hggdh: all testing results like this need to be posted to the tracking bug, so they aren't lost on IRC. I've copied this one to the bug. I'll have a look and see whether we should pass it or not [13:54] sconklin: I was giving yuou advance warning. When I finish a test sequence, I write in anything interesting, including bugs. No need to tell me this should go into the tracking bug [13:54] hggdh: ok, thanks, sorry to jump on that [13:55] maybe I can have an answer before you are done [13:55] this is my hope. I cannot tie it to the kernel so far [13:59] Hi All, [13:59] First want to report some bugs in ubuntu-natty kernel (for ARM) [14:00] the default driver module for Touchscreen ads786 is not the right one [14:01] And now am trying to build this module , but not succeeded [14:02] Steps followed : Downloaded the kernel source from rep. by apt-get install linux-source [14:02] hggdh: Any idea which test was running when this occurred? Maybe it's in the logs and I don't know where to look [14:03] make modules SUBDIRS=drivers/input/touchscreen/ [14:03] hey everybody [14:04] im new to this... which languages do i need to know to customize my kernel ? [14:04] English & C :) [14:05] so far so good. :P [14:05] :) [14:06] im not so familiar with ubuntu though. like i said im new to linux, but im not bad with C and java [14:07] Then i suggest you to familiar with linux/ubuntu first then get into Linux Kernel [14:08] and whats your advice on how to become familiar with ubuntu fast? [14:08] i feel like its gonna take ages [14:09] any e-book or online texts recommended ? [14:09] not at all [14:09] Download ubuntu ISO isntall it and play with it [14:09] sconklin: at this point in time it was the LPT lite [14:10] it really does not help that a lot of programs get to print warning/error messages without identification :-( [14:11] khanx, get familiar with basic unix/linux commands [14:12] this will make your " Playing with Linux" more easier [14:14] any help pls [14:15] Or any one can provide me ads786.ko (touchscreen module) for 2.6.38.8-omap [14:16] siji, if you have ARM specific questions you may get quicker answers on #ubuntu-arm, the peeps who mostly have the h/w hang out there [14:16] apw, :) [14:16] apw, lol [14:17] apw, i have no clue how to only build modules in subdirs of the kernel tree [14:17] he he [14:17] never had to [14:17] make M= isn't it? [14:17] or make module.ko [14:17] ohsix, means [14:17] there's "make help" too [14:17] so i'm surely a bad candidate to support that :) [14:18] ok [14:19] (ohsix, looking into make help ) [14:19] building only in a subdir, hmmm, that is in the docs somewhere something like fakeroot debian/rules prepare-; make O=debian/build/build- [14:19] or /.ko at the end [14:19] not that i ever do it, i normally just build the lot [14:19] apw, ya it worked for me [14:20] but it hasnt built whole kernel modules [14:20] I have defined he directory name [14:21] ./touchscreen which is having more than 10 driver modules, but it built only three [14:21] And unfortunately my module was not there in those three ;( [14:22] well it will only build the ones which are turned on in your config too [14:22] Am not a expert in kernel , so i feel something wrong in the .conf or some where else [14:22] apw, understood [14:22] I hve tried many times by configuring it [14:22] if you want foo.ko built there will be something like obj-$(CONFIG_FOO): foo.ko in the Makefile, which tells you the CONFIG to look for in your .config [14:23] apw, ok [14:23] let me have a look into it [14:23] apw, well, the initial issue that drove siji to try the recompile was that the module shipped in the package breaqks with undefined symbols on loading [14:24] apw, which i find extremely weird for a module built in the same package [14:25] ogra_, apw , then by building the kernel from the same source(ubuntu repository) again will make any sense for me ? [14:25] ogra_, odd indeed but not impossible i think [14:25] if the dependancies in the Kconfig are broken [14:27] apw, broken locally in my setup only ? [14:27] the kteam-bugs is 831M [14:27] -EWRONGCHAN [14:28] siji, likely generally if they don't work with the ones in the archive [14:28] is there a bug or error messge i can look at [14:28] apw, ok [14:29] the bug is with ads7846 driver module (Touchscreen - used TSC2046 in the Expansion board,and it connect with DM3730 by SPI port) [14:30] ogra_, sorry hasnt got time to report the bug report . apologise .. [14:31] dont apologize, file it instead :P [14:32] :) [14:34] ogasawara, am working on jjohansen's AA 13 patch set. [14:35] ogasawara: ack, I'm lookin at ike's [14:35] cool === ericm|ubuntu is now known as ericm-Zzz === ericm_ is now known as ericm-Zzzz [14:43] morning [14:55] ogra_, done :) [14:56] good learning for me :) [14:57] siji, tell apw the bug number [14:57] o [14:57] ok [14:57] apw, Bug #824589 [14:57] Launchpad bug 824589 in ubuntu "inproper Touch Screen Dirver module(for arm beagleboard) for TSC2046 in the Expansion board,and it connect with DM3730 by SPI port ." [Undecided,New] https://launchpad.net/bugs/824589 [14:58] yes the same [15:17] ogra_, do you have a pointer to that possible arm smb issue [15:18] apw, omap3 has no SMP :) [15:18] ogra_, omap4 [15:19] err, you said smb :P [15:19] Though I would probably prefer not being called an issue... :) [15:19] sorry, running a meeting atm [15:19] ogra_, He was trying to say smp [15:32] jjohansen, about ? i've replied to your TOCTOU patches, i think there is an issue with the Maverick one [15:32] apw, He is here though I need to ask [15:32] apw: oh? that is possible [15:33] jjohansen, and i am wondering if you are doing backports for lucid and hardy or not [15:34] apw: ah yeah, sorry /me couldn't find those last night and hasn't gotten to it this morning yet [15:39] jjohansen, ok will wait on those rather than doing them [15:40] apw: okay, I'll take a look at the Maverick one again too [15:40] jjohansen, my review comments are on the mailing list [15:41] apw: thanks [15:46] smb, apw, meeting done, so what was the issue [15:47] ogra_, Question was whether you got a pointer to the bug concerning poor performance in smp mode on the panda [15:47] the bug number you mean ? [15:47] yep [15:47] i have no further pointers than what are noted in the bug [15:48] one sec [15:48] bug 709245 [15:48] Launchpad bug 709245 in linux-ti-omap4 "ARM SMP scheduler performance bug" [High,Confirmed] https://launchpad.net/bugs/709245 [15:48] ogra_, K, thanks [15:49] Currently upgrading the board here [16:00] pgraner, looks like the ISO builder ran last night? [16:04] bummer, still getting apt-get trap divide error [16:33] does anyone here who is (a) in ubuntu-core-dev and (b) not in techboard or ubuntu-release have a bug task they'd like to nominate for a stable release? [16:33] right now you can just do it; I want to make a change to the stable release ownership in LP and would like to make sure that doing so does not break it this [16:34] but I can't test it myself since I'm overprivileged [16:34] tgardner, ^^ is the only one i think who fits that criteria [16:34] cjwatson, hmm, that might be me. [16:35] if you have a suitable task to nominate, then yeah, you have the level of privileges I'm looking for [16:35] tgardner, you can safely nominate any of our cves to like karmic and the tools will slam them invalid later [16:36] cjwatson, are you wanting to mess with our ability to make nominations? its already difficult in that very few of us _can_ do it. [16:36] s/cves/cve bugs/ [16:36] tgardner: I want to make sure I'm *not* messing with your ability to do so [16:36] cjwatson, ah [16:36] actually, here [16:36] tgardner: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/752694 - can you see an "approve/decline" thing there? (don't click it yet) [16:36] Ubuntu bug 752694 in grub2 "Grub2 installs on the wrong device if in a RAID config" [High,Confirmed] [16:37] cjwatson, yes [16:37] tgardner, http://people.canonical.com/~kernel/reports/kernel-cves.html any of those bugs can be safely nominated to anything, and we'll cope [16:37] tgardner: OK, reload - can you still see it, and if you hit approve, does it work? [16:38] (I've just changed natty's owner from ubuntu-core-dev to ubuntu-release) [16:38] cjwatson, it appears to have worked [16:38] excellent, thanks! [16:39] I'll go and change the other stable releases now then. Please let me know if you notice regressions in your ability to nominate things [16:39] cjwatson, will do. [16:40] apw, since FF approaches, perhaps its time to get an empty oneiric LBM package in the archive before it gets difficult. [16:42] i think ff offically is in 15 mins, but lbm has always been treated at part of the kernel [16:43] but yes i think its time [16:43] (BTW, the fix for bug 451390 ought to have improved the nomination situation for you guys) [16:43] Launchpad bug 451390 in launchpad "limited upload rights no longer give series nomination accept/decline rights" [Critical,Fix released] https://launchpad.net/bugs/451390 [16:44] cjwatson, ohhh nice ... musdt find a way to test that [16:51] tgardner, i'll put the LBM on my todo for tommorrow [16:51] apw, ack [16:52] tgardner, you will be pleased to know i have spent several hours un-fooking my email so i have some hope of keeping up on it [16:52] apw, be sure to check the repo in your morning in case I manage to get to it yet today. [16:52] tgardner, ack [16:53] apw, I'm pretty ruthless about keeping my inbox down to just a couple of emails by EOD [16:54] yep and working on a similar scheme but it needs for my filtering to be sane and work, which has been the big battle [16:54] i think i have just about won [16:54] * tgardner likes procmail [16:59] ogasawara, does '(drop after 3.0)' make sense for a clean cherry-pick ? [16:59] they should drop themselves on rebase anyhow [17:00] tgardner: sorta unneeded I guess as apw says they'll drop anyways [17:00] apw, ogasawara: you mean for the P series kernel, or for the once more that might happen for oneiric ? [17:01] (for P i am assuming) [17:01] tgardner: they should drop for P, possibly for O if they hit stable and I rebase [17:01] ok, I'm just being OCD [17:01] tgardner: I assume you're referring to ike's patches [17:02] ogasawara, yep [17:03] tgardner: if it makes you feel better, I'm indifferent so if you want to fix up and push over the top, feel free :) [17:03] ogasawara, I'll get over it. some good drugs will help. [17:05] * smb wonders whether there some involved already... [17:19] tgardner, you got any rtl8192cu's ? if so do they work ? [17:19] apw, I can't hear you.... [17:20] damn [17:20] apw, everything I jave is Intel or Atheros [17:20] have* [17:21] apw, think I got one [17:22] apw, Is that one of your new micro dongles [17:22] smb, i got one of those micro 80211.n things ... yeah [17:22] seems to be recognised but not auth ... bah [17:22] think I got a similar one now [17:22] it worked [17:22] but in natty you needed cw-2.6.39 [17:25] doesn't seem to work in oeniric, damn [17:29] * tgardner unleashes his inner OCD on ogasawara [17:30] heh, what now? [17:30] k-team email [17:32] smb, did your fixes for LBM versioning get out ? [17:42] apw, I did versioning? [17:44] The last thing I remember doing was that failure to package the right modules [17:44] Though that was limited to lucid I think [17:44] sconklin, bjf, herton: bug 818175 done, marked qa-testing-passed [17:44] Launchpad bug 818175 in linux "linux: 2.6.38-11.48 -proposed tracker" [Medium,In progress] https://launchpad.net/bugs/818175 [17:44] hggdh: thank you! [17:45] smb, limited to lucid ok [17:46] hi, i downloaded the latest kernel. i have no clue where to start examining it lol [17:47] isnt there something like a schema === sarahsharp1 is now known as sarahsharp [17:49] khanx: no [17:50] khanx: the best place to start is Documentation/ and kernelnewbies.org [17:55] thanks for the tip =) [18:00] the natty update on this t30 has fecked it [18:00] * tgardner --> lunch [18:14] * smb wonders where to test build ti-omap4 kernels... [18:33] smb, gomeisa or tangerine ? [18:34] tgardner, Any would be ok, but there does not seem to be a oneiric-armel schroot on either of them [18:34] smb, I've been using the cross compiler since the armel schroot doesn't work [18:35] tgardner, Ah, ok. Will do the same then [18:35] smb, echo "debuild -B -edo_tools=false -us -uc"|schroot -c oneiric-amd64 [18:35] thanks [18:35] smb, opps, forgot arch: echo "debuild -B -edo_tools=false -us -uc -aarmel"|schroot -c oneiric-amd64 [18:36] Heh, ok was thinking that would need to be somewhere [18:36] smb, besides, the cross compiler method is way faster [18:37] * ogasawara shakes fist at tgardner's contagious OCD and updates ike's patches [18:37] tgardner, Completely agreed. Was not sure the cross compiles is installed or not [18:38] * tgardner enjoys his infectious OCD tendencies [18:38] can you direct them at seeing why i get 3 backlight adjustments for each hard key press :> [19:20] herton: ok, I can reliably reproduce the issue now [19:20] herton: bug updated [19:22] mdeslaur: cool, will give you another test kernel shortly [19:22] herton: great, thanks [20:01] mdeslaur: done, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/814325/comments/23 [20:01] Ubuntu bug 814325 in linux "fuzzy and corrupted display with update in natty-proposed" [High,Triaged] [20:03] herton: trying now, thanks [20:09] herton: sorry, r4 has the regression :( [20:10] mdeslaur: ok, seems only way is reverting that patch for now [20:19] sconklin: just making sure, I am -- right now -- up-to-date on SRUs, correct? [20:20] hggdh: hold on while I double check. I think so [20:21] hggdh: yes, you are caught up :-) [20:28] \o/ [20:28] thank you, sir [20:39] * herton -> eod [20:55] * ogasawara bails for a bit === Quintasan_ is now known as Quintasan === yofel_ is now known as yofel