| RAOF | Ok. New mesa 9.1 snapshot question - is anyone seeing a gem object leak on sandybridge? | 01:20 |
|---|---|---|
| === tomreyn_ is now known as tomreyn | ||
| mlankhorst | RAOF: no but I haven't looked for one either, so maybe? :p | 08:35 |
| tjaalton | i'm running 9.1 on snb, how can I tell if it leaks or not? | 08:35 |
| RAOF | tjaalton: while true ; do egrep "^[[:digit:]]+ objects, [[:digit]]+ bytes" /sys/kernel/debug/dri/0/i915_gem_objects ; sleep 1 ; done | 08:37 |
| RAOF | That'll print out the total size; if it keeps increasing over time (not necessarily strictly increasing, but just tending to increase) then you're probably hitting it. | 08:38 |
| RAOF | Other symptoms include: after a while everything will crash as the OOM killer flails around before killing compiz. | 08:39 |
| * mlankhorst hands RAOF watch -n1 | 08:40 | |
| * RAOF gratefully receives it | 08:40 | |
| tjaalton | the egrep part is missing something | 08:41 |
| tjaalton | egrep: Pariton [ tai [^ | 08:41 |
| RAOF | Sorry, yeah. | 08:41 |
| tjaalton | sorry for the locale | 08:41 |
| tjaalton | :) | 08:41 |
| RAOF | It's missing the trailing ‘:’ from the second :digit: | 08:41 |
| tjaalton | oh | 08:41 |
| tjaalton | ok, I'll let it run for a while | 08:46 |
| * mlankhorst needs to figure out why nouveau is failing :( | 08:47 | |
| mlankhorst | not upstream, just wrote some patches for the new synchronization framework, fails on my poor nva8 | 08:50 |
| tjaalton | RAOF: doesn't seem to be increasing here | 09:07 |
| tjaalton | seems to stay around ~770 objects, ~510000000 bytes | 09:08 |
| RAOF | tjaalton: Ta. Urgh. | 09:27 |
| mlankhorst | RAOF: if you know what application, valgrind to the rescue? | 10:30 |
| RAOF | mlankhorst: Plausibly. Tis the mighty compiz, though. | 10:30 |
| mlankhorst | I mean it would probably show up as a normal allocation, so just crank up logging on valgrind and kill it after a bit | 10:31 |
| RAOF | Yeah. As long as it's not free()ing the bo without bo_unref()ing it, it should leak valgrind-tracable memory. | 10:32 |
| RAOF | Now, to see if this abuse of prime does anything interesting... | 10:32 |
| mlankhorst | did it? | 12:56 |
| shadeslayer | hm, can anyone install the fglrx driver on raring? | 13:21 |
| shadeslayer | update-initramfs just goes kaput | 13:21 |
| shadeslayer | http://paste.kde.org/670244/ | 13:22 |
| shadeslayer | I also see a bunch of bug reports | 13:22 |
| ogra_ | shadeslayer, thats modprobe, not update-initramfs | 13:35 |
| ogra_ | (talk to #ubuntu-kernel probably) | 13:37 |
| shadeslayer | ogra_: okay :) | 14:35 |
| shadeslayer | could someone also advise me on this : http://paste.ubuntu.com/1644035/ | 14:36 |
| shadeslayer | I don't quite understand what's going wrong there | 14:41 |
| === xnox is now known as foxtrot | ||
| === foxtrot is now known as xnox | ||
| === yofel_ is now known as yofel | ||
| mdeslaur | I just upgraded to raring, and I have to reboot about 10 times for X to come up. Is "Cannot run in framebuffer mode. Please specify busIDs" a known bug, and is there a bug number? | 19:31 |
| tjaalton | hehe | 19:32 |
| tjaalton | it's well known | 19:32 |
| tjaalton | mdeslaur: for a workaround, remove -modesetting & -fbdev | 19:34 |
| tjaalton | pastebin your logfile to be sure | 19:34 |
| bryce | tjaalton, does the sleep in /etc/lightdm/lightdm.conf also work around it still? | 19:35 |
| tjaalton | bryce: ah, maybe | 19:35 |
| tjaalton | haven't tried that one | 19:35 |
| bryce | might be a less invasive workaround | 19:35 |
| mdeslaur | tjaalton: http://paste.ubuntu.com/1645897/ | 19:35 |
| mdeslaur | tjaalton: remove that from what, exactly....are those packages? | 19:35 |
| tjaalton | [ 4.157] setversion 1.4 failed | 19:35 |
| tjaalton | bingo | 19:35 |
| mdeslaur | ie: xserver-xorg-video-fbdev | 19:35 |
| tjaalton | mdeslaur: xserver-xorg-video- | 19:35 |
| tjaalton | yes | 19:36 |
| mdeslaur | tjaalton: ok, I only have -fbdev | 19:36 |
| mdeslaur | ah, no, I've got the other one too | 19:36 |
| * mdeslaur is in fat-finger mode | 19:36 | |
| mdeslaur | tjaalton: is there a bug #? | 19:36 |
| bryce | https://bugs.launchpad.net/ubuntu/+source/libdrm/+bug/982889/comments/5 | 19:36 |
| ubottu | Ubuntu bug 982889 in xorg-server (Ubuntu) "X trying to start before plymouth has finished using the drm driver" [Critical,Triaged] | 19:36 |
| bryce | mdeslaur, bug 982889 | 19:37 |
| mdeslaur | thanks bryce | 19:37 |
| tjaalton | mdeslaur: did you use quantal before? | 19:38 |
| mdeslaur | tjaalton: yep, never got that on quantal | 19:38 |
| tjaalton | weird | 19:38 |
| tjaalton | but maybe it's slightly quicker to boot now, which makes it easier to hit | 19:39 |
| mdeslaur | yeah, it's called progress...darn people with they speed optimizations :P | 19:40 |
| mdeslaur | hehe :) | 19:40 |
| bryce | or kernel got slower... | 19:41 |
| fmms | is chris willson here? | 20:20 |
| fmms | https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/1116587 | 20:22 |
| ubottu | Ubuntu bug 1116587 in xserver-xorg-video-intel (Ubuntu) "Lenovo T400: With external monitor the system is rarely working for more than 10 minutes" [Undecided,Confirmed] | 20:22 |
| bryce | fmms, he is ickle on #intel-gfx | 20:24 |
| fmms | bryce, thanks | 20:32 |
| === Quintasan_ is now known as Quintasan | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!