Sarvatt__ | hrm I'm not sure when it should go in, looks like main is frozen before beta freeze is lifted? | 00:01 |
---|---|---|
Sarvatt__ | RAOF: err, compiz memory leak happening with the nvidia blob too not just intel | 00:55 |
bjsnider | you got that right | 00:55 |
bjsnider | have to restart compiz every few hours | 00:55 |
RAOF | On alt-tab? | 00:55 |
Sarvatt__ | sarvatt 1911 0.5 0.7 133924 30920 ? Sl Apr06 37:33 compiz is with the blob here, I never alt+tab | 00:56 |
bjsnider | neither do i | 00:57 |
Sarvatt__ | was surprised the blob was doing it too though, rules out some of the places I was looking.. gonna need a valgrind | 00:57 |
RAOF | Is there a bug filed, and can you valgrind it? | 00:57 |
bjsnider | fixing cairo didn't help | 00:57 |
RAOF | Fixing cairo was never *meant* to help compiz | 00:58 |
RAOF | So we've got two known OMG bugs on the go, then :) | 00:59 |
Sarvatt__ | I hit the min/max under compiz one a crapload and didn't realize it | 01:00 |
RAOF | The endless-waiting-in-_XReply one? | 01:01 |
* Sarvatt__ has a bad habit of min/maxing windows passing time :) | 01:01 | |
Sarvatt__ | yeah | 01:01 |
RAOF | Ok. Make a coffee, finish fixing mono, track down crazy _XReply bug. | 01:02 |
Sarvatt__ | bjsnider: do you use unity or a classic session? | 01:11 |
bjsnider | unity | 01:13 |
bjsnider | i thought that cairo was being run as a thread in compiz and with the cairo problem fixed the mem usage would be, but no joy | 01:13 |
RAOF | bjsnider: No, the cairo problem was a fixed 5MiB cost for every process that linked to cairo (ie: every GUI app). | 01:14 |
bjsnider | well, i think in my case looking at the numbers, compiz has been gobbling up lots of ram since i started using natty 2 weeks ago. the other processes are ok | 01:15 |
Sarvatt__ | gonna try to look at it in the morning and wanted to be sure you were using unity, I sure as heck can't reproduce it with the blob on a classic session at least | 01:16 |
bjsnider | well, that narrows it down | 01:17 |
bjsnider | i can see compiz ending up at >500 mb after a few hours | 01:17 |
bjsnider | i don't have a swap file so i have to head that off before it goes too far | 01:18 |
ScottK | Sarvatt__: We won't unfreeze. Go ahead and upload and the release team will figure out the best time to get it in. | 01:50 |
ScottK | Sarvatt__: I just uploaded it. | 02:04 |
ScottK | It's in queue for release team review. | 02:08 |
Sarvatt__ | ScottK: thanks a ton for that | 03:15 |
ScottK | Sarvatt__: It's in. | 03:40 |
Sarvatt_ | hrm got one http://paste.ubuntu.com/593017/ | 09:07 |
Sarvatt_ | just close the lid with g-p-m set to blank the screen under unity on intel | 09:07 |
Sarvatt_ | comes back up in a weird state where you can interact with things partially but compiz is hung | 09:09 |
RAOF | Partially? | 09:09 |
Sarvatt_ | yeah cursor is working, it changes when i hover over a resize area but cant change windows or anything | 09:10 |
RAOF | Right. | 09:11 |
Sarvatt_ | i just attached gdb to compiz there while it was in that state | 09:11 |
RAOF | I *think* that's another manifestation of the same bug. | 09:11 |
RAOF | X is happily motoring along, compiz is sitting waiting for X to reply to its request. | 09:11 |
Sarvatt_ | hmm either http://cgit.freedesktop.org/xcb/libxcb/commit/?id=f0565e8f06aadf760a9065a97b8cf5ab9cbd18de helped or I just got lucky being able to reproduce it 3 times in a row there earlier :) | 09:43 |
tjaalton | there was talk on the xcb list last month about a new release, but nothing came out of it, yet | 09:44 |
RAOF | sarvatt: Hm. If I read that commit correctly it should result in compiz dying with an IO error. | 09:46 |
RAOF | Sarvatt_: Yeah; that'll result in the X connection being shut down, which compiz is unlikely to be happy with. | 09:47 |
RAOF | Sarvatt: If you want a nice reproducer, revert bzr r1091 from unity. That'll make maximize kill things again ;) | 09:53 |
Sarvatt_ | must just need to have it closed for a longer period of time to reproduce it then, had it closed >30 minutes each time before | 09:53 |
Sarvatt_ | oh maximize is fixed now? | 09:54 |
Sarvatt_ | woohoo | 09:54 |
RAOF | Yeah. Well, worked around, at least. | 09:54 |
Sarvatt_ | ugh, intels gonna need another upload now too | 10:28 |
Sarvatt_ | http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=686018f283f1d131073ef5917213e6a8ac013f26 fixing all the 915gm hangs | 10:28 |
=== yofel_ is now known as yofel | ||
seb128 | could someone reassign, set bug settings as appropriate for bug #757906 | 12:35 |
ubot4 | Launchpad bug 757906 in unity (Ubuntu) "compiz assert failure: compiz: intel_tex_image.c:726: intelSetTexBuffer2: La declaraciĆ³n `!texImage->Data' no se cumple. (affects: 6) (dups: 6) (heat: 56)" [Medium,New] https://launchpad.net/bugs/757906 | 12:35 |
tjaalton | seb128: done | 12:37 |
seb128 | tjaalton, thanks | 12:37 |
seb128 | tjaalton, I've dumped some similar bugs I found googling about the error in the comments | 12:38 |
seb128 | not sure how useful they are though ;-) | 12:38 |
tjaalton | seb128: we'll see :) | 12:39 |
=== Sarvatt is now known as Sarvatt|2 | ||
=== Sarvatt_ is now known as Sarvatt | ||
lool | Hey | 17:06 |
lool | syndaemon is using 100% here, even after logging out and back in; did anybody see this so far? | 17:06 |
lool | looks like https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-input-synaptics/+bug/754470 | 17:07 |
ubot4 | Launchpad bug 754470 in xserver-xorg-input-synaptics (Ubuntu) "syndaemon consumes 100% CPU (affects: 31) (dups: 1) (heat: 150)" [Undecided,Confirmed] | 17:07 |
lool | cnd: ^ | 17:12 |
cnd | lool, I haven't seen it | 17:13 |
cnd | I thought syndaemon was deprecated... | 17:13 |
cnd | I guess not since I'm running it :) | 17:15 |
Sarvatt | the disable touchpad while typing option in the gnome capplet uses it | 17:15 |
cnd | lool, can you install the dbgsym package and gdb it to see where it's stuck? | 17:16 |
lool | cnd: Actually I rebooted after a kernel upgrade and it doesn't happen for me anymore | 17:21 |
lool | I got it three times after logging out and killing it, but now that I rebooted it's gone | 17:23 |
lool | cnd: I'll gdb it if I reproduce though | 17:23 |
cnd | lool, ok, thanks! | 17:23 |
Sarvatt | updating libxcb is always an adventure.. https://launchpadlibrarian.net/69144345/buildlog_ubuntu-natty-i386.libxcb_1.7%2Bgit20110412.42c4adef-0ubuntu0sarvatt_FAILEDTOBUILD.txt.gz | 17:40 |
Sarvatt | at least it doesnt need autoreconf patches refreshed every time now | 17:40 |
bryceh_ | Sarvatt, did that -intel patch get taken care of by anyone already? Should I roll up a ppa and solicit testers? | 18:21 |
Sarvatt | bryceh_: nope not yet, it'll need a bit of cleaning up for 2.14.0 because of has_kernel_flush | 18:23 |
bryceh_ | ok | 18:23 |
bryceh_ | Sarvatt, did everything get in yesterday that needed in? mesa and whatnot? | 18:32 |
Sarvatt | bryceh_: yep! | 18:32 |
bryceh_ | ok great | 18:32 |
tjaalton | i bought an esprimo v5535 with sis mirage 3+ (672) on it, oh my.. | 19:04 |
bryceh_ | tjaalton, get it for a good price? | 19:05 |
tjaalton | bryceh_: that would've meant "free" :) | 19:05 |
bryceh_ | heh | 19:05 |
bryceh_ | guess it'll keep you busy at least | 19:06 |
tjaalton | well, it's a bit faster than the old thinkpad t23, so once flash runs on it I'll pass it on to the kids | 19:06 |
bryceh_ | I got an old via system hereabouts (one of those walmart PCs back when they were selling linux) | 19:07 |
tjaalton | does openchrome work with it? | 19:07 |
bryceh_ | unfortunately the cpu isn't well supported by the kernel so last I tested (which was a while ago) I couldn't even get it booting | 19:08 |
bryceh_ | figure I'd need to roll my own kernel with a custom config. maybe someday | 19:08 |
tjaalton | huh, that's.. bad | 19:09 |
tjaalton | via cpu? | 19:09 |
bryceh_ | C3 iirc | 19:09 |
tjaalton | right, so the optimizations make it unbootable | 19:10 |
tjaalton | or the cpu family support in -generic | 19:10 |
bryceh_ | yeah at the time I looked at it, it just needed a config setting that we don't include by default | 19:10 |
bryceh_ | but even when it does work, the system is slow as snot | 19:11 |
tjaalton | hehe | 19:11 |
tjaalton | this one seems to be noisy | 19:11 |
tjaalton | the fan is on full blast even when idling | 19:11 |
bryceh_ | ew | 19:12 |
tjaalton | i suspect overheating issues | 19:12 |
tjaalton | in the future | 19:12 |
bryceh_ | hot gpu or does it just not support power modes? | 19:12 |
tjaalton | it was the same on win7 | 19:12 |
bryceh_ | ah | 19:12 |
tjaalton | i'll try with only the battery.. | 19:12 |
bryceh_ | it's crazy how many gpu's I've burned through so far working on X | 19:13 |
bryceh_ | 4-5 I think | 19:13 |
bryceh_ | (mainly ati) | 19:13 |
tjaalton | though seems like the live-session doesn't support cpu throttling | 19:13 |
tjaalton | oh, just overheated them? | 19:13 |
bryceh_ | tjaalton, any theories on bug 747205? | 19:15 |
ubot4 | Launchpad bug 747205 in libdrm (Ubuntu) (and 1 other project) "[arrandale] Black screen on all outputs when external VGA is connected (affects: 7) (dups: 2) (heat: 50)" [High,Triaged] https://launchpad.net/bugs/747205 | 19:15 |
tjaalton | bryceh_: well, it's worse with compiz. but the register dumps are what ickle wants now | 19:17 |
tjaalton | i mean with metacity it's almost equally broken, just that you get back to normal by unplugging the monitor | 19:18 |
bryceh_ | right | 19:20 |
BlackZ | hello, are nvidia drivers expected to be broken on natty? | 19:57 |
Sarvatt | only nvidia-173 and nvidia-96 | 19:58 |
BlackZ | Sarvatt: I can't get my video card working with nvidia-current | 19:59 |
Sarvatt | what GPU is it? can ya describe how its broken? like can you not install the package at all? | 20:00 |
BlackZ | Sarvatt: nvidia geforce 7500 LE; I can't run the default Ubuntu desktop session (even unity doesn't run completely); I can successfully install the package and build the kernel module | 20:01 |
Sarvatt | BlackZ: yeah that one needs the nvidia-173 driver and not nvidia-current unfortunately :( | 20:02 |
Sarvatt | oh wait, I'm sorry | 20:02 |
bjsnider | no, it shuld work | 20:02 |
bjsnider | 6k and onwards | 20:02 |
bjsnider | BlackZ, have you got any crazy ppa add-ons? | 20:03 |
BlackZ | bjsnider: no, fresh natty install | 20:03 |
Sarvatt | BlackZ: can you pastebin a /var/log/Xorg.0.log (or .old) from a failed startup? | 20:04 |
BlackZ | Sarvatt: http://pastebin.ubuntu.com/593259/ | 20:06 |
tjaalton | um, looks normal | 20:07 |
Sarvatt | yeah that looks fine, how about ~/.xsession-errors? | 20:07 |
BlackZ | Sarvatt: http://pastebin.ubuntu.com/593263/ | 20:09 |
Sarvatt | sudo apt-get purge nvidia-173 for starters | 20:10 |
BlackZ | Sarvatt: it's not installed | 20:11 |
BlackZ | Sarvatt: however I have no problems at booting time; I have problems when starting a session in gdm (e.g. the default Ubuntu one) | 20:12 |
BlackZ | Sarvatt: if I use the nouveau drivers all works good | 20:13 |
Sarvatt | i'm not sure whats going on with your sessions there but it looks like the nvidia driver is working fine, do you have any packages missing by any chance? sudo apt-get install ubuntu-desktop offer to install anything? | 20:15 |
BlackZ | Sarvatt: no, it doesn't | 20:16 |
BlackZ | Sarvatt: however I only have this problem on natty; on maverick the proprietary nvidia drivers works good | 20:21 |
bjsnider | BlackZ, did you install all updates as of today? | 20:28 |
BlackZ | bjsnider: yes | 20:28 |
jcastro | Sarvatt: other than the occasional artifact in the dash the thinkpad X120E is running great with the open drivers, thanks for sorting out snagging the patch from upstream | 20:28 |
bjsnider | Sarvatt, how much ram would you say compiz should typically be using in natty? | 21:23 |
JanC | I think RAM usage depends on what plugins you have enabled and probably also on display size, number of workspaces, & such? | 21:34 |
bjsnider | JanC, right. in this case, it's just stock default unity | 23:08 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!