[01:37] <bryce> tepsipakki: hmm, I solved the earlier mesa makefile problem.  Now just to sort out a linker error in xvfb...
[02:35] <bryce> hrm
[02:36] <bryce> well I'm out of brainpower on this one for the day.  Timo, if you care to poke at it, I've copied my current work to http://bryceharrington.org/Ubuntu/
[02:38] <bryce> it looks sort of like the stuff xvfb depends on was removed by patch 133
[02:38] <bryce> I haven't been able to find an explanation of why the GL/mesa/X/* stuff was dropped with this patch
[02:39] <bryce> I suspect perhaps in this case fedora did not care about xvfb so didn't notice the breakage?  not sure
[02:39] <jcristau> fedora probably updated the glx stuff to whatever it looks like in HEAD
[02:39] <jcristau> and only changed hw/xfree86/ to cope
[03:38] <bryce> whoohoo, I think I finally solved that gdm issue with the failsafe mode :-)
[03:38] <bryce> I decided to just kill off gdm after getting xinit up, and lo and behold, problem solved.
[06:25] <tepsipakki> jcristau: well, according to the changelog the problem was triggered by trying to install via parallels desktop on osx
[06:25] <tepsipakki> and it was marked as "RH specific", so I thought it wasn't worth it
[07:40] <bryce> bedtime for me.  g'nite!
[10:36] <mvo> joy! x does not even start anymore now with nvidia+comiz with the latest xserver-xorg-core
[10:37] <tepsipakki> finally :)
[10:37] <tepsipakki> then we can mark all those bugs as invalid :)
[10:38] <mvo> and file a request-for-removal for nvidia-glx-*
[10:38] <tepsipakki> yeah
[10:40] <tepsipakki> mvo: does the logfile show anything useful?
[10:41] <mvo> tepsipakki: it seems to die when LIBGL_ALWAYS_INDIRECT=1 glxinfo is run
[10:41] <mvo> then it segfaults
[10:42] <tepsipakki> the server?
[10:42] <mvo> somewhere in XMEsaDestroyVidual
[10:42] <mvo> yes
[10:43] <tepsipakki> hmm, 12ubuntu3 has five more patches than the WIP 12ubuntu5
[10:43] <tepsipakki> backported from 1.3.99
[10:43] <mvo> http://paste.ubuntu.com/49/http://paste.ubuntu.com/49/
[10:43] <mvo> http://paste.ubuntu.com/49/
[10:44] <tepsipakki> oh, they were superseded by the fedora mesa/exa-patches
[10:44] <mvo> I would really like to test a xserver-core with minimal patching, I guess you do not have a idea what patches are absolutely required to build it?
[10:44] <mvo> and which are "just" fixes/tweaks etc
[10:45] <mvo> I want to avoid that nvidia comes back to us and says: if you take out your patches, all works fine :)
[10:46] <tepsipakki> I'm not sure if any of those are absolutely required
[10:46] <mvo> I tried without any and it does not build (fails somewhere in mesa)
[10:46] <tepsipakki> ah
[10:46] <mvo> then I tried to apply some that looked like mesa patches and tried again and it failed at a different spot
[10:47] <mvo> then I gave up for the day
[10:47] <tepsipakki> right, you need patches 125-7
[10:48] <tepsipakki> at least
[10:49] <mvo> thanks
[10:51] <tepsipakki> although to make the minimal xorg.conf work you need also patches 05, 94, 119
[10:56] <mvo> aha, the login problem seems to be just on my system, that is good
[10:58] <mvo> now the build dies in rensize.c and complains about missing GL_DEPTH_STENCIL_MESA
[11:21] <ubotu> New bug: #137465 in xorg (main) "control alt backspace reboots pc sometimes" [Undecided,New]  https://launchpad.net/bugs/137465
[12:08] <mvo> holly cow, after a lot of struggling I made xserver-xorg work nearly without patches and compiz+glxgears does *not* crash
[12:08] <mvo> its freaking slow though
[12:14] <tepsipakki> yep
[12:15] <mvo> ^--- bryce (when you wake up :)
[12:15] <mvo> tepsipakki: I guess it will not be easy to figure *what* exactly makes it break
[12:15] <tepsipakki> that would be patches 107, 110, 120
[12:15] <tepsipakki> to make it speedier
[12:15] <tepsipakki> right..
[01:13] <ubotu> New bug: #137376 in xorg (main) "[gutsy]  Logging into Gnome makes xorg restart" [Undecided,New]  https://launchpad.net/bugs/137376
[01:45] <jcristau> tepsipakki: i think it's the same as http://gitweb.freedesktop.org/?p=xorg/xserver.git;a=commitdiff;h=1afdf8b0a92437dffe84fa98b6083b3d8fd55e27
[04:06] <ubotu> New bug: #137515 in xorg (main) "[gutsy]  I upgrade from Test Tribe 4 to Test Tribe 5 and my monitor get in stand-by" [Undecided,New]  https://launchpad.net/bugs/137515
[05:35] <bryce> morning
[06:04] <mvo> hey bryce! bad news (or good ones, depending on the POV). it seems like our patches are responsible for the nvidia+compiz+glx breakage
[06:05] <bryce> yup, I saw the discussion
[06:06] <bryce> mvo, my guess would be since you also report performance reduction, that in this process some accelerator got disabled, and it's in that that the bug is
[06:06] <bryce> like maybe it's switched to software rendering or something
[06:07] <mvo> don't put to much into my performance reduction thing, its the first time that reidrected rending works, so it may well be caused by redirected direct rending and not by some missing patch 
[06:07] <mvo> (works for me on nvidia that is)
[06:07] <mvo> its much faster than software rendinging, I don't think that is the reason
[06:08] <mvo> but it does sometime feel like it takes a 1sec break from time to time
[06:09] <bryce> hmm
[06:39] <mvo> the sky is falling: http://lwn.net/Articles/248227/
[06:50] <bryce> hehe
[06:50] <ubotu> New bug: #137561 in xserver-xorg-video-mga (main) "Ubuntu Feisty & Gutsy : mga driver can't display video with 2d acceleration (xv)" [Undecided,New]  https://launchpad.net/bugs/137561
[06:54] <tepsipakki> mvo: try enabling patch 120, that should boost performance.. then you'd also see if it breaks the blob
[06:55] <mvo> tepsipakki: I will try to find time for this today, its not easy as there are some other tribe-6 bugs assigned to me :/
[06:55] <ubotu> New bug: #137563 in xorg (main) "[gutsy]  [regression]  After suspend-to-ram, the pen on a toshiba tabled pc does not work" [Undecided,New]  https://launchpad.net/bugs/137563
[09:55] <ubotu> New bug: #137604 in xorg (main) "Black Bar Across Screen with latest Xorg Update" [Undecided,New]  https://launchpad.net/bugs/137604
[11:25] <ubotu> New bug: #137522 in ubuntu "[Gutsy]  Incorrent refresh rates with nvidia driver by default (Twinview issue) (dup-of: 92599)" [Undecided,Incomplete]  https://launchpad.net/bugs/137522
[11:45] <ubotu> New bug: #137471 in crack-attack (universe) "Crack attack crashes my session (dup-of: 130325)" [Undecided,Invalid]  https://launchpad.net/bugs/137471
[11:45] <ubotu> New bug: #137626 in xserver-xorg-video-ati "6.7.192 resolution not correctly detected" [Undecided,New]  https://launchpad.net/bugs/137626
[11:45] <ubotu> New bug: #137627 in xserver-xorg-video-ati "6.7.192 movie output only on the primary screen" [Undecided,New]  https://launchpad.net/bugs/137627