=== Azelphur_ is now known as Azelphur [00:37] Sarvatt: Nah, the -nr patch itself wasn't the issue. The issue was updating -intel's patch for 2.12+legacy. [00:41] Sarvatt: Oh, and does unity crash the X server for you (in mesa, dri2 flush paths) with 1.9 + intel 2.12? [00:47] do i have to run a unity session to crash it? [00:47] i just ran unity and got a crapload of errors and the bar is blank, first time running it [00:48] it didnt crash though [00:49] RAOF: sorry, shoulda pinged you about that :) [00:50] You might. [00:50] I haven't yet done much investigation. [00:50] how did you crash it? [00:50] lemme try a session, need to reboot for this new kernel anyway [00:50] Run 1.9 + 2.12+legacy, from gdm select UNE session. [00:51] Boom! X crash, mesa, invalidate buffers. [00:51] http://paste.ubuntu.com/463747/ [00:51] got your 2.12+legacy packaged anywhere? [00:51] brb rebooting into unity [00:52] In raof/aubergine, yeah. [01:00] no crash [01:00] but it doesn't work [01:00] it's in a little 640x480 or so area inside my real resolution [01:01] and you cant see anything but a few buttons at that size [01:03] Hm. Might be +leagacy breakage, then. [01:03] or even legacy. [01:06] Legacy is sitting in the raof/aubergine PPA if you want to try it. [01:06] You'll need to rebuild against 1.9 obviously. [01:10] got a backtrace on the crash by any chance? [01:10] In fact, you'll need to build locally *anyway*, as the PPA hasn't yet built. [01:10] even just the xorg.0.log one [01:11] http://paste.ubuntu.com/463755/ [01:11] Is not very good. [01:11] are you using UMS? [01:11] No, not then. [01:15] Wait a sec, I'll get a proper backtrace. [01:18] http://paste.ubuntu.com/463759/ is a better one. [01:21] looks like mesa to me [01:23] does it happen with this in ~/.drirc? http://paste.ubuntu.com/463762/ [01:26] doubt i'll be able to reproduce it, no 965's handy :( [01:29] The problem is that drawable->driContextPriv->driverPrivate is NULL at that point. [01:30] Hah.\ [01:30] Well, that theory's nice and easily confirmed. [01:30] Intel swap events be broken! [01:31] Quadrapassel does the same thing. [01:33] anything non clutter crash it? :) [01:34] Not so far :) [01:37] CLUTTER_DEBUG=disable-swap-events work? [01:37] clutter has *so many* debugging env variables [01:37] I'm just trying with 2.12 from experimental. [01:37] tried edgers already? [01:37] Ba baw. [01:37] No, haven't tried edgers. [01:38] Nope. CLUTTER_DEBUG=disable-swap-events quadrapassel aint a winner. [01:40] My guess is that edgers will work, because it'll have a newer mesa built. [01:40] ahhhhh ok i thought you were using edgers + that intel [01:41] Nah. [01:41] might want to try dropping 102-disable-page-flipping-v2.patch too for the heck of it to be closer to whats in git [01:42] Tried that; bug's still there. [01:43] Well, rather, bug where the system locks after some period of time is still there; dunno about the clutter crash :) [01:46] do you have a compiled mesa demos handy by any chance? [01:47] or uncleaned old 7.8 mesa source tree? [01:47] can ya see if glxgears_fbconfig crashes you too? :) [01:50] oh man i miss my irc bouncer, brb again legacy is done compiling [01:59] I can't even start X with legacy on 945 [01:59] it just silently dies after initializing HW cursor [01:59] only tried KMS though [02:01] Have you checked the gdm log? There's a good chance it's dying with a missing symbol, which would be written to the gdm log, not the xorg log. [02:01] Also, that's rather annoying. 'Twas working here on 965! [02:01] oh duh [02:01] X: ../../dix/pixmap.c:118: AllocatePixmap: Assertion `pScreen->totalPixmapSize > 0' failed. [02:01] Oh, crud. [02:02] Did I not push the refreshed nr patch? [02:02] nr patch was disabled [02:03] On your X server, or 101_copy_fb in the DDX? [02:03] ddx [02:10] i think intel legacy is the longest ddx compile now [02:12] It builds 3 drivers + an acceleration architecture, cut it some slack :) [02:13] RAOF: yep that was it, can't reproduce the unity/quadrapassel crash on though on edgers with 945 :( [02:14] now to break things! [02:14] Well, edgers is now installing, so I can check soon. [02:16] what mesa were you using again? 7.8.2 or the maverick one? [02:16] The maverick one. [02:21] is it me or is mesa 7.9 really late? i think the intel people are waiting for the glsl2 stuff before pushing it [02:24] so much has changed in it and it seems like 7.8 is dying [02:24] I was expecting it to be branched, given than 7.9.0 should be considered 7.9rc1 on past form. [02:25] idr has been doing the releasing and if they're waiting for the glsl2 stuff i don't know that we'll even see it in time for maverick? [02:27] it was in lockstep with the intel quarterly releases for awhile there I mean [03:49] Stupid X says: Lunch time! [08:51] hey guys [08:51] can anyone speak to me about what is needed to triage those pesky i845g bugs [08:52] i am working on two so far and was wondering what you guys would need === yofel_ is now known as yofel === Azelphur_ is now known as Azelphur [13:38] Hi [13:38] A question [13:39] nvidia-current-dev from the ubuntu-x ppa is incompatible with sdl1.2-dev on lucid. [13:39] any chance this will be changed? [13:40] libsdl1.2-dev ? [13:40] The reason seems to be that sdl1.2-dev depends on mesa [13:40] yes libsdl1.2-dev [13:41] you can try the latest version of nvidia-current and nvidia-current -dev from maverick [13:41] it should work [13:41] thnks [13:41] np [13:41] dumb question noobie i guess, where and what is maverick? [13:42] maverick is the next release of Ubuntu [13:42] OK i see. [13:42] Didn't even know it had a name yet. [13:42] thanks i will check it out [13:42] have a nice day [16:42] Is it possible to use kubuntu with two monitors with the default driver? [16:42] I think I have an nvidia card, and I have the nuoveau driver. [16:42] yes [16:43] (modulo driver bugs, of course) [16:43] What do I have to do? My second screen is blank and the system settings screens module doesn't find it. [16:43] 'This module is only for configuring systems with a single desktop spread across multiple monitors. You do not appear to have this configuration.' [16:44] pastebin the output of xrandr somewhere [16:45] http://dpaste.com/218467/ [16:51] jcristau: Anything relevant there? [16:52] either you have 2 identical monitors, or the driver is confused [16:52] The monitors are the same AFAIK [16:53] well this says the driver thinks it has enabled 2 monitors [16:54] And that should be enough to configure it to be a 'double width workspace'? (you can see I don't know the jargon) [16:55] 'xrandr --output DVI-I-2 --right-of DVI-I-1' should extend the desktop [16:55] instead of cloning it [16:55] but if your second screen is blank then something's wrong already [16:56] The liveCD cloned it, but when I rebooted the right screen was blank, [17:00] Hi all! Is there any chance of getting a new package build of mesa to fix LP #600243? [17:00] alf__: Bug 600243 on http://launchpad.net/bugs/600243 is private [17:02] No it is not: [17:02] libegl1-mesa-dev needs to depend on other -dev packages [17:02] alf__: what are you building that doesn't have xserver-xorg-dev in the build deps? [17:02] Rebooting that machine didn't help anyway... [17:03] Sarvatt: GL/GLESv2 benchmarks [17:05] thats kind of silly, egl can be used without X too [17:07] pushed that patch to git [17:07] Any other bright ideas for me to get dual screens? [17:07] alf__: thanks [17:08] Or even debug why I don't even get cloning after installing, while I do before [17:08] steveire: check the kernel logs for both, see if you spot a difference [17:08] How? Where are they? [17:08] dmesg [17:10] alf__: i have a hard time caring about egl on hurd ;) [17:11] jcristau: I share your sentiments, just wanted to be thorough :) [17:16] Sarvatt: Yes, indeed EGL should be usable without X. But why are all these dependencies in the .pc files in the first place? [17:23] alf__: i can't upload it unfortunately (and i've been trying to get libdrm sponsored for a month already), sorry :( [17:24] Sarvatt: No worries [17:25] Sarvatt: it's silly that you're still not core dev [17:29] been working on it in maverick, i didn't actually upload much to ubuntu in the past [17:30] Sarvatt: do you only need to upload a new libdrm? [17:30] or is there anything else? [17:31] http://sarvatt.com/downloads/merges/libdrm/ [17:31] well alf__'s mesa fix [17:32] jcristau: Installed version: http://pastebin.com/rDDLkwia liveCD version: http://pastebin.com/cbAyRsQj [17:32] I don't know what kind of differences I need to look for [17:38] mesa needs a merge anyway but i dont have the battery power to test build it at the moment, would be easier to just add his mesa-common-dev dep changes to the current one [17:39] steveire: don't know either. best would be to file a bug, and/or try with a newer kernel [17:39] ppa builds are backed up 3+ days now too [17:41] steveire: have you tried a maverick livecd by any chance? [17:42] Sarvatt: ok, I can upload alf's change. Did you test libdrm? [17:42] Sarvatt: I have not [17:42] tseliot: yeah, it's been in x-updates as well [17:44] Sarvatt: ah, I tested it on my main computer then [17:46] Sarvatt: ok, I'll sponsor both things then [17:46] thank you a ton tseliot! [17:46] np [17:49] Sarvatt, jcristau, tseliot: Thanks a lot! [17:50] alf__: thanks to you ;) [18:13] all uploaded [18:14] Sarvatt: shall I update only the mesa git branch? [18:30] jcristau: can you check if I still have an account on alioth, please? [18:31] it looks like I don't... [18:33] uid=229885(tseliot-guest) gid=229885(tseliot-guest) groups=229885(tseliot-guest),41008(pkg-xorg),81008(scm_pkg-xorg) [18:34] I forgot the "-guest" part, sorry [18:34] it's been while since I've used alioth's git branches [18:35] the connection timed out, I guess I made too many attempts... [18:36] should be fine in 10 minutes or so, they're using fail2ban on alioth :/ [18:36] ok, thanks === Azelphur_ is now known as Azelphur [19:50] Where is the button on https://bugs.launchpad.net/ for reporting a new bug? [19:54] I got the no-redirect link. How do I answer the question 'In what package did you find this bug' if the problem is that a dual screen setup works in a livecd and not after installation. [19:54] just run ubuntu-bug xorg [19:56] I'll put xorg in that box then. [20:05] am I really wrong? - https://bugs.launchpad.net/bugs/578038 [20:05] Launchpad bug 578038 in libxext (Ubuntu) "libext6 does not create symlink in /usr/lib on amd64 (affects: 1) (heat: 6)" [Undecided,Incomplete] [20:06] * Sarvatt must not be understanding the situation [20:07] Sarvatt: submitter is mistaken [20:07] that bug should be marked invalid [20:09] hmm, i thought they had to be in bug control to change the status away from invalid, guess not [20:10] looks like i was allowed to move it back [20:18] thanks jcristau, he probably wouldn't have believed me and it would have been an open/close war if i closed it again myself :) [20:20] jcristau: can you set a bug to wont fix status? if not i can just do that if he does again if anything [20:21] no idea [20:38] I can't start gnome with xorg-edgers on lucid, getting: [20:38] nautilus: symbol lookup error: /usr/lib/libcairo.so.2: undefined symbol: pixman_image_composite32 [20:38] can't compile cairo after installing libpixman(-dev) from xorg-edgers either [20:39] Sarvatt, do you know anything about what that could be? [20:40] on amd64 === |Alexia_Death| is now known as Alexia_Death [20:41] oh there was a new cairo build 16 minutes ago... [20:43] that's a broken dependency. [20:46] right, but why? [20:48] *shrug* [20:48] because shit happens [20:49] well, wierd... [20:49] gonna see if the new cairo upload helps the situation [20:54] damnit, it didn't help [20:58] does xorg-edgers work for anyone on lucid? [21:08] jcristau, broken due to cjwatson's apt breakage from earlier this morning? === chrisccoulson__ is now known as chrisccoulson [22:04] bryce2, hi! any idea about the cairo/pixman breakage? [22:15] ernstp, notta clue [22:16] ernstp, I saw an email from cjwatson that everything was broken from building earlier today. dunno any details [22:18] bryce2, for me it's been broken like that for a couple of weeks [22:19] oh, must be something unrelated [22:19] (xorg-edgers on lucid, the missing pixman_image_composite32 symbol) [22:19] I'm curious if xorg-edgers works for anyone on lucid... ? [22:21] means you have an old pixman [22:22] that symbol is included in pixman 0.18 [22:23] I have pixman 0.19 from xorg edgers [22:23] when that occurs [22:23] then make sure you don't also have an older one in /usr/local [22:23] because this sure sounds like pebcak [22:24] jcristau, oooh, there is something in /usr/local indeed! [22:25] jcristau, thanks a lot :-)