[00:03] <bryceh> Sarvatt, I totally don't mind putting an ati card in it and having a screaming fast non-intel graphics system ;-)
[00:05] <bryceh> "The only people who care about Linux are elitist neckbeards who intentionally want to make things harder then they need to be."
[00:05] <bryceh> neckbeards?
[00:26] <bjsnider> bryceh, it looks like he tried maverick. maybe the problem is the kernel isn't new enough
[00:28] <bryceh> bjsnider, 3rd paragraph of http://www.phoronix.com/scan.php?page=news_item&px=ODk3MA itemizes what all is needed
[00:29] <bjsnider> but as charlie points out "...none were available short of building from source on our own. We do not feel this meets any reasonable standard for 'available'."
[00:31] <bryceh> 5th paragraph addresses that
[00:33] <bjsnider> yeah but that paragraph is simply saying "it's too hard" which is charlie's point.
[00:38] <bryceh> bjsnider, you don't think it's too hard?
[00:39] <bjsnider> that's an understatement. it's more than too hard
[00:39] <bjsnider> but charlie's saying if it's too hard then drivers cannot be said to be "available"
[00:41] <bryceh> ah
[00:41] <bryceh> heh, well that's Intel for you
[00:42] <bryceh> I'd share that rant...  I hate that they never backport fixes or hardware enablement, so it's a constant game of having to upgrade to latest kernel+libdrm+mesa+-intel+etc. or do messy backporting ourselves and then get accused of shipping frankenkernels or whatnot
[00:43] <bryceh> -ati is sooo much better about all of this
[00:43] <bjsnider> i wouldn't know about that
[00:43] <bjsnider> i'd say nvidia is so much better though
[00:43] <bryceh> even the proprietary drivers do a good job about being reasonably self-contained, although obviously the binary incompatibilities kill us here and there
[00:44] <bjsnider> it's so easy to build the blob even i can do it
[00:44] <bryceh> yep
[00:47] <bjsnider> charlie wouldn't have an argument if all he was being asked to do was install the latest nvidia driver. but if that was the case he wouldn't be complaining
[00:50] <bryceh> yeah.  hope it gets some attention to the issue
[02:11] <bryceh> P67:  http://www.amazon.com/Gigabyte-GA-P67A-UD7-Intel-P67-Atx/dp/B004G60AKO/ref=sr_1_1?ie=UTF8&s=electronics&qid=1294107053&sr=8-1
[02:23] <bryceh> Sarvatt, what if we had it collect dmesg.0 and dmesg.1.gz from /var/log?
[02:23] <bryceh> http://www2.bryceharrington.org:8080/X/Reports/ubuntu-x-swat/totals-natty-workqueue.svg - down to 3 bugs again :-)
[02:40] <bryceh> Sarvatt, in the -ati package in edgers I see you disabled 101_select_between_classic_and_gallium_dri.patch - is that just because of a failure to patch or was there some other problem?
[12:41] <Sarvatt> bryceh: because it was using an older mesa base that didn't have the right dri driver names to work with that
[12:42] <Sarvatt> just fixed that with this mornings updates, thats what I meant about a mesa packaging update yesterday
[17:49] <dbarth> hiya
[17:50] <dbarth> i'm trying to get a nice backtrace for a dri/r300 issue
[17:50] <dbarth> i can't figure out which are the right debug packages to add
[17:51] <dbarth> previously i had libgl1-mesa-dri-dbg but that didn't contain the right symbols (ie no line numbers)
[17:51] <dbarth> and now i have libgl1-mesa-dri-dbgsym and that doesn't work either
[17:51] <dbarth> any idea?
[17:54] <Sarvatt> pastebin your incomplete backtrace? it'll still list the libs you need dbg packages for
[17:58] <dbarth> ah cool
[17:59] <dbarth> Sarvatt: it's here:
[17:59] <dbarth> http://pastebin.ubuntu.com/544957/
[18:00] <dbarth> i guess bryceh was mostly interested in r300_dri.so and libGL
[18:00] <dbarth> referring to https://bugs.launchpad.net/unity/+bug/691653
[18:00] <ubot4`> Launchpad bug 691653 in mesa (Ubuntu) (and 2 other projects) "compiz crashes when using alt-tab (the radeon driver kills it) (affects: 1) (heat: 6)" [Undecided,Incomplete]
[18:03] <cnd> bryceh, we're trying to plan xi 2.1 landing and development and it's getting more important for us to know when 1.10 will land in natty
[18:03] <cnd> do you know when that will be?
[18:03] <cnd> pre-alpha2?
[18:05] <Sarvatt> dbarth: i'm pretty sure that was fixed on mesa 7.9 branch already, i'm in the middle of another bug at the moment but will look asap
[18:07] <dbarth> Sarvatt: thanks; let me know as a bug comment maybe if i'm gone later
[18:08] <bryceh> heya dbarth
[18:10] <bryceh> cnd, hard to predict xorg's release schedule accurately, but I anticipate we'll get the rc candidate in within the next two weeks
[18:10] <bryceh> cnd, unless for some reason raof feels we should stick with 1.9.  I anticipate that decision to be made firm next Monday.
[18:10] <cnd> bryceh, so that's before alpha 2?
[18:11] <bryceh> cnd, right, alpha 2 is end of feb iirc
[18:11] <cnd> bryceh, alpha 2 is beginning of feb
[18:12] <bryceh> cnd, ah, ok anyway I anticipate we'll have it in by end of this month at the absolute latest
[18:12] <cnd> ok, cool
[18:12] <cnd> bryceh, who will be doing the patch merge for xorg-server?
[18:13] <cnd> it sounds like we may want to leave the maverick gesture patches in until alpha 2, and then drop them and pull in xi 2.1 right after alpha 2
[18:13] <bryceh> cnd, I'm assuming RAOF will since he likes doing that sort of thing
[18:14] <cnd> heh
[18:14] <bryceh> but really I think any of us could do it
[18:14] <cnd> bryceh, ok, once the decision is made on the server version I'll try to sync with RAOF about this
[18:14] <cnd> thanks!
[18:14] <bryceh> cnd,  no prob
[18:15] <bryceh> btw, sounds like RAOF is on vacation this week
[18:15] <cnd> ahh, good to know :)
[21:16] <LLStarks> bryceh, any plans for improving legacy intel for natty?
[21:16] <LLStarks> or still vesa hell?
[21:18] <bryceh> LLStarks, you mean 8xx?  think we've mostly written it off since it seems to be a moneypit of time
[21:19] <LLStarks> no shadowfb or whatever was in the pipeline?
[21:19] <LLStarks> i thought there was a decent solution that barely missed the cut-off window
[21:28] <LLStarks> unrelated question. is i915 gallium still dead?
[21:41] <bryceh> far as I know
[21:41] <bryceh> the intel guys haven't been very favorable towards gallium generally
[22:44] <Sarvatt> bryceh: btw check #intel-gfx :)
[22:44] <Sarvatt> thats some timing
[22:47] <Sarvatt> it looks like fedora 14 is hitting the same r300g problem dbarth is having in bug #691653, https://bugzilla.redhat.com/attachment.cgi?id=464327
[22:47] <ubot4`> Launchpad bug 691653 in mesa (Ubuntu) (and 2 other projects) "compiz crashes when using alt-tab (the radeon driver kills it) (affects: 1) (heat: 6)" [Undecided,Incomplete] https://launchpad.net/bugs/691653
[22:47] <Sarvatt> better backtrace there, thats from compiz 0.8.6
[23:05] <bryceh> magic registers for 8xx eh?
[23:09] <bryceh> Sarvatt, how'd you find that?  What's the bug # to go with that backtrace?
[23:09] <bjsnider> Sarvatt, do you know if randr support is coming to the blob in natty?
[23:10] <Sarvatt> googled radeon_r300_winsys_buffer_from_handle, ton of compiz crashes in red hat bugzilla
[23:10] <bryceh> Sarvatt, ahh
[23:10] <Sarvatt> none have any useful info in them, just a bunch of reports. i thought i saw something about that crash on 7.9 branch but cant find it if i did
[23:11] <Sarvatt> bjsnider: doubt it very much but dont know for sure
[23:12] <bryceh> Sarvatt, ok
[23:12] <bryceh> if one of the reports looks at all active, might be worth linking to ours
[23:13] <Sarvatt> none do :(
[23:13] <bryceh> Sarvatt, oh I see the problem
[23:13] <bryceh> #0  radeon_r300_winsys_buffer_from_handle (rws=<value optimized out>, whandle=0x7fffe69964b0, stride=0x7fffe69963f4, size=0x7fffe69963f0) at radeon_r300.c:123
[23:13] <bryceh>         ws = <value optimized out>
[23:13] <bryceh>         _buf = 0x0
[23:13] <Sarvatt> https://bugzilla.redhat.com/show_bug.cgi?id=656755 is the best one
[23:13] <ubot4`> bugzilla.redhat.com bug 656755 in compiz "[abrt] compiz-0.8.6-3.fc14: Process /usr/bin/compiz was killed by signal 11 (SIGSEGV)" [Medium,New]
[23:13]  * Sarvatt nods
[23:13] <bryceh>     if (size)
[23:13] <bryceh>         *size = _buf->base.size;
[23:13] <bryceh> simple null pointer deref
[23:15] <Sarvatt> bryceh: ok if i do a bit of tinkering in the wayland ppa?
[23:15] <Sarvatt> now that mesa in edgers is using natty packaging it can just be copied in there
[23:16] <Sarvatt> all the fixing you did seems to be in stuff specific to the old packaging where i had extra crap enabled
[23:16] <bryceh> Sarvatt, sure feel free
[23:17] <Sarvatt> might as well add wayland to edgers too I guess, its only 2 packages
[23:18] <Sarvatt> well 3 with cairo, need to figure out whats going on with alf's no glew stuff that went in upstream cairo recently
[23:19] <bryceh> yeah would love to see it pulled into xorg-edgers
[23:20] <bryceh> yeah the cairo bits are proving to be the main irritation.  Keeps changing versions and I have to update stuff  8-P
[23:21] <bryceh> oh btw I don't have the .so files packaged right for the libs.  Maybe you have more packaging-fu than me on how to install the .so.0.0.0 files?
[23:21] <Sarvatt> _alf probably backported it all to cairo 1.10 branch for linaro
[23:23] <Sarvatt> what libs?
[23:23] <bryceh> libxkbcommon
[23:23] <bryceh> and libwayland
[23:23] <bryceh> I think it just needs a symlink installed
[23:24] <bryceh> also I see these errors, which I've not sorted out yet but are on the todo list:
[23:24] <bryceh> E: wayland: no-shlibs-control-file usr/lib/libwayland-server.so.0.0.0
[23:24] <bryceh> E: wayland: no-shlibs-control-file usr/lib/libwayland-client.so.0.0.0
[23:24] <bryceh> E: wayland: postinst-must-call-ldconfig usr/lib/libwayland-server.so.0.0.0
[23:24] <bryceh> W: wayland: package-name-doesnt-match-sonames libwayland-client0 libwayland-server0
[23:24] <bryceh> E: wayland: missing-dependency-on-libc needed by ./usr/bin/wayland-compositor and 2 others
[23:24] <Sarvatt> whoa, its in the natty archives!?
[23:25] <jcristau> shared libs need to be in their own package
[23:25] <bryceh> libxkbcommon is, yeah... I uploaded it there a month or two ago
[23:25] <bryceh> jcristau, ah
[23:25] <bryceh> jcristau, I knew you'd know ;-)
[23:25] <jcristau> so one package for libwayland-client.so.0, one for libwayland-server.so.0, and one for the compositor.  at least.
[23:25] <jcristau> you need that so you can change the package name when the soname changes
[23:27] <bryceh> ok
[23:29] <Sarvatt> bryceh: https://launchpad.net/~afrantzis/+archive/cairo-gl
[23:29] <Sarvatt> alf__ rocks :)
[23:29]  * Sarvatt always gets his nick wrong since he doesn't hang out here much
[23:33] <bryceh> Sarvatt, hmm, looks strikingly similar to my cairo-gl package
[23:34] <bryceh> ooh, hide_glew_symbols.patch looks interesting
[23:34] <Sarvatt> yeah thats what i was pointing out, sorry
[23:37] <Sarvatt> he went one better and completely ripped glew out of cairo completely in git
[23:42] <Sarvatt> just noticed i'm from camaroon now, sweet! http://cgit.freedesktop.org/cairo/commit/?id=7a023a62f7517ad0d54f4d59c99909fadcc05e82
[23:43] <Sarvatt> so mesa 7.10 is for sure releasing on friday
[23:47] <bryceh> camaroon?
[23:47] <Sarvatt> gmail.cm :)
[23:50] <bryceh> ah
[23:55] <Sarvatt> got native ubuntu up on a tegra tablet I got last week, looks like I need to bring xserver 1.6 to natty for it since thats the only video abi nvidia supports with tegra :)
[23:56] <Sarvatt> arm ppas sure would be nice to have
[23:56] <jcristau> lucky you.
[23:57] <Sarvatt> i thought their video card blobs were bad, this tegra crap is in a world of its own