=== anubhav_ is now known as anubhav === smb` is now known as smb [07:47] * smb shuffles in [07:54] * apw waves to smb [07:54] * smb waves back to apw [08:17] ogasawara, about ? [10:04] ogasawara, ok then ... i've fixed up the lack of attribution of the rebase to 3.0 and pushed on an abi bump so it'll build [12:35] apw: okay, so, the pending/released flipflop we have is due to running certain scripts on our end in the wrong order. basically we need to do gitscan, then usn-mismatch, then the usn-publication-sync. can you do a quick gitscan and I'll merge and sync correctly today? [12:36] if we do the usn-pub-sync first it thinks the usn db shows a given cve as being published and closes it. [13:05] apw: cool, thanks for doing the rebase [13:05] ogasawara, i didn't re [13:05] ogasawara, i didn't rebase ... you did [13:06] kees, i think now that the 0711 has a version for it and its released i won't argue with the usn [13:06] apw: hrm, I hadn't done the actual rebase to 3.0 final [13:06] apw: I was gonna do it first thing this morning, eg now [13:06] apw: maybe tgardner did it? [13:07] apw: true, but after making the usn-publisher not change versions, we ended up in a kind of goofy situation. perhaps I can further improve it to ignore versions greater than the USN's package version [13:07] the committer was you i thought ... and on rebase normally the committer becomes the rebase-er i thought [13:07] so ... likely was tim anyhow but its looking ok [13:08] ogasawara, i'd like to quickly test with overlayfs and push that in before it gets uploaded if that ok with you [13:08] apw: works for me [13:08] apw: I've got some more modular config review changes to push too [13:08] ogasawara, go ahead and push them whenever [13:10] apw: pushed [13:10] ogasawara, i have built kernels with overlayfs in, if that boots in the livecd i'll push it and then test aufs [13:10] cool [13:11] i have kernels with and without [13:11] * apw shouts at his download link ... get on with it i need that CD image now [13:12] kees, ahh i see ... if the version doesn't match my version i will fix ti and move it pending [13:12] kees, but if its matching the version it would be, even if its only in -proposed then i won't touch it [13:12] as i actually can't tell and assume you do any pending -> something transitions [13:14] * kees nods [13:29] ogasawara, does it seem like a good day to you for an upload ? [13:29] tgardner: yep, just waiting for apw to finish up with testing some overlayfs patches he wants to shove on top [13:30] tgardner: I assume it was you who did the 3.0 rebase? if so, thanks! [13:30] ogasawara, cool. I pushed the rebase last night so I could try it out on emerald [13:34] ogasawara, bjf: headed into the office now. can't get in before 0700. biab. [13:39] fri kernel upload? brave! :) [13:40] kees, na not as much as you'd think as it takes days to build, and does't take effect till meta goes in [13:40] kees, so its actually a perfect day to upload [13:41] oh, yeah, i guess with an abi bump, that is perfect [13:41] yeah [13:41] kees, so we have a couple of cves outstanding against hardy, one of which at least is going to be a heap of mess required to backport anything close to something thats workable [13:42] whats our policy on that kind of thing, if we decide the risk is very great [13:47] which cve is it? if it's not a serious issue, we can choose to ignore it or work around it. it's a case-by-case thing. [13:48] kees, the one i am particularly unsure about is the epoll recursion one, which likely is a heap to fix, i've not yet tried but it feels like it may well be... so was interested in the process [13:48] i try to be risk adverse, which normally means fixing bugs, but if things are really ugly, not fixing is the safe plan [13:49] i'll let you know when i've tried and failed :) [13:49] heh [14:41] tgardner: you rebuilding oneiric chroots on tyler? [14:41] ogasawara, yep [14:42] * ogasawara goes to hammer on tangerine then [14:42] ogasawara, they were in an intermediate state, so I thought I'd just rip 'em out and start over [14:42] tgardner: works for me [14:42] ogasawara, didn't think you were on tyler [14:42] yet [14:43] tgardner: nope, only just noticed when I went to kick off a build [15:12] wow, we're not in the list of top 5 packages for bugs reported against last week [15:12] we must be losing our touch :) [15:13] ogasawara: I can 'fix' that [15:13] heh :) [15:15] that's actually great to know. I've been worrying about the regressions we did have, but perhaps we're having fewer and tracking those better - so it seems worse than it is [15:30] * ogasawara back in 20 [16:24] tgardner: can you take a look at the omap4 build? [16:25] tgardner: i was told to disable the abi check since this was a handcrafted pkg [16:25] tgardner: but it seems it was checking there [16:30] ppisati, yeah, I'm in the middle of figuring it out, but got distracted. I'll get it re-uploaded sometime today. [16:30] tgardner: k [16:31] ogasawara, so is my S3 systemtap script broken for oneiric? [16:31] cking: bug 754711 mentions it is, but i've not actually confirmed [16:31] Launchpad bug 754711 in linux "[Dell Studio XPS 1340] Doesn't enter suspend mode" [High,In progress] https://launchpad.net/bugs/754711 [16:32] ogasawara, ok - I will look at that in the next 24 hours [16:32] cking: no hurry, it can wait till next week [16:33] tgardner: i think it misses three files [16:33] tgardner: http://pastebin.ubuntu.com/650086/ [16:36] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/809878 [16:36] Ubuntu bug 809878 in linux "Wireless network & graphics card stop working after upgrade to 2.6.38-10" [Undecided,New] [16:40] Daviey, did we confirm it was the java update which was the problem in this case ? [16:45] bjf: it looks like my kernel-driver- tagging of oops reports was incomplete as I only got those with EIP and not RIP [16:45] apw: We haven't identified what introduced it yet. [16:46] Daviey, ok cause i just heard people blaming it as if it was definatly it [16:46] apw: blaming "it"? [16:46] The two obvious candidates are either openjdk or kernel. [16:46] Daviey, blaming openjdk change for starting to tickle the existing kernel bug [16:47] Daviey, you were going to back that out to confirm i believe [16:47] apw: Who did you hear? [16:47] Daviey, kate just said "we would have caught the issue with openjdk ...." [16:47] and i wanted to confirm it was confirmed that it was the openjdk update which was the 'cause' [16:47] apw: Ah. you are sneeking on the call? [16:48] (obviously the kernel bug is real and is the root cause) [16:48] not sneaking i was invited [16:48] apw: Oh sure, didn't mean it like that. [16:48] apw: I'm not entirely sure it's worth the time investment to discover it.. although sucky. [16:57] sforshee, aff, I don't understand that intel bug [16:58] it has 1+ year old and it is a very basic feature [16:58] and the Chris Wilson doesn't say nothing :( [16:59] ppisati, refetch HEAD for ti-omap4. I think it'll build now, but I'm still gonna upload to the c-k-t PPA first just to make sure. [16:59] mfilipe, which bug are you talking about? [17:02] jjohansen: the talk you gave at ubuntu dev week about bisecting, did that get translated into a wiki by chance? [17:02] jjohansen: no worries if not, just wanted to point someone at it for how to bisect. I can always point them to the classroom logs. [17:02] ogasawara: hrmm, no its on the todo [17:03] sforshee, https://bugs.freedesktop.org/show_bug.cgi?id=38750 [17:03] Freedesktop bug 38750 in DRM/Intel "[Arrandale] Intel Core i3 External Monitor Wavy Output" [Normal,New] [17:03] ogasawara: logs are here https://wiki.ubuntu.com/MeetingLogs/devweek1107/KernelDebugging [17:04] jjohansen: thanks [17:04] mfilipe, ah, that one :) [17:04] yeah, it's annoying, I think they understand what needs to be done but just haven't done it, not sure why [17:05] sforshee, yep, he didn't say nothing yet and has MUCH people want the solution [17:46] ogasawara, ok i've pushed up the overlayfs changes, they seems to build, and boot ok, and i am able to use CDs with overlayfs and aufs2 as their union soln. [17:47] apw: awesome, I'll get it prepped for upload then [17:47] ogasawara, cool, thanks [17:59] apw: did you want your SOB added to the overlayfs patches? I've gotta mark some patches from the delta review as SAUCE anyways, so I could fix them up while I'm at it [18:14] ogasawara, ahh yes i meant to do that [18:14] apw: ack, I'll add it and push [18:14] good spot thanks [18:31] * bjf -> lunch === bjf is now known as bjf[afk] [18:31] * cking --> EOD [18:40] hey all.. anyone from the kernel team have any suggestions on how to handle Bug #799630 and its relatives? [18:40] Launchpad bug 799630 in drbd8 "package drbd8-source 2:8.3.7-1ubuntu2.1 failed to install/upgrade: drbd8 kernel module failed to build" [Low,Incomplete] https://launchpad.net/bugs/799630 [18:51] adam_g, why is there a dkms package for drbd at all? The driver has been in mainline for quite some time. [18:52] sforshee: it merged 2.6.33, so it was out-of-tree in lucid [18:52] anyone whos back ported a kernel will break its updates [18:53] adam_g, if you're using a backport kernel is the driver not included? [18:53] i believe there is a similar issue right now with the lirc IR drivers [18:54] sforshee: it is, but on updates the dkms module attempts to rebuild itself against the current kernel that includes drbd, so it fails [18:57] adam_g, I see. Some of the errors look like redefinitions of stuff already defined in the kernel, and I don't understand why drdb is defining them at all. [18:57] Some of the others look like changes to kernel functions [18:58] so to fix it someone would have to go to the trouble to make drbd understand the differences between the versions at compile time [19:00] right [19:01] thats the question. is that something we are interseted in doing to support back ported kernels? or is there packaging magick we can employ to avoid the issue all together? [19:04] cnd, ping [19:04] pgraner, hey [19:04] cnd, remember at UDS you fixed a box for me with the Alps touchpad? [19:05] cnd, what did you do? [19:05] cnd, I think I found the machine but was going to confirm to see if you fix is there [19:06] pgraner, IIRC I added an /etc/modprobe.d/ file to tell it to use a specific PS2 proto to talk to the device [19:06] but I'm having trouble recalling the specific issue and why that solved it [19:07] if you need info you'll have to refresh my memory more :) [19:07] cnd, it was the right click was actually a middle, and a a tap was right mouse click [19:07] cnd, or something like that [19:07] cnd, and you had to use the proto=bare [19:08] ok [19:08] cnd, whats the easiest way to tell if its an alps? [19:09] hmmm [19:10] trying to remember [19:10] it's serial/ps2 so it doesn't have a hid vendor/device id [19:10] but there is some form of id [19:10] cnd, udevadm? [19:10] I need to look in the kernel code [19:11] adam_g, I don't know the answer to those questions [19:12] cnd, I was looking at it the other day. It seems each particular driver just does some magic to try to figure out if it supports a given device [19:12] I didn't see any universal way to identify a vendor/model [19:12] pgraner, the way the ps2 mouse module works is there's a core piece of code that calls out to all the loaded modules [19:12] and each module tries to identify the device [19:12] there's no standard way of identification [19:13] each manufacturer has their own special ps2 query command to get the version [19:13] and say it's a new device that linux doesn't know about, then it will just show up as a bare ps2 mouse I think [19:14] so the only way to know for sure is to figure it out from the product specs [19:14] and/or other people's bug reports :) [19:14] for alps, chances are that if the trackpad isn't working right, then it's an alps :) [19:15] cnd, pgraner: If you can get the OEM Windows install on a system it can tell you what kind of device it is too [19:17] sforshee, cnd, all demsg tells me is that is: i8042: PNP: PS/2 Controller [PNP0303:KBC0,PNO0f13:MSA0] at 0x60,0x64 irq 1,12 [19:18] pgraner, if you take out the modprobe.d file so it doesn't lock to the bare proto, it might be recognized as alps [19:18] comment out the modprobe.d file and then rmmod and modprobe psmouse [19:18] sforshee, cnd, found it: input: AlpsPS/2 ALPS GLidePoint as /devices/platform/i8042/serio1/input/input10 [19:19] sforshee, its on my netbook :( [19:24] * jjohansen lunch [19:27] heh === bjf[afk] is now known as bjf [20:37] * tgardner --> EOW === yofel_ is now known as yofel [21:10] * herton --> eow