[05:37] <Sarvatt> hmm, autodetection routine for sis in xserver with no xorg.conf seems a little screwy
[05:37] <bryce> where'd you get sis hw?
[05:38] <Sarvatt> http://sarvatt.com/downloads/sis.txt
[05:38] <Sarvatt> creates 2 identical sis devices?
[05:38] <Sarvatt> dug up an old server :D
[05:38] <bryce> ah heh
[05:38] <bryce> yeah that does sound improper
[05:39] <Sarvatt> maybe thats because theres pci.ids
[05:41]  * bryce rolls eyes at lp #401107
[05:42] <Sarvatt> ahh yep it is, deleted the sis.ids and it only added the 1
[05:43] <Sarvatt> sheesh!
[05:46] <Sarvatt> wheres the "wont be fixed in the next 3 years" bug status :D
[05:46] <bryce> no kidding
[05:47] <bryce> >maybe< running as non-root will come for -intel soonish, but when you think of all the other random drivers that aren't likely to get KMS any time soon...
[05:53] <Sarvatt> yeah sounds like he read an article on phoronix about moblin which is only targetting intel graphics so they can do it :D
[06:37] <bryce> sweet http://www.ubuntu.com/news/canonical-open-sources-launchpad
[07:28] <Sarvatt> nice!
[09:04] <tjaalton> wow, even soyuz is open
[09:43] <Ng> ooh, new intel X love
[09:43]  * Ng does the crack-of-the-day dance
[09:45] <bryce> :-)
[09:46] <bryce> sorry, not very much changed, just a handful of bug fixes
[09:46] <bryce> but at least now we're officially at 2.8.0 in time for alpha-3
[09:46] <bryce> the mesa 7.5 update should be moderately more interesting
[10:10] <Ng> bryce: bug fixes is fine, and still worth testing :)
[12:16] <Company> hey
[12:16] <Company> 2 questions:
[12:16] <Company> 1) i updated to the edgers repo and it didn't pull the new kernel
[12:16] <Company> so i ran with the default jaunty kernel
[12:17] <Company> and that didn't go very well - read: it quickly hardlocked
[12:17] <Company> should i file bugs about that or should maybe the intel drivers just pull a newer kernel?
[12:18] <Company> 2) is there a way to get debug symbols for the edgers ppa?
[12:19] <Company> i'm hacking on pixman atm, would like to debug its usage inside X and am too lazy to compile it myself but gdb without debug symbols kinda sucks
[12:19] <jcristau> hacking on it but too lazy to build it yourself?
[12:19] <jcristau> that kinda doesn't make sense :)
[12:23] <Company> i'm hacking on pixman, not X :p
[12:23] <jcristau> ah, ok :)
[12:24] <tjaalton> 1) not a bug, 2) no idea
[12:25] <Sarvatt> sudo apt-get install linux-image-2.6.30-10 linux-headers-2.6.30-10-generic linux-headers-2.6.30-10
[12:25] <Sarvatt> everything with debug symbols in ubuntu has debug symbols in the PPA
[12:26] <Company> i know how to solve the kernel problem - i just wonder if it should be done by default instead of leaving me with a kinda broken system
[12:27] <Sarvatt> no it shouldnt
[12:27] <Sarvatt> its optional
[12:27] <Company> not really, because it doesn't work
[12:35] <Sarvatt> you can pastebin your logs from when you locked up and we might be able to help, it doesnt work and it locked up doesnt really say much
[12:36] <Sarvatt> if you can pastebin them we might be able to help rather
[12:39] <Company> i'd rather not debug bleeding edge X on outdated kernels
[12:39] <Company> there's more useful things to do
[12:40] <Sarvatt> the PPA isnt just for intel, forcing the kernel and breaking jaunty packages that people might be using for everyone just because it makes intel better isnt the best idea
[12:41] <Company> yeah, but you could make intel depend on the kernel
[12:44] <Sarvatt> if it was required sure, but it runs fine with 2.6.28 on all of the machines i've tested it on. do you have a /var/log/Xorg.0.log.old showing where it didnt work maybe?
[12:47] <Company> nope, restarted X a bit too often since then
[12:49] <Sarvatt> can ya paste a trace of an example of what you need symbols for so I can help ya find which packages you want  to install?
[12:51] <Sarvatt> libpixman-1-0-dbg xserver-xorg-core-dbg and libgl1-mesa-glx-dbg most likely at least
[12:52] <Company> oh, -dbg
[12:52] <Company> i was looking for -dbgsym packages
[14:26] <Sarvatt> hmm, thinking about putting an updated pixman on xorg-edgers, there was a post asking for testing of the fix for server 1.6 branch that we might be using and i have a feeling people will be wanting to upgrade pixman before karmic release for the arm optimizations
[15:31] <Sarvatt> jbarnes: any chance the fifo watermark kernel fixes might make it in time for 2.6.31-rc4? :D
[16:52] <jbarnes> Sarvatt: yeah hopefully
[16:59] <Sarvatt> is it like, just people with dual channel memory having problems? everyone with flickering i've talked to is running dual channel but its probably just coincidental
[17:01] <jbarnes> Sarvatt: no the patch that went upstream just has a lot of bugs
[17:01] <jbarnes> I think it'll affect a lot of people
[17:24] <apw> jbarnes, that sounds bad, we got a fix we can apply separatly until it gets in?
[17:24] <apw> bryce, how are we doiing with radeon userspace support?
[17:24] <apw> (for kms)
[17:25] <bryce> heya apw
[17:25] <apw> morning
[17:25] <jbarnes> apw: yeah there are a couple of patches you can apply until upstream gets the bits
[17:25]  * jbarnes digs up urls
[17:26] <bryce> in fact was just chatting with sarvatt about that last night.  Seems it is going to require the radeon-rewrite branches of mesa, et al.  After we get alpha-3 out I think we'll bite the bullet and pull those development trees in.
[17:26] <jbarnes> http://lists.freedesktop.org/archives/intel-gfx/2009-July/003471.html
[17:26] <jbarnes> http://lists.freedesktop.org/archives/intel-gfx/2009-July/003472.html
[17:27] <jbarnes> https://bugs.freedesktop.org/show_bug.cgi?id=19304 - last comment has an untested fix for 845G
[17:27] <ubott2> Freedesktop bug 19304 in Driver/intel "[845G] FIFO underruns" [Major,Reopened]
[17:27] <ubott2> Freedesktop bug 19304 in Driver/intel "[845G] FIFO underruns" [Major,Reopened] http://bugzilla.freedesktop.org/show_bug.cgi?id=19304
[17:27] <Sarvatt> funny enough all the people i've come across with the flickering problems are on 945
[17:27] <Sarvatt> 6 people in #ubuntu+1
[17:27] <jbarnes> http://lists.freedesktop.org/archives/intel-gfx/2009-July/003471.html is probably the fix for them
[17:28] <Sarvatt> thanks for the link, there were so many patches flying around i didnt know which to pass along :D
[17:29] <apw> bryce, sounds like stormy waters ahead
[17:29] <apw> jbarnes, thanks, will put them on my list to check when we rebase next
[17:30] <jbarnes> oh those are in drm-intel-next now fwiw
[17:31] <bryce> apw, :-)
[17:32] <bryce> apw, actually the ATI guys are pretty conservative about when they pull in new stuff, so it's possible the code's already plenty stable, just not "perfect", so waters may merely be turgid
[17:36] <apw> heheh lets hope so
[17:41] <bryce> apw, oh btw, don't know if they've filtered through to you, but as we've been getting -intel bugs resolved upstream with proposed kernel patches, I've been moving them to the kernel queue and marking them xorg-needs-kernel-fix
[17:41] <apw> ahh no not noticed but th
[17:41] <apw> that sounds good.  will review those on the next rebase and see which ones are included
[17:42] <bryce> cool, yeah I suspect a lot can just be closed as already in the kernel
[17:42] <apw> i see there is a fair lump of fixes in mainline already
[17:42] <bryce> I think ogasawara mentioned she'd put together a report of these
[17:42] <apw> but linus' weeks seem to be rather longer than 7 days on average
[17:43] <bryce> heh, sometimes I feel the same way
[17:43] <apw> will ask her about it
[17:43] <apw> yeah i know waht you mean
[20:08] <MT-> What does xdmx offer?
[20:19] <MT-> figured it out
[20:59] <Sarvatt> hmm thats a new one
[20:59] <Sarvatt> dpkg-shlibdeps: warning: debian/xserver-xorg-core/usr/lib/xorg/modules/libexa.so contains an unresolvable reference to symbol xf86ProcessOptions: it's probably a plugin.
[20:59] <Sarvatt> dpkg-shlibdeps: warning: 94 other similar warnings have been skipped (use -v to see them all).
[21:20] <Sarvatt> if anyone else is using arm and wants to test out pixman with neon support -- http://sarvatt.com/downloads/armel/