=== yofel_ is now known as yofel === pzanoni_ is now known as pzanoni === chrisccoulson is now known as chrisccoulson2 === chrisccoulson2 is now known as chrisccoulson [19:51] bRoas [20:51] tjaalton: http://kernel.ubuntu.com/~sarvatt/patches/116_use_shared_galliumcore.diff actually does work with 8.0 branch, just need to fix up the mesa-egl crap [20:51] oh, cool [20:55] http://paste.ubuntu.com/823052/ should be the last problem that needs fixing [20:59] is this only on the ubuntu branch? [20:59] yup [20:59] ok, so no need to bug kibi about it :) [21:00] oh right, it's due to moving the libs around [21:01] maybe ya did something in the merge i'm glossing over and might remember it better :) [21:03] you mean i botched the merge? totally plausible [21:03] since this shouldn't be any different really [21:03] than what it was [21:04] oh the mesa-egl crap completely isnt in debian [21:05] no wonder it works there [21:05] haven't seen that before and thought we inherited it in the merge [21:09] well it's possible that I missed something from rules when merging from debian [21:10] rules merge looks fine to me looking at a399f4dab2de48673bf6da34dd10c4ff176dfbd3..HEAD :( [21:18] i bet it's dh_makeshlibs that's to blame [21:20] Sarvatt: check what's in the mesa-egl/ld.so.conf -file [21:22] tjaalton: wait, i just managed to fix it but dont remember what i did [21:22] too many context switches [21:22] haha [21:22] git diff! [21:23] reverting this hunk of debian/rules [21:23] - dh_shlibdeps -s -l/usr/lib/$(DEB_HOST_MULTIARCH)/mesa:\ [21:23] -/usr/lib/$(DEB_HOST_MULTIARCH)/mesa-egl [21:23] + dh_shlibdeps -s \ [21:23] + -l/usr/lib/$(DEB_HOST_MULTIARCH)/mesa:\ [21:23] + /usr/lib/$(DEB_HOST_MULTIARCH)/mesa-egl [21:23] uh oh [21:23] yay she built [21:23] * Sarvatt does a happy dance [21:24] of course [21:24] stupid me [21:24] -rw-r--r-- 1 sarvatt sarvatt 4.5M Jan 30 21:19 libgl1-mesa-dri_8.0.0+git20120126+8.0.19f88670-0ubuntu0sarvatt_i386.deb [21:24] it grew almost 2MB though [21:24] and all that for making it a tad nicer looking :) [21:24] still better than the 27MB one without dricore and galliumcore [21:24] i'll fix it in git [21:25] cool [21:28] tjaalton: so mesa's good to go after the merge :) [21:28] eeexcellent [21:29] merge now or wait for vdpau stuff to go in ya think? [21:29] * Sarvatt would prefer not bothering with vdpau because mesa's build system is already so freaking fragile but oh well [21:29] +1 [21:30] does any app actually use these shiny video apis? [21:30] yeah maybe postpone it for now [21:30] sure does, but dunno how the gallium drivers work [21:31] at least nvidia blob vdpau is widely used (on my PVR for instance ;) [21:31] i have no clue either [21:32] (if it even works) [22:04] tjaalton: oh yeah, you said you had a problem with xa on amd64? i'm building on i386 so didnt notice that [22:05] maybe I didn't, can't remember anymore [22:05] *did [22:09] I we've had some inquiries about whether or not we support vdpau/vaapi; while there's not *that* much that uses it right now, there does seem to be a reasonably complete gstreamer element with support. [22:14] RAOF: you use ati at all to test it? [22:16] oh actually I have an ati in a desktop I run headless, can test it out there [22:33] I do have an ATI to test it; I need to spin that machine up anyway, to update it. It'll take *some time* [22:34] ah no worries, vdpau stuff isn't in git [22:34] What cards does it run on, anyway? [22:45] doing the wayland merge [22:45] RAOF: I dunno at all [22:46] :) [22:46] will look after this, imagine just radeon/nouveau [22:46] r300+ [22:46] Oh, I mean: what chip generations. Good, my 5450 should cover that :). [22:48] should the wayland update wait until after A2 too? [22:48] the wiki says soft freeze in effect today but haven't seen any announcements [22:48] well yesterday for RAOF :P [22:49] wayland is not on the images, so it shouldn't matter [22:49] what about libxcb? [22:49] it surely is :) [22:49] The archive still seems to be open; go for it. [23:02] 1177 root root 1 0.11s 1.34s 172K 0K 0K 0K -- - R 1 75% Xorg [23:02] this isn't normal :S [23:03] tjaalton: Hey, I heard you wanted a Workstation Key? [23:04] Anybody else want one (or Fusion for that matter)? [23:05] Prf_Jakob: well, I could test the new stuff yeah [23:05] Ok, I'll fix one for you. [23:05] I only have precise to run it on though [23:05] which hw/driver? [23:06] intel [23:06] sandybridge [23:06] 7.10, 7.11 or 8.0? [23:07] oh mesa? 7.11 for the time being, hopefully 8.0 tomorrow [23:07] I actually had a reports of 7.10 being slow on sandybridge, 7.11 and 8.0 should be fine. [23:07] Oh you are probably good. [23:08] as long as the kernel module builds on 3.2 ;) [23:08] yeah thats typically the problem I run into, hunting down kernel patches :) [23:09] patches so the guest additions build against newer kernels i mean [23:11] Running w/o tools work fine, if I have problems its with the host kernel modules Workstation wants to build. [23:11] I could do with a key, if you'd like me to test stuff :) [23:13] Just WS or does anybody want a Fusion key as well? [23:13] Then again I guess nobody here runs Mac on thier computers ;) [23:13] * Sarvatt paid for fusion, its actually decently priced :) [23:14] Yeah [23:14] WS is a wee bit expensive. [23:21] thank you for offering though Prf_Jakob [23:23] bryce, RAOF, tjaalton: which ppa should I shove new wayland and mesa in? [23:23] x-updates? [23:23] RAOF: would it make any sense if that GPU hang i ran into was connected to the weird hsync/vsync issue with GnomeRR? i seem to only see it if i use GnomeRR to switch from internal-display-only to external-display-only [23:24] broder: It would indeed make sense; GnomeRR is a wee bit crazy about that. In particular, it'll switch the CRTC from internal to external, whereas the xrandr tool generally won't. [23:25] Sarvatt: x-staging imo [23:25] It's entirely plausible that GnomeRR exposes a bug that you would be difficult to trigger in any other way. [23:25] and 7.11.2 in x-updates? [23:26] oh 7.11.2 in x-updates for oneiric would indeed be a good idea [23:26] Sarvatt: I concur; we might as well continue using x-staging. [23:26] alrighty! [23:26] I'd *like* 7.11.2 to be suitable for oneiric-updates; I don't know that it is, though. :/ [23:27] Sarvatt, x-staging sounds good to me too [23:27] * Sarvatt is seriously sick of looking at mesa for tonight so task for tomorrow or something :) [23:28] Sarvatt: I can shove stuff somewhere too [23:28] also, wayland should be good to push straight to precise, right? [23:28] still have to adapt all this crap to mesa master to get xorg-edgers building again, major changes in master after automakification [23:29] well the merge is pushed to git, it needs a changelog still [23:29] we have a pretty huge diff in debian/, this stuff needs to be pushed to debian [23:32] multiarch stuff yep [23:35] ok got http://kernel.ubuntu.com/~sarvatt/wayland/ and pushed it to git, totally untested until i eat dinner :) [23:35] hum, wonder why we've had to add the COPYING.GPL-2 there [23:44] tjaalton: wayland isn't in the images? huh?! [23:44] oh [23:44] forgot edgers is "special" due to newer cairo [23:45] wayland-demos will need an update too, rrgh [23:45] libwayland0 (= 0.1.0~0.2-0ubuntu3) [23:46] that needs some drastic changes and it's in universe though [23:46] with the weston rename [23:49] are the cd images going to be in crisis mode if they grow 2MB? [23:50] thats how much bigger libgl1-mesa-dri is compressed in 8.0 [23:50] We can see. [23:50] :) [23:50] Failing anything else, I believe we've got the leeway to grow to 750MB. [23:51] i'll see if any of the other packages increased too [23:51] http://paste.ubuntu.com/823262/ [23:52] thats rc2, i just hadnt bumped the changelog yet [23:53] 30-Jan-2012 08:28 712M Desktop CD for 64-bit PC (AMD64) computers (standard download) [23:53] oh cool [23:53] Yeah; we're already oversize :) [23:53] maybe they should stop calling it "CD" on the download page [23:54] and saying its a bug that its oversized :P [23:54] * Sarvatt installs newer mesa to his wife's laptop [23:55] that's always a good stress test [23:55] lots of reports about nvidia-current being hosed in current precise btw [23:57] although two of the people hitting it were hit by busted unity-greeter in the end and blaming nvidia [23:58] /X [23:58] https://bugs.launchpad.net/ubuntu/+source/unity-greeter/+bug/921998 is pretty nasty [23:58] Launchpad bug 921998 in unity-greeter (Ubuntu Precise) (and 1 other project) "unity-greeter crashed with SIGSEGV in indicator_object_get_entries() (affects: 23) (dups: 7) (heat: 124)" [High,Confirmed]