[00:38] any i686 or pae oneiric kernels newer than 4/19? iwl3945 is broken. [01:04] LLStarks: there is a mainline kernel build from today http://kernel.ubuntu.com/~kernel-ppa/mainline/ [01:04] not i686 [01:06] LLStarks: i really would like to get one i386 kernel too.. the config is not optimal or the kernel code needs a fix. one of those 2 options would give a package [01:46] LLStarks, Kano: kernel.ubuntu.com/~jj/linux-image-2.6.39-0-generic_2.6.39-0.2_i386.deb [01:46] oh. did you just roll that? [01:46] any headers? [01:47] yep, no promise on how long that one will stay there, I have been planning to test on oneiric kernel on my x120 and you just kicked me to do it [01:47] yep [01:47] jjohansen: i need it for mainline because it is binary compatible with debian [01:47] linux-headers-2.6.39-0-generic_2.6.39-0.2_i386.deb [01:48] Kano: ah, sorry. I am not going to be getting to kicking off a mainline build yet. Maybe I can get that rolling later tonight [01:48] would be fine, or rc5 [01:49] the bt config should be better now [01:49] jjohansen, do i need generic? [01:51] cnd, any simple way of doing middle click with synaptics? [01:51] a multitouch gesture or a saner way of clicking both buttons simultaneously [01:52] * RAOF 's synaptics touchpad accepts two and three finger taps as button 2 & 3 respectively. [02:03] raof, vendor string? [02:05] does the touchpad depend on the kernel or userspace driver? === emma is now known as em === fairuz_ is now known as fairuz === ogasawara_ is now known as ogasawara === cking_ is now known as cking === smb` is now known as smb === chrisccoulson_ is now known as chrisccoulson [10:53] cking, we miss you here [10:56] pgraner, that's kind - you miss my acerbic wit? ;-) [10:57] I don't miss the millbank sandwiches [10:57] cking, hehe we miss our iso testing banter [10:58] there's always mumble ... [11:34] cking: ping [11:36] jjohansen, pong [11:37] cking: mumble? [11:37] yep [11:37] hold on === mdz` is now known as mdz === yofel_ is now known as yofel === michaelh1 is now known as michaelh1|away === jjohansen is now known as jj-afk [14:42] hi, why is todays daily kenrel still only amd64? [14:43] didnt somebody say it is fixed [14:45] Kano, to which version in particular do you refer? [14:52] * tgardner sure likes some of the maintenance tools that bjf has written [15:04] bjf rocks [15:05] cking, careful with the accolades lest he get a swelled head :) [15:07] how about bjf's scripts are kinda good ;-) [15:07] heh [15:07] * ogasawara finds it awful hard to work this morning being in sunny san diego [15:08] ogasawara, sounds like fun! [15:08] ogasawara, why didn't you take the whole day off? [15:08] cking, no iso testing for release for you > [15:08] tgardner: I thought in case of any release fires, I might want to stick around to see the carnage :) [15:08] no fires, yet [15:09] ogasawara, looks like its all done. [15:09] apw: congrats on natty btw! [15:09] apw, I missed all the fun in millbank [15:11] ogasawara, i think you did the last push on natty really, so its yours as much as ine [15:11] mine [15:11] cking, heh we missed you too [15:12] hey guys :) [15:21] apw, I didn't miss the sandwiches though [15:22] sandwiches, blech! [15:25] we had _other_ things one day, yep other things [15:25] solid or liquid? ;-P [15:26] apw, indescribable other things? [15:26] mostly yep :) [15:26] joy [15:31] * apw notes that he is still able to use irc, how unusual for release day [15:32] why CVE 2010-4243 has been marked as Invalid for all releases except for Natty? [15:32] ppisati: fs/exec.c in the Linux kernel before 2.6.37 does not enable the OOM Killer to assess use of stack memory by arrays representing the (1) arguments and (2) environment, which allows local users to cause a denial of service (memory consumption) via a crafted exec system call, aka an "OOM dodging issue," a related issue to CVE-2010-3858. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2010-4243) [15:32] bug 768408 [15:32] Launchpad bug 768408 in linux "CVE-2010-4243" [Undecided,Fix released] https://launchpad.net/bugs/768408 [15:32] apw, Either there is more bandwidth, or less pulling, ... [15:34] ppisati, looks like I was working on that last week. I need to get back to it and finish up. [15:35] tgardner: k [15:36] tgardner: i was just wondering why it's Invalid in all that releases since the description says "...before 2.6.37" [15:36] ppisati, well, its possible I haven't done my homework yet. [15:37] k [15:37] no problem [15:38] ppisati, Though it also has happened that they just did not care that much and before 2.6.x meant only a limited timeframe [15:38] They usually don't look too hard how far things really go [15:39] hey guys nice work on release :) [15:39] (they meaning whoever puts the text into the mitre reports) [15:39] hey apw :) time to see with out the sru kernel on my desktop whats broken lol [15:39] ppisati, I only vaguely remember working on it. It may be that the vulnerability did not exist in 2.6.35 or earlier. can't remember for sure. [15:40] tgardner: ok, i'll check === eagles0513875 is now known as zz_eagles0513875 === bjf[afk] is now known as bjf [15:49] apw, still with us ? [15:52] bjf hi [15:53] apw, so just to let you know I have not forgotten your input, if you go to http://people.canonical.com/~kernel/reports/1-day-new.html [15:53] apw, scroll to the bottom of the page, there are two links "DARK" and "LIGHT", click "LIGHT" [15:53] apw, it's still a work in progress but you get the idea [15:53] apw, and it sets a cookie so you only need to set it once [15:54] yeah thats really nice [15:54] bjf, thanks! [15:55] apw, hey! anything for you man! :-P [15:57] bjf, I much prefer the DARK theme === kirkland` is now known as kirkland [15:58] tgardner, me as well, but *some* people ... [15:59] :) [15:59] some people like to MOAN [16:03] bjf, no ubuntu color theme then? ;-) [16:04] cking, can't give away all the secrets [16:05] oh, and I thought one just hacked the appropriate css/themes/*/style.css file.. [16:05] cking, i take patches :-) [16:06] ..and feature requests? I want "Ubuntu colors" :-) [16:06] just kiddin [16:07] nice choice of lime-green. If only we could colorise the kernel like this.. [16:07] cking, you could always file a bug [16:08] what's the project name? [16:09] fwts has a nice logo, I think your webby page should have too.. [16:10] cking, after my rotation on the design team i'll have more skillz [16:12] bjf becomes a DX god === zz_eagles0513875 is now known as eagles0513875 [16:13] purple and green text with grey bold text on a stripey black and yellow background and flashing orange links seems cool to me. === eagles0513875 is now known as zz_eagles0513875 [16:19] apw, do you have a minute to consult about vesafb? === zz_eagles0513875 is now known as eagles0513875 [16:24] tgardner, yep, sup === eagles0513875 is now known as zz_eagles0513875 === zz_eagles0513875 is now known as eagles0513875 [16:27] apw, ah, was looking at bug #772042. I assume its OK for vesafb to legitimately fail to load? [16:27] Launchpad bug 772042 in linux "WARNING: at /build/buildd/linux-2.6.38/arch/x86/mm/ioremap.c:98 __ioremap_caller 0x3fe/0x440()" [Undecided,New] https://launchpad.net/bugs/772042 [16:27] * apw looks [16:29] tgardner, i would expect it to be possible that vesafb insertion could fail as you may not have such a thing indeed. i am not expecting a WARNING though [16:30] apw, its got an SiS VGA adapter which has likely already mapped that I/O space. [16:30] I think the WARN_ONCE is in the wrong place in the ioremap() function, but thats a different issue. [16:30] then yes, that failure sounds sensible, i am not convinced the error should be that severe [16:31] it should just be returning, EBUSY sort of thing [16:31] apw, right. I think the warning should be placed _after_ a successfull ioremap [16:31] its really just a benign sanity check [16:31] yeah i am sure its a warn to make it easy to detect two drivers overlapping [16:33] apw, ok, since the ioremap is failing, I think the warning is bogus. I'll send a patch to Ingo et al to see what they think. [16:33] sounds reasonable to me [16:36] tgardner, apw: is it really reasonable for vesafb to ioremap that much memory ? [16:36] vmap allocation for size 4272398336 failed [16:37] sforshee, hmm, I didn't notice the amount. that seems bogus. [16:37] tgardner, yeah, especially since it's a 32-bit kernel [16:37] sforshee, underfalw ? [16:37] underflow * [16:38] could be [16:38] it's FEA7A000 in hex [16:42] SiS VGA adapter, isn't that an oxymoron? [16:46] sforshee, looks like we need some debug info that dumps the screen_info passed in by the BIOS. [16:47] it doesn't look like its in the boot log by default. [16:47] tgardner, agreed, doesn't look like vesafb does anything to mess up the size that badly [16:49] sforshee, you want to work on this with Timo? pgraner wants me to do some other stuff for next week. === eagles0513875 is now known as zz_eagles0513875 [16:50] tgardner, will do === zz_eagles0513875 is now known as eagles0513875 [16:55] bjf, the only colour wich isn't great on the light theme is the yellow, its too light [16:56] bjf, oh and the white, which i couldn't see at all :) [16:56] yeah, it needs to be black [16:56] apw, i think there are others, they are so light you can't see them, :-) [16:56] apw, it's on my todo list [16:56] bjf, an oppotunity for $white = 'black' [16:56] heh [16:58] oh! just got email that I've won 1.86 million euros!, been nice knowing you [16:58] bjf, You probably have to go to Uganda to collect. good luck. [17:04] * eagles0513875 waves to apw === jj-afk is now known as jjohansen [17:26] Hullo? [17:29] Don't know if anyone is here at the moment but I thought I'd ask about/mention the bug which is stopping the backlight from working under the 2.6.38 kernels. Reading through the bug logs seems to imply that the issue was solved and patched in rc6 (or 8 I forget). Anyway, it's still present, I've tried upgrading to 2.3.39 but still not working so either the patch wasn't committed or it doesn't work for me (I'm u [17:29] sing an intel GMA 4500m). === eagles0513875 is now known as zz_eagles0513875 [17:34] https://bugzilla.kernel.org/show_bug.cgi?id=27702 === zz_eagles0513875 is now known as eagles0513875 [17:38] Kalanac: Error: Could not parse XML returned by bugzilla.kernel.org: Connection reset by peer. (https://bugzilla.kernel.org/xml.cgi?id=27702) [18:28] * tgardner --> lunch [19:49] * bjf is heading out for a walk, biab === bjf is now known as bjf[afk] === michaelh1|away is now known as michaelh1 === bjf[afk] is now known as bjf [20:47] * bjf is back [21:31] hi, who's in charge of the kernel track scheduling? [21:33] jcastro, no one has been appointed. bjf - would like to organize that? [21:33] would you* [21:33] ok so my question is, is it ok if I file the kernel stuff under the hardware track or other, since we don't really have "kernel" per se [21:34] jcastro, seems reasonable [21:34] jcastro, isn't there a foundations track? we've only got a few blueprints that actually need sessions, and they are kind of foundation related. [21:35] that would be fine too [21:35] jcastro, lets do that [21:35] it's just right now they're in the summit system and they don't have a home so they get put in the sidebar and I want to make sure they hit the schedule [21:36] tgardner: ok I'll bug bjf when he's around for the rename, don't want to break links I'm not aware of, etc. [21:36] jcastro, i thought pganer already suggested "other" [21:36] jcastro, but I could be mistaken [21:36] bjf: ok, but the BPs themselves aren't named other-kernel-o-whatever [21:36] just kernel-o-whatever [21:37] bjf: so maybe just file them where they would make the best sense, be it foundations, hardware, or other [21:38] jcastro, https://wiki.ubuntu.com/KernelTeam/Specs [21:38] jcastro, we only have 6 [21:39] jcastro, we were hardware last time [21:39] jcastro, so I need to go rename them again, why does it seem like we have to rename thise a dozen times for every UDS [21:39] however you want to do it is fine with me, I'm just saying that you'll have to manually schedule if you just keep plain "kernel-o" [21:40] bjf: I dunno, I don't make those decisions, if it makes you feel better I am stuck renaming things every day no matter what we do. :-/ [21:41] jcastro, it would seem that by the time we put out the call, there should be a wiki page which explicitly states where each team should put their blueprints [21:41] jcastro, anyway, you are suggesting I pick somewhere (hardware or other) so they can be auto scheduled [21:41] jcastro, i can make that happen [21:41] right [21:42] jcastro, consider it done, if for some reason you see any after a couple minutes from now that are not in "other", let me know and i'll fixe them as well [21:43] * jcastro nods [21:43] bjf: I know it sucks, but we'll have the schedule done by tomorrow, normally we don't even get this far until like the weekend before, so at least you'll know your schedule before you get on that plane! [21:44] jcastro, we still need to schedule our morning roundtable [21:44] do you want one per day? [21:44] I can do that for you [21:44] jcastro, yes, first session, every day [21:44] jcastro, that would be a big help [21:44] bjf: about how many people on your team these days? [21:44] I can put you in the same room each day [21:45] tgardner, should we count hwe folks for our morning roundtable? [21:46] bjf, I suppose some would like to join us [21:46] jcastro, make it ~25 [21:48] jcastro, all blueprints changed to "other-kernel-o-*" [21:49] ok, your roundtables are now set [21:49] I've stuck you guys in "Krudy" every day, 9am as your home room [21:49] jcastro, awesome! [21:56] jcastro, dude! an icon on the session that launches a pad, that's nice! [21:57] bjf: not ready yet, but yeah, once we have it, you can just paste the notes right in there [21:58] so when you get to your session, blam, you have all your stuff and agenda already set out [21:58] it's brilliant. [21:58] jcastro, have we stress tested the server at all ? [22:01] bjf: IS has been slammed with release, hopefully I can poke them tomorrow and see what's up [22:01] jcastro, i just don't want to get there and have it roll over and die under load [22:02] it's running great for the openstack guys at their conference, but I don't have numbers on how many people are hitting it [22:28] jcastro: I see this year there is a "security" top-level? last year we did other-n-security etc... what should we do for roundtables and things this year? [22:45] jcastro: hm, ignore me. we used both other-n and security-n last time. :) === bjf is now known as bjf[afk] [23:21] kees: for roundtables you can make them by hand in summit, it's easier than bp's, or just ping me tomorrow and I'll make them for you [23:23] jcastro: no worries; I've already made and placed them. :) [23:23] I was just trying to figure out the name for them. [23:24] I do: [23:24] community-o-monday-roundtable [23:24] and the kernel ones are similar