[02:59] mlankhorst: I've been looking at your -ati merge; there seem to be a couple of problems in it. Firstly, it looks like you've dropped some changelog entries (at least) - the archive currently has 1:6.14.99~really6.14.4-0ubuntu1, but that's nowhere in debian/changelog. [03:00] Secondly, the version number is unhelpful; it should be 6.14.99~really6.14.4 again, as (as far as I can tell), that's what it's based on. [05:50] RAOF: ugh no idea how changelog entries got lost though, fortunately there will be a 7.0 soon so maybe we could sync versioning then again properly [05:50] HURRAY! [06:02] mlankhorst: assuming there will be a 7.0 before we release 12.10 when theres a month before f18 might be crazy :) [06:03] f18? [06:07] Fedora [06:08] 1.5 months tops until we can get libdrm updated and can just do a new git snapshot at least [06:09] (mesa release schedule) [06:10] i'm sure there will be a rc1 on mesa 8.1 at the beginning of august so even sooner [06:11] anholt commited a ton of automakification late last week, hope its buildable out of tree now [06:12] havent been able to make debian packages for over a month now because of how we build multiple versions of it at once in the source package :( [06:18] theres UMS fixes post 6.14.5 so definitely will be another UMS release? [06:27] Oh, dear. How much do I care that xf86-video-msm FTBFS? [06:28] err [06:28] thats the one that was a drop from 2 years ago wasnt it? [06:29] Something like that. [06:29] Grr! What moron builds a tarball with -Werror?! [06:30] might be patches to fix it in collabora git somewhere [06:31] i dont even see that on git.debian.org even though i remember fixing it up to build against xserver 1.7 years ago [06:32] I guess I should test-build before uploading to quantal-proposed... [06:33] Which means it's time for everyone's favourite! Xserver builds in qemu-user-static! Yay! [06:34] 1.5 hour break while it builds? woohoo [06:34] thats how long it takes on an i7-2600 :) [06:35] Let's see how well the laptop ivb manages it. [06:36] ya upgrade? [06:37] No; that's the SDP I got. [06:37] oh [06:47] RAOF: why qemu? pandaboard? :p [06:47] mlankhorst: I haven't finished setting that up; I only just got an .au power cable for it today :) [06:47] oh [06:48] the adapator seemed to be pretty standard though, i have at least 10 european power cables for that lying around here somewhere [06:48] Yeah, it's standard. I just didn't happen to have a cable free. [06:49] I'm in a continual state of cablelesness; I went out and bought two jug plugs and two flat figure 8 cables today. [06:49] NOW I HAVE THE POWER (to power things) [06:50] the power to empower [06:51] suppose I should get off this pandaboard, wake up and get some work done. brb :) [06:52] RAOF: you got a panda too? [06:52] crazy the UE are all getting free pandaboards for some reason :) [06:53] omapdrm might work in quantal on it! [06:55] whenever libdrm and xf86-video-omap get synced [06:56] Yup. The whole desktop team got pandads, I think. [06:57] It will be *marvellous* to have an actual armhf device to build/test on. [06:57] On the other hand, I can no longer disclaim responsibility for ARM stuff ;) [06:58] libdrm-omap1. Why, hello! [07:00] got to manually build http://anonscm.debian.org/gitweb/?p=collab-maint/xf86-video-omap.git;a=summary against it [07:01] Ok. I think everything but -msm is staged in quantal-precise now. Please feel free to check the quantal section of http://people.canonical.com/~ubuntu-archive/pending-sru.html to see if I've missed anything. [07:01] Ah. A non-framebuffer omap driver? Cool. [07:10] is that driver open source? [07:27] RAOF: Interesting, I remembered to check for .gov.au before applying for visa, else some site would have gotten my personal details :) [07:32] scum sites :s [07:33] Odd! [07:40] RAOF: was thinking of going on vacation there in august :) [07:40] Good plan; we're awesome! [07:40] ☺ [07:41] Although perhaps a little bit cold, depending on where you go. And not as cold as frigid Europe, anyway :) [07:41] as long as I go north it's fine, you probably call 15°c cold :p [07:43] Well, it got up to 6°C here today. [07:44] sometimes gets -20°c here before sun comes up if unlucky during winter [07:44] It does get genuinely cold sometimes. Not icy European cold, but sufficiently cold to rug up and say “ooh, that's nippy” [07:49] Woo! Xserver almost finished. [08:05] A mere 54 minutes! [11:07] RAOF: hehe >:D === yofel_ is now known as yofel [14:40] hiya [14:40] a friend of mine ran into https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/1017243 - is there anything else he should be adding to the report? [14:40] Launchpad bug 1017243 in Mesa "Program received signal SIGBUS, Bus error." [Medium,Confirmed] [15:03] hey [15:06] dholbach: sounds like a masive map though, if you shrink the file does it work? :) [15:07] mlankhorst, maybe you can ask on the bug report? I personally don't have the problem - a friend of mine ran into it :) [15:08] you asked what to add, but ill have to brb [15:08] making food [15:09] ah, enjoy :) [15:10] mlankhorst, ok, I passed on the question and it seems to work with small files like /usr/share/icons [15:10] but here the man is: toabctl :) [15:11] toabctl, seems like mlankhorst just went to prepare some food [15:11] mlankhorst, my files are about 3 mb large. after 3 files, i get the error [15:12] it's reproducable with the attached python test program [16:24] mlankhorst, i'll leave. i'll be back tomorrow if you need more infos. [16:36] RAOF, no missing changelogs, I just corrected mlankhorst's version number before uploading it, just wasn't pushed to git. [16:42] oh back [16:45] Yikes lost train of thought. :( [17:55] bryceh: [17:56] mlankhorst, yep [17:56] woops [17:56] ?? [17:56] http://people.canonical.com/~mlankhorst/prime-wip.patch can you poke a hole in it? I think it sucks.. [17:57] but it's ok for a first attempt, I suppose. [17:57] sure [17:59] mlankhorst, in the dma-buf.c routines, I take it null pointer checks would be superfluous? [18:00] not yet, sadly [18:01] I think it's more something about general api at this point [18:02] gotcha [18:02] so i was thinking why not export the sync bo as dma-buf too [18:02] (nvbo should also maybe be checked before deref) [18:03] oh no that's all fine for hack [18:03] I just care about generic api at this point [18:07] well, one thing I'm wondering with the api, I notice you're adding another parameter to a bunch of functions that I gather is a function of the number of bufs. Is that really variable, or could it be a property of the fence? [18:08] i.e., instead of passing as a parameter to the routines, could it be added to the nouveau_fence struct? [18:09] bryceh: well i need to know in advance how much room is available [18:10] mlankhorst, ok, well other than that, looks good to me. Does it work? [18:10] not yet [18:10] i needed today to work out some details [18:10] started out entirely different [22:16] any packages we need to get in before the alpha-2 window closes? [22:23] bryceh: I think we have the important ones now so most graphics corruption is gone :) [22:34] bryceh: We *could* copy xserver 1.12 from -proposed to quantal; everything but an obscure arm driver is ready. Although just-pre-alpha-2 might not be the *ideal* time to do that... ☺ [22:34] which one btw? [22:37] xf86-video-msm [22:38] It's for the Galaxy Nexus. Or Nexus One, or some other Nexus phone. [22:41] RAOF, if we were definitely going to be shipping 1.12, I might argue having it in A2 would be beneficial [22:41] but given that we're hoping to get 1.13 in, seems like no reason to hurry. [22:41] btw with 1.12 are we going to temporarily lose any prop drivers? or have they all been updated already? [22:42] We don't lose nvidia, and I don't think we lose fglrx. [22:45] legacy drivers? [22:53] Not sure.