[01:25] <bryceh> RAOF, dunno if you've run across it but there is some sort of OOM situation that crops up during livecd testing
[01:25] <bryceh> RAOF, which has had a tendancy to show itself as an X crash somewhere in the error handling logic
[01:25] <RAOF> bryceh: I don't do much livecd testing, so I haven't hit it.  Fun!
[01:26] <bryceh> I've put in some patches for those crashes, but the root cause of the OOM situation remains a mystery
[01:31] <bryceh> RAOF, anyway, it's been sort of wack-a-mole.  Fix a crash today, wait a couple days, and see where X crashes next.  :-)
[01:31] <bryceh> RAOF, so keep an ear open for people with bugs like this one - https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/720445
[01:31] <ubot4> Launchpad bug 720445 in xorg-server (Ubuntu) "Xorg crashed with SIGSEGV in _start() (affects: 1) (heat: 10)" [Medium,New]
[01:32] <bryceh> specifically, the pattern in the Xorg.0.log - http://launchpadlibrarian.net/64482277/XorgLogOld.txt
[01:32] <bryceh> it's a sudden spew of "[   191.259] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Cannot allocate memory" followed by a crash somewhere in the error handling stack
[01:33] <RAOF> We don't have memory details in the apport crashes, do we.
[01:34] <bryceh> unfortunately no
[01:35] <bryceh> I don't think I had that problem when installing from CD.  maybe using usb keys it's different?
[01:36] <RAOF> Possibly?
[01:36] <bryceh> RAOF, so far, it's entirely been canonical people (mostly in the testing teams) who have reported these kinds of bugs, not community folk in the general public
[01:36] <bryceh> so that seems odd
[01:37] <RAOF> Something has to be leaking memory somewhere, and it has to be pretty early into boot.
[01:37] <RAOF> It's even possible that it's the union filesystem we use.
[01:37]  * bryceh nods
[01:38]  * RAOF wants a faster laptop.  Maybe with 8 processors, or something.
[01:38] <bryceh> given that we don't (afaik) have such a memory leak outside the livecd environment, it makes me doubtful that it's X leaking, however anything's possible I suppose
[01:38] <RAOF> It could be something in the live environment that X isn't handling.
[01:38] <RAOF> Any of these crashes with anything but intel?
[01:38] <bryceh> all have been with intel
[01:39] <bryceh> it might not be significant though, I think the vast majority of the testing hw is intel gfx
[01:39] <RAOF> Yeah.
[01:43] <RAOF> …-Wl,-rpath,:::::::::::::::.  Really, compiz?  Really?  You need to check all those empty linker paths?
[03:10] <bryceh> RAOF, I'm getting towards EOD... any bits or patches you need uploaded?
[03:12] <bryceh> ᑕᓗᖑᓱ ᑕᓗᖑᓱ ᑕᓗᖑᓱ
[03:12] <RAOF> mesa would be nice.
[03:12] <RAOF> Um, ???
[03:12] <bryceh> been having xkeyboard-config fun today ;-)
[03:13] <RAOF> What was that string? :)
[03:13] <bryceh> it was "blah blah blah" in Canadian Inuktitut
[03:13] <RAOF> :)
[03:13] <bryceh> which I've never heard of before but has nice letter forms
[03:14] <bryceh> looks easy to draw in snow or polar bear hides
[03:14] <RAOF> :)
[03:14] <RAOF> I haven't yet tested the final mesa build on my netbook (stupid i386!), though it should be good.
[03:14] <bryceh> got a .dsc or git branch ready for mesa?
[03:15] <RAOF> http://cooperteam.net/Packages
[03:15] <bryceh> ok, I think with mesa especially, I shouldn't upload at EOD
[03:15] <bryceh> but I'll load it on a few systems tomorrow and test it out, and if it looks good will upload it
[03:16] <RAOF> Yeah, good call.
[03:17] <RAOF> I might have the various other pieces of the r600g switch ready by then, so we could do that at the same time, or in a subsequent upload.
[03:17]  * bryceh snags and starts a build
[03:17] <bryceh> sounds good
[03:18] <bryceh> I tend to favor being prolific with uploads when we're not in release crunch; it can give folks extra versions to test, which can help in narrowing down when issues appeared
[03:18] <RAOF> Right.
[03:23] <bryceh> RAOF, changelog should mention the -1ubuntu[23] changes too
[03:23] <bryceh> the two patches can be dropped, but the libudev-dev change needs mentioning
[03:24] <RAOF> The 1ubuntu[2,3] changes are in the changelog.
[03:25] <RAOF> Ah, but I forgot to include the Debian revisions in the .changes, however.
[03:25] <bryceh> oh I mean in the 'Remaining Ubuntu changes' section
[03:26] <RAOF> Oh, right.
[03:46] <alex_mayorga> bryceh: ping
[03:50] <bryceh> yeah
[03:52] <alex_mayorga> I've updated bug 713781
[03:52] <ubot4> Launchpad bug 713781 in xserver-xorg-video-nouveau (Ubuntu) (and 1 other project) "[natty] Video corruption on kernel 2.6.38-1-generic and nVidia Corporation GT216 [GeForce GT 230M] (rev a2) (affects: 1) (heat: 10)" [Undecided,New] https://launchpad.net/bugs/713781
[03:52] <alex_mayorga> please let  me know if anything else is needed
[03:53] <alex_mayorga> still not working on 2.6.38-3
[03:53] <alex_mayorga> should I rename it
[03:53] <bryceh> alex_mayorga, thanks.  I can take care of that
[03:54] <alex_mayorga> should I gather something on the working 2.6.37?
[03:54] <alex_mayorga> bryceh: thanks to you for caring :)
[03:55] <RAOF> Is that the one I responded to…
[03:55]  * RAOF wonders, and checks.
[03:55] <bryceh> RAOF, yep
[03:55] <bryceh> RAOF, let me know if you'd rather do the upstreaming
[03:56] <bryceh> alex_mayorga, no that's not necessary, although it might be nice to know what version last worked properly for you
[03:56] <RAOF> Nah, I'm happy for you to upstream.
[03:56] <alex_mayorga> 2.6.37-12-generic
[03:57] <alex_mayorga> bryceh: that's in comment #5
[03:57] <bryceh> great thanks
[04:01] <bryceh> alex_mayorga, one thing to try, if you can reproduce it on a livecd or liveusb, is to try booting the i386 image. Could be a 64bit specific issue.
[04:02] <alex_mayorga> bryceh: which one should I try? daily?
[04:02] <alex_mayorga> would need to find a USB though
[04:02] <bryceh> a daily would be fine, whatever works for you
[04:04] <alex_mayorga> I'd give it a try if I find a big enough USB
[04:18] <bryceh> ok, forwarded as https://bugs.freedesktop.org/show_bug.cgi?id=34371
[04:18] <ubot4> Freedesktop bug 34371 in Driver/nouveau "[natty] Video corruption on kernel 2.6.38-1-generic (and on -3) and nVidia Corporation GT216 [GeForce GT 230M] (rev a2)" [Normal,New]
[04:19] <bryceh> alex_mayorga and RAOF, please add yourselves to the CC of that bug ^^ 
[04:20] <alex_mayorga> bryceh: done, thanks!
[12:30] <lilstevie> I am having a problem with X not giving me any output into why it isnt outputting to the display, is there anything better than invoking X with -verbose 20?
[12:31] <tjaalton> which release?
[12:32] <lilstevie> 1.9.0
[12:32] <tjaalton> so ubuntu 10.10?
[12:33] <lilstevie> yes
[12:33] <tjaalton> livecd or did it break by itself?
[12:33] <lilstevie> http://pastie.org/1570123 <- this is the output
[12:33] <lilstevie> armel
[12:33] <lilstevie> never had it running
[12:34] <tjaalton> fix the kernel first
[12:34] <tjaalton> there are no input devices it seems
[12:34] <lilstevie> well thats because GPIO buttons and touch screen are the only input devices
[12:35] <lilstevie> what is wrong with the kernel?
[12:35] <tjaalton> still, it's either kernel or udev failing
[12:35] <tjaalton> not X
[12:36] <lilstevie> ok
[12:36] <tjaalton> no device has ID_INPUT set by udev
[12:36] <lilstevie> hm ok
[12:36] <tjaalton> so either the kernel has no such devices or udev fails otherwise
[12:37] <tjaalton> hum no, it did find the touchscreen
[12:37] <lilstevie> ok so that is stopping X from working
[12:37] <lilstevie> the backlight of the lcd flashes on and off
[12:38] <tjaalton> dunno why config/udev spams the log like that
[12:39] <tjaalton> if there's natty alpha2 image for arm, you could try that if it fails the same way
[12:39] <lilstevie> heh
[12:39] <tjaalton> bbl->
=== yofel_ is now known as yofel
[15:38] <bigon> hi when compiling examples of champlain from GIT on natty I get
[15:38] <bigon> http://pastebin.com/9jhJ2WNJ
=== JanC_ is now known as JanC
[17:42] <Darxus> Might be worth cherry picking this, required for wayland outside of X with nouveau:  http://cgit.freedesktop.org/nouveau/linux-2.6/commit/?id=8e645575d469bf08c9d5d98a101ef4cfce6a9180
[17:42] <Darxus> Although I don't know if everything else needed is already in the natty kernel.
[17:51] <tjaalton> roughly 80 bugs against fglrx-installer (out of the 370 filed) where the kernel module failed to build
[17:51] <tjaalton> because there were unofficial kernel packages installed, and dkms barfed
[17:52] <Sarvatt> tjaalton: you missed the cleanup of those about 2 months ago too, was closer to 800 :)
[17:52] <Sarvatt> across all the dkms packages
[17:52] <tjaalton> Sarvatt: and no-one thought about fixing the root problem?-)
[17:53] <tjaalton> anyway, it should be doable
[17:53] <tjaalton> bbl->
[17:55] <Sarvatt> tjaalton: what do you mean? we did at the time
[17:56] <Sarvatt> oh you mean just the ones with unofficial kernel packages?
[17:56] <Sarvatt> missed that part, sorry
[18:01] <tjaalton> yeah, now that there are mainline builds etc that people are testing..
[18:02] <tjaalton> closed over 30 "fixed upstream" bugs today, another batch waiting for tomorrow
[18:57] <bryceh> tjaalton, thanks for tending to those
[21:30] <tjaalton> bryceh: was a refreshing experience :)
[21:31] <bryceh> tjaalton, I think I can even see the dent it made - http://www.bryceharrington.org/X/Reports/ubuntu-x-swat/totals.svg
[21:32] <bryceh> or maybe not, that's a humongous graph :-)
[21:33] <tjaalton> yeah I've been cleaning some bugs along these past two weeks, at least the "remainder" is going down :)
[21:33] <bryceh> tjaalton, btw I'm sure I've shown you this already but in case not - http://www.bryceharrington.org/X/Reports/ubuntu-x-swat/totals-natty-workqueue.svg
[21:33] <bryceh> I ended up not really worrying about _total_ bug reports, but rather focusing on bug reports against the current development version
[21:33] <tjaalton> but speaking of natty, I should test wacom a bit. noticed that there's some crasher reported recently
[21:33] <tjaalton> yeah
[21:33] <bryceh> which I've found to be a lot more tractable
[21:34] <bryceh> and the 'workqueue' variant shows bugs that aren't already incomplete/upstreamed/fix-committed so it's just bugs we have to actually look at
[21:35] <bryceh> the lines are clickable too so you can go see the exact bugs (more or less)
[21:35] <tjaalton> yep, nice lists/graphs
[21:35] <bryceh> also did a tabular view of all bugs in the workqueue - http://www.bryceharrington.org/X/Reports/ubuntu-x-swat/workqueue-natty.html
[21:36] <bryceh> anyway, I'm probably the only person in the world that geeks out this much over bug graphs/lists.  ;-)  But I think it's been really useful for keeping on top of bug reports this release.
[21:37] <tjaalton> i've noticed, the speed of seeing bugs reported upstream (by you) is breathtaking ;)
[21:37] <tjaalton> um, bad language
[21:40] <bryceh> hi jaytaoko
[21:40] <bryceh> jaytaoko, btw, just fyi I've removed the glxinfo call from the apport hook.
[21:41] <bryceh> jaytaoko, I am guessing you guys don't need it for compiz/unity reports; in any case glxinfo has been moved to universe so not everyone has it installed, and so the apport hook gets errors in those cases.
[21:41] <bryceh> jaytaoko, if you do need any of the info that glxinfo produces, maybe you can roll it into the unity support test tool