[00:21] RAOF: actually, the fix for geode's ZTV module is a lot simpler than that: we simply need to stop including "linux/videodev.h" and that's it. our driver is already V4L2. :) [00:21] Q-FUNK: Heh, ok. [00:22] I admit, I didn't look any further than “ooh, V4L1. No-one will miss it if I just disable it” :) [07:36] Anyone feel like sponsoring a final round of rebuilds for the X server? [07:41] RAOF: sure [07:42] tjaalton: http://cooperteam.net/Packages [07:43] Thanks. [07:50] RAOF: yeah, uploaded [07:51] Ta. [12:08] if one has nvidia h/w one does not get 3c in nouveau, is there a recipe for getting the gallium drivers for that ? (and do they indeed sort that out) [12:18] 3d, yes, install libgl1-mesa-dri-experimental (or such, don't have the package name handy) [12:24] apw: ^ [12:53] tjaalton, thanks === yofel_ is now known as yofel [13:10] tjaalton, actually jockey offered that to me, and installed it, but it doesn't seem to work at all [13:11] apw: breaks stuff? which card do you have? [13:12] tjaalton, no breakage, just classic desktop and unity saying 'software renderer in use' [13:12] 05:00.0 VGA compatible controller: nVidia Corporation GF100 [GeForce GTX 480] (rev a3) [13:12] apparently [13:12] hmm, wonder if it's too new [13:12] tjaalton, could well be ... such is life [13:13] Xorg.0.log should tell [13:13] kms works though? [13:13] yeah it switches into high res i recon [13:13] k [13:13] nouveau is kms only right ? [13:14] yes [13:14] [ 27.543] (--) NOUVEAU(0): Chipset: "NVIDIA NVc0" [13:14] ok, but does it load the DRI2 driver? [13:14] or even attempt that [13:14] [ 27.426] (II) Loading extension DRI2 [13:14] nothing to say if it worked or not [13:15] put the full log on pastebin [13:15] * apw should warn that this is a very new machine and quite possibly has stupidly new h/w [13:16] the kernel won't even boot the secondary cpus, not that that should matter [13:16] pastebinit Xorg.0.log [13:16] http://paste.ubuntu.com/563843/ [13:16] thanks [13:16] [ 28.528] (EE) NOUVEAU(0): Error initialising acceleration. Falling back to NoAccel [13:16] guess that's it then [13:19] and the line above [13:24] apw: the gallium driver for nvc0 was merged upstream on nov 12th, dunno if it's in 7.10 or not [13:24] ah ... very very fresh then [13:24] tjaalton: afaict it's not [13:25] jcristau: yep, latest commit from october [13:25] * apw will have to purchase an older one then and ram that in instead [13:26] apw: xorg-edgers might have a newer mesa for you :) [13:29] yep [13:57] how well does X cope with two cards inserted, different manufactueres, i'd be happy selecting one or other in some way; making a dual card test setup [13:58] it sometimes work. [13:58] jcristau, sounds like something to avoid then [15:05] apw, bryceh: FYI, I just submitted # 693093 [15:06] apw: thx [15:06] bug #693093 [15:06] Launchpad bug 693093 in linux (Ubuntu) (and 1 other project) "[i945gme] 2.6.37-10.24: Black Screen on Boot (affects: 10) (heat: 136)" [High,Confirmed] https://launchpad.net/bugs/693093 [15:07] marjo, thats an old bug ... [15:07] apw: sorry [15:08] apw, bryceh: # 714635 [15:09] apw, bryceh: #714635 [15:09] bug 714635 [15:10] Launchpad bug 714635 in linux (Ubuntu) "[i945gme] 2.6.38-2: Black screen on boot (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/714635 [15:14] marjo, did you get the intel_reg_dumper outptu ? [15:14] apw: yes, attached it already [15:15] marjo, to the right bug ? [15:16] marjo, oh its mixed in with the apport stuff, didn't know you could do that [15:29] marjo, we probally should get the Xorg.log.0 onto that bug as well [15:33] apw: done [15:39] tjaalton, i'd say that that X startup has gotten stuck, HW cursor is not the last thing one sees on my boxes [15:41] apw: yeah [15:43] marjo, i wonder if you are able to try an strace on the Xorg process once it is stuck [15:43] to find out whether it thinks it is in the kernel [15:44] apw: sure, can you please give me specific directions? [15:44] ps -ef | grep /X [15:44] take the pid from that ... second column [15:46] could also try with 'Option "HWCursor" "off"' [15:49] apw: next? [15:49] you wait while i recover the machine i tried the command out on (stupidly) from within X, no no no, supid [15:49] :) [15:50] yeah i know, /me takes the dunce of the day hat and puts it on [15:51] need to run, back in 3,5h [15:52] ok ... back. [15:52] marjo, ok this should do the trick: sudo strace -p `pidof X` [15:52] and see what that says [15:52] not something one should run in an xterm, no not a good idea [15:54] apw: "attach: ptrace(PTRACE_ATTACH, ...): No such process" [15:55] apw: i'm doing this through ssh [15:55] ps -ef | grep X [15:55] whats the output of that look like [15:57] marjo, ^^ [15:57] 1000 1298 1181 0 10:57 pts/0 00:00:00 grep --color=auto X [15:57] so you have no X running at all then [15:58] apw: ack [15:59] add that info to the bug, i would not not expect X to exit without telling us why it did so [15:59] bryceh, ever heard of that ^^ [16:00] can happen if it aborts e.g. because of a missing symbol [16:00] would show up in gdm log [16:01] apw: what's weird is I can go into recovery mode/low graphics mode [16:01] marjo, could we have the gdm log as well please [16:01] apw: will do in a few minutest, ok? [16:01] minutes, ok? [16:02] no rush [16:02] apw: thx [16:06] apw: i see a few *.log and -slave.log files [16:06] apw: any specific one to upload to bug report? [16:06] marjo, probabally :0.log [16:07] apw:ok, will do [17:21] apw: :0.log attached to bug #714635 [17:21] Launchpad bug 714635 in linux (Ubuntu) "[i945gme] 2.6.38-2: Black screen on boot _after_ visible splash (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/714635 [17:50] win. [17:50] /usr/bin/X: symbol lookup error: /usr/lib/xorg/modules/drivers/intel_drv.so: undefined symbol: intel_batch_wait_last [17:54] marjo: sounds like you have an obsolete version of the driver [17:56] jcristau: which driver? [17:57] the one that's missing a symbol [17:57] the xserver-xorg-video-intel package [18:03] jcristau: when you say obsolete, do you mean that i don't have an updated one? or is there a new one that i did NOT pick up? [18:05] i mean that bug was fixed in 2:2.14.0-1ubuntu2 [18:05] jcristau: installed xserver-xorg-video-intel and now things are working! [18:07] jcristau, apw: thx much! [18:07] marjo, yay ... [18:08] jcristau, thanks for the help getting to the right log :) [18:08] np [18:08] jcristau: was there a previous bug I can refer to in my bug report? [18:08] none mentioned in the changelog afaict, so no idea [18:09] jcristau: that's too bad [18:09] so i'll just say fixed with xserver-xorg-video-intel 2:2.14.0-1ubuntu2, right? [18:13] so i'll just say fixed with xserver-xorg-video-intel 2:2.14.0-1ubuntu6, right? [18:24] marjo, yes fixed with xserver-xorg-video-intel 2:2.14.0-1ubuntu2 [18:25] marjo, that was an obvious issue we caught right away; we simply hadn't caught it before uploading due to insufficient testing === fosdemlogger is now known as austrialogger === austrialogger is now known as apachelogger [21:49] Mornin all. [21:52] heya RAOF [21:53] jcristau: Good $TIME_OF_DAY :) [21:53] :) [21:54] Hm. apw has an nvc0 card? Congratulations for volunteering to test the acceleration we may or may not merge in for those cards! [21:55] it's a wheezy evening. [21:55] I presume wheezy is currently in a mad state of flux? [21:56] yeah [21:57] sid got libreoffice today [21:58] probably new compiler and libc soon.. [21:58] Any coordination with our libreoffice maintainer? [21:59] aiui your libreoffice packages were based on the ones in experimental [21:59] Heh. Now sid's unfrozen I'll need an actual sid chroot for testing rather than everything going through the experimental one. Let's see if it's debootstrappable! [22:04] RAOF, seems so ... its scheduled for an unscrewing and shoving in a box shortly for replacement with something supported, but happy to test [22:05] apw: How well did it work (3D excepted) while it was in there? [22:05] Did you notice that there was no 2D acceleration, for example :) [22:05] RAOF, its still in there, runs in 2D well enough [22:05] heh no, but the machine has a lot of CPU :) [22:06] I suspect that correlates well with brand new GPUs :) [22:07] RAOF, but if you have some test bits in a bucket somewhere please point me at it [22:07] xorg-edgers has the test bits. [22:10] does ubuntu-x have a dedicated triage team like the the ubuntu mozilla squad? [22:11] RAOF, ahh ok will try and get those one it, when i've bought something that it can drive other than the 37in tv in the front room [22:11] No. We've got bryceh and his amazing band of arsenal scripts, though. [22:12] apw: Yeah, no hurry. I'm not in a great rush to grab new libdrm, nouveau packages to speed up but possibly break new cards. [22:12] because i've been thinking about joining the ubuntu-x-swat team and i'm not sure if i have enough credentials [22:17] or quota for your inbox ;) [22:18] meh, wanted to try alpha2 & nouveau, but the dvd-drive seems broken [22:18] LLStarks: You're welcome to start triaging, although we're actually pretty on top of it at the moment. [22:20] What do you want to do in the swat team? [22:25] be mindful of my hardware and continue giving good reports regarding how stable and unstable bits affect said hardware [22:26] i know a little c, but i am not well-versed in the graphics stack to do any coding [22:28] That can change :) [22:30] A good way to learn how to give good reports is to look at what happens when bugs get upstreamed - what questions get asked, etc. [22:30] Also, if you want to invest a little time then bisecting regressions is an excellent way to help get them fixed, and can get you a little familiarity with the build process. [22:31] i've been bisecting -intel as needed, getting proper kernel drm bits is more of a pain for me [22:32] Yeah, that's quite awkward. [22:34] i've been looking for a good jumping in point for learning about the stack. drm, glx, xserver, etc [22:35] usually a jumping in point is a bug that annoys you so much that you fix it. [22:37] had that the other day with libswscale. feels good man. [22:37] The stack is actually a lot less complex than it looks from the outside. [22:43] heya RAOF [22:44] bryceh: Good evening. Or afternoon. [22:44] Or whenever! [22:50] ok, alpha2 boots fine with a working media, but compiz fails after installing the nouveau gallium driver [22:50] compiz: nv50_pc_emit.c:863: emit_flop: Assertion `STYPE(i, 0) == 0x09' failed. [22:50] crashes there [22:51] and the livecd still offers to install the blob [22:51] thought it was disabled for now [22:51] :( [22:51] Yeah, so did I. [22:51] Oh, actually - Jockey doesn't know about server ABIs. [22:51] right [22:51] But I thought we *were* disabling it on the livecd. [22:51] and actually, talked with pitti about it and he [22:52] damn us layout [22:52] anyway pitti said he would add that support once the drivers itself had all the deps/breaks in place [22:53] the blobs, that is [22:53] Ah, good. [22:54] RAOF: is mesa 7.10+git possible for natty? [22:54] will try edgers tomorrow to see if compiz works with it [22:55] We can certainly try. We'll pull in 7.10.1. [22:56] sure, and this crasher is something that should be .1 material anyway [22:56] assuming you trust the nouveau people to care about stable branches [22:56] hehe [22:56] anyway, 2D is very snappy on my 8600GT [22:57] Well, if we can identify the fix, applying it hopefully won't be *too* hard. [22:57] Yeah. Nouveau's 2D is excellent. [22:59] maybe ill just upgrade it lucid->natty anyway, compiz aint _that_ important [22:59] but tomorrow [22:59] I think you might find that compiz will work as long as you're not loading unity. [23:00] oh, hmm [23:01] k, will play with it in the morning [23:01] night! ->