=== Bernardo|away is now known as Bernardo
[05:33] <Bernardo> hi
=== Bernardo is now known as Bernardo|away
=== Bernardo|away is now known as Bernardo
[09:44] <Bernardo> hi
[09:44] <lucazade> hi Bernardo
[09:45] <lucazade> looks like we're ready to fix :)
[09:46] <Bernardo> :) let's hope tseliot can help us.
[09:47] <lucazade> absolutely
[09:47] <tseliot> I'm investigating the issue and I have a device now
[09:47] <Bernardo> hi tseliot (or is it Alberto?)
[09:48] <tseliot> my real name is Alberto
[09:49] <Bernardo> you now have a machine to test?
[09:50] <tseliot> I have several but only one available for testing
[09:50] <Bernardo> ok, anyway, if you want us to test anything for you we're here
[09:51] <lucazade> yes
[09:51] <tseliot> do you experience corruption without my patch when shadowfb is disabled?
[09:51] <Bernardo> let me check
[09:51] <lucazade> no
[09:52] <lucazade> using it now
[09:52] <tseliot> can I see your xorg.conf, please?
[09:52] <tseliot> and the output of lsmod
[09:52] <Bernardo> You mean with "ShadowFB" "true"?
[09:52] <Bernardo> ok
[09:52] <lucazade> http://dl.dropbox.com/u/1338581/Gma500/xorg.conf
[09:53] <Bernardo> my lsmod - http://pastebin.com/xpQJGYTq 
[09:53] <Bernardo> my xorg.conf - http://pastebin.com/KhKvYX3v
[09:54] <tseliot> ah, you blacklisted vga16fb
[09:54] <Bernardo> yes
[09:54] <tseliot> why?
[09:55] <Bernardo> to be able to load psb in framebuffer mode
[09:55] <Bernardo> and I have no corruption with ShadowFB
[09:55] <lucazade> http://dl.dropbox.com/u/1338581/Gma500/lsmod
[09:55] <tseliot> ok
[09:56] <Bernardo> without blacklisting vga16fb I have a corrupted screen if I load psb without the no_fb=1 option
[09:56] <lucazade> some artefacts during plymouth right?
[09:57] <Bernardo> yes, no splash screen, only text mode
[09:57] <Bernardo> brb
[09:58] <tseliot> ok, I'll do it here too as I want to make sure that we have a similar environment
[09:58] <lucazade> i was able to see plymouth in graphic mode (using vga16fb) 
[09:58] <tseliot> me too but then ttys are busted
[09:59] <lucazade> ah true
[10:03] <Bernardo> we probably need to put libdrm-poulsbo in /lib to have a working plymouth
[10:05] <tseliot> Bernardo: how did you start X (with -core) when you gave me the core dumps? Was it with startx?
[10:05] <tseliot> err
[10:05] <tseliot> I mean from the binary
[10:05] <tseliot> /usr/bin/Xorg etc.
[10:06] <Bernardo> no, I edited /etc/kde4/kdmrc to have it pass the -core parameter to Xorg
[10:07] <Bernardo> if I start with startx, I always get the "SVDO is not trainded" error, and X won't start
[10:07] <tseliot> ah, you're using kde
[10:07] <tseliot> and the problem affects konsole, right?
[10:08] <Bernardo> yes
[10:08] <Bernardo> also firefox
[10:08] <tseliot> ok, let me install kde then
[10:09] <Bernardo> lucazade is using gnome, he can tell you what are the obvious symptoms there
[10:09] <lucazade> more or less the same happen on gnome, terminal can't scroll more 2 lines
[10:10] <lucazade> and i can't see any icon
[10:12] <lucazade> i'm able to start session with startx, don't have the issue of SVDO is not trainded
[10:12] <lucazade> or better it doesn't hang
[10:13] <lucazade> on gnome i had to remove compiz becuase it use some glx 1.3 features that hangs gnome 
[10:13] <lucazade> in that way was able to enable composite
[10:15] <Bernardo> bbl
=== Bernardo is now known as Bernardo|away
[10:22] <tseliot> lucazade: do you have the same problem if you use gdm and let it start x for you?
[10:23] <lucazade> yes the same
[10:38] <tseliot> ok
=== Bernardo|away is now known as Bernardo
[11:15] <lucazade> brb
=== Bernardo is now known as Bernardo|away
[12:15] <tseliot> RAOF: do you still have the source of your dkms packages for nouveau somewhere?
[12:28] <ricotz> tseliot, you can try to get it out here https://edge.launchpad.net/~xorg-edgers/+archive/ppa/+sourcepub/1074879/+listing-archive-extra
=== Bernardo|away is now known as Bernardo
[13:10] <tseliot> ricotz: thanks
[13:15] <Bernardo> hi
[13:42] <ricotz> tseliot, np
[14:05] <|Bernardo|> tseliot: you earlier had asked if I had bitmap corruption with your patch and "ShadowFB" on
=== |Bernardo| is now known as Bernardo
[14:05] <Bernardo> tseliot: I thought I hadn't as icons show without problems, but I just noticed the "loading" cursor for firefox still gets corrupted
[14:06] <tseliot> Bernardo: no, I meant without my patch and with shadowfb set to false
[14:07] <Bernardo> ok, without your patch and shadowfb false I have no corruption
[14:20] <tseliot> Bernardo: oh, I have corruption here
[14:21]  * bryceh waves
[17:39] <Sarvatt> jcristau: mind if I update upstream-experimental/debian-experimental branches of pixman with 0.18.2?
[17:40] <jcristau> Sarvatt: go ahead
[17:47] <Sarvatt> oh hmm huge diff in ChangeLog for some reason
[17:50] <lucazade> hi Bernardo
[17:50] <lucazade> any news?
[17:59] <Sarvatt> ok pushed it, hope I did everything right :) no new symbols or build changes 
[18:00] <jcristau> looks good to me
[18:01] <jcristau> now you can just dch -r, push that and i'll build/tag/upload :)
[18:05] <Sarvatt> "Prepare changelog for upload." ?
[18:05] <jcristau> or whatever you think is appropriate :)
[18:05] <jcristau> but that works
[18:10] <Sarvatt> done, sorry first time doing all that and I didn't want to screw anything up :) thanks for the help
[18:11] <jcristau> uploaded.  thank you. :)
[18:14] <Bernardo> hi
[18:15] <Bernardo> hi lucazade, no news, but I think that tseliot managed to reproduce the bitmap corruption
[18:15] <lucazade> ok thanks for the info
[18:58] <Sarvatt> jcristau: huh how did libdrm 2.4.20-2 actually build in experimental, nouveau symbols are messed up
[18:59] <Sarvatt> http://paste.ubuntu.com/433475/
[19:06] <jcristau> Sarvatt: e.g. nouveau_bo_pin comes from 03_revert_abi_change.diff
[19:07] <Sarvatt> ahh I'm an idiot, thats right :)
[19:07] <Sarvatt> was just merging and dropping that since we need libdrm that works with 2.6.34 in maverick
[19:08] <jcristau> i'm still pondering moving new libdrm to sid
[19:08] <jcristau> dunno whether execbuf2 is something we want for squeeze
[19:08] <jcristau> when there's a 2.6.34 kernel in experimental we'll probably put the new libdrm-nouveau and ddx there as well
[19:27] <Sarvatt> anyone around that can sponsor a libdrm upload? i've already pushed it to git, it's pretty urgent since we have a 2.6.34 kernel in maverick already - http://sarvatt.com/downloads/libdrm/
[19:41] <Sarvatt> intel has been in such a bad shape the past week, sheesh
[19:42] <Sarvatt> new quarterly release must be due soon
[19:45] <tjaalton> Sarvatt: shouldn't you be partying at UDS?-)
[19:46] <Sarvatt> I couldn't go!
[19:46] <tjaalton> aww..
[19:46] <tjaalton> sucks
[19:46] <tjaalton> same here
[19:48] <Sarvatt> yeah tell me about it, flight pushed back 3 times and I got hit by another car on the way to the last one, wife had to go to the hospital and I couldn't get a new flight that arrived before wednesday night. this was a sucky week
[19:48] <Bernardo> :(
[19:48] <Bernardo> Sorry to hear that, thought you were on this side of the atlantic now
[19:49] <Bernardo> I hope your wife is ok
[19:51] <tjaalton> yeah
[19:51] <tjaalton> ..
[19:52] <tjaalton> Sarvatt: anyway.. I can sponsor it later today if I'm awake ;)
[19:52] <tjaalton> libdrm that is
[19:52] <tjaalton> now tv ->
[21:08] <Sarvatt> no worries, xserver-xorg-video-nouveau needs to be updated too anyhow for it and i dont want to mess with that at the moment :)
[21:09] <Sarvatt> would just need a rebuild normally but i think we need http://cgit.freedesktop.org/nouveau/xf86-video-nouveau/commit/?id=e2146d3b29a4bea3d584c145e3190c3313692ed9
[21:11] <Sarvatt> someone complaining about the symbols being dropped without a soname bump on debian-x already, yay nouveau! :)
[21:39] <Sarvatt> does squeeze have any hope of being released before october? 
=== apachelogger is now known as electrolysis
[21:43] <jcristau> i don't know.  probably not :/
=== electrolysis is now known as PyObjectPtr
[22:46] <Sarvatt> just wondering if we'll be syncing from experimental the whole cycle, not sure if stuff in sid stops getting moved to squeeze at some point