| mlankhorst | ok so wine testbot is on life support :( | 10:45 |
|---|---|---|
| Chipaca | hi guys. intel arrandale in R here reporting “not direct rendering capable.”, falling back to llvm :( | 11:18 |
| jcristau | x log and dmesg? | 11:18 |
| Chipaca | http://paste.ubuntu.com/1506381/ | 11:18 |
| Chipaca | http://paste.ubuntu.com/1506383/ | 11:19 |
| Chipaca | jcristau: that's /var/log/Xorg.0.log and /var/log/dmesg respectively | 11:19 |
| jcristau | [144099.016] (WW) intel(0): cannot enable DRI2 whilst forcing software fallbacks | 11:20 |
| Chipaca | jcristau: yeah, I think I understand what it means, but not how it got there | 11:21 |
| Chipaca | (i.e., why is it forcing software fallbacks?) | 11:21 |
| Chipaca | ("well don't do that then!" :) ) | 11:21 |
| jcristau | yeah, dunno | 11:22 |
| Chipaca | new kernel in today's updates, am going to reboot and retest. This is something that starts happening after a while of uptime (maybe after suspend?), so it'll be a little bit before I can re-report | 11:24 |
| Chipaca | o/ | 11:24 |
| jcristau | i'd start with removing the i915 and drm options from the kernel cmdline | 11:24 |
| mlankhorst | it sounds more like a xorg option though | 11:29 |
| jcristau | intel->force_fallback = | 11:29 |
| jcristau | drmCommandNone(intel->drmSubFD, DRM_I915_GEM_THROTTLE) != 0; | 11:29 |
| mlankhorst | ah right | 11:29 |
| jcristau | also: | 11:30 |
| jcristau | if (!can_accelerate_blt(intel)) { | 11:30 |
| jcristau | intel->force_fallback = TRUE; | 11:30 |
| mlankhorst | prerelease hardware crap, probably | 11:30 |
| jwi1 | i915_enable_rc6 is out to to kill you... | 11:30 |
| jcristau | dunno which of those it can be, having a note in the log would be helpful | 11:30 |
| mlankhorst | Chipaca: lspci -vvv, graphics card listing? | 11:30 |
| mlankhorst | jcristau: well after I had a similar problem they added a message to dmesg :-) | 11:31 |
| tjaalton | arrandale is old, don't think it's meant to be used with rc6 anyway.. silly to force it | 11:31 |
| Chipaca | fwiw, with X freshly restarted I again have drm. Do you want the lspci after it stops working? | 11:31 |
| mlankhorst | hm just try without all the kernel options first | 11:32 |
| Chipaca | mlankhorst: how do i list the graphic card? | 11:32 |
| Chipaca | mlankhorst: http://paste.ubuntu.com/1506405/ of it now | 11:32 |
| Chipaca | to repeat, it works fresh after a reboot, then suddenly unity crashes and doesn't come back up again (restarting X fixes it, sometimes) | 11:33 |
| Chipaca | I have not yet pinpointed the bug, as you see; was hoping it was something known :) | 11:33 |
| mlankhorst | I would try without the kernel options first, if it's not enabled by default either the intel devs are evil, or they think it's not ready yet for your card :/ | 11:34 |
| tjaalton | yeah, probably the gpu gets in a "state" due to the driver not really supporting those options, because of instability or whatever | 11:37 |
| jwi1 | the lack of any hangcheck vomit in dmesg is a little surprising. | 11:41 |
| jwi1 | picking up the fix for https://bugs.freedesktop.org/show_bug.cgi?id=55984 probably wouldn't hurt (raring still has uxa by default, right?) | 11:43 |
| ubottu | Freedesktop bug 55984 in DRM/Intel "[ilk regression] gpu hangs on ironlake with 3.6 + -next + -fixes code" [Normal,Needinfo] | 11:43 |
| jcristau | ah, so it's just a regular hang, and reset not working, then. | 12:35 |
| mlankhorst | ok so erm new xserver-xorg-video-nouveau | 12:59 |
| shadeslayer | not exactly sure what happened but for some reason I've lost compositing on raring | 13:19 |
| shadeslayer | /var/log/Xorg.0.log | 13:19 |
| shadeslayer | erm | 13:19 |
| shadeslayer | http://paste.ubuntu.com/1506621/ | 13:19 |
| shadeslayer | http://paste.ubuntu.com/1506622/ < dmesg | 13:20 |
| jcristau | uh that backlight spam in xorg.log is insane | 13:21 |
| jcristau | shadeslayer: define 'lost compositing' | 13:21 |
| jcristau | also what's with people setting random module options in kernel cmdline today? | 13:22 |
| tjaalton | hehe | 13:22 |
| shadeslayer | jcristau: earlier, with those options, kwin was using opengl | 13:23 |
| shadeslayer | but now it falls back to xrender | 13:23 |
| * jcristau blames kwin. | 13:23 | |
| shadeslayer | ( not to mention no backlight control ) | 13:23 |
| shadeslayer | somehow I doubt that | 13:24 |
| mlankhorst | erm sometimes it happens | 13:24 |
| mlankhorst | grep OpenGL ~/.kde/share/config/kwinrc | 13:25 |
| shadeslayer | OpenGLIsUnsafe=false | 13:25 |
| shadeslayer | it worked perfectly before I rebooted into the new kernel | 13:25 |
| mlankhorst | should also say Backend=OpenGL, but it doesn't :P | 13:25 |
| shadeslayer | Backend=XRender | 13:25 |
| shadeslayer | like I said | 13:25 |
| mlankhorst | but can change that in systemsettings | 13:25 |
| shadeslayer | I tried | 13:26 |
| shadeslayer | but it falls back to XRender | 13:26 |
| mlankhorst | does it say anything? | 13:26 |
| shadeslayer | not really no | 13:26 |
| shadeslayer | "Failed to activate desktop effects using the given configuration options. Settings will be reverted to their previous values. | 13:28 |
| shadeslayer | Check your X configuration. You may also consider changing advanced options, especially changing the compositing type." | 13:28 |
| mlankhorst | what if you start kwin manually from command line after setting that backend line with kwin --replace ? | 13:28 |
| jcristau | also, glxinfo | 13:29 |
| shadeslayer | need a couple of minutes | 13:31 |
| shadeslayer | hah, kwin crashed | 13:33 |
| shadeslayer | libGL error: failed to load driver: i965 | 13:33 |
| shadeslayer | what | 13:33 |
| mlankhorst | LIBGL_DEBUG=verbose glxinfo ? | 13:34 |
| shadeslayer | http://paste.kde.org/639572/ | 13:36 |
| jcristau | sounds like you don't have access to /dev/dri/card0 | 13:37 |
| tjaalton | issues with policykit then? | 13:37 |
| jcristau | no | 13:37 |
| jcristau | would be consolekit if anything afaik | 13:37 |
| tjaalton | ah | 13:37 |
| tjaalton | well, probably the same reason as for the acpi spam | 13:38 |
| shadeslayer | okay, how about this, there are quite some upgrades that I need to do, I'll do those and get back to you? | 13:38 |
| shadeslayer | I pinged because I thought maybe it was a known issue | 13:38 |
| tjaalton | works fine here | 13:41 |
| jcristau | tjaalton: do you guys have some random wiki or forum or doc advising to set i915 module options? | 13:42 |
| shadeslayer | hrh | 13:42 |
| shadeslayer | heh | 13:42 |
| shadeslayer | regarding i915 module options | 13:42 |
| shadeslayer | I got those from a blog that I saw a couple of weeks ago | 13:42 |
| shadeslayer | because I have dual graphics cards and I disable the ATi card | 13:43 |
| tjaalton | jcristau: no :) | 13:45 |
| tjaalton | forum users maybe get silly ideas | 13:45 |
| mlankhorst | tjaalton: well if you google it you find some of those.. | 13:45 |
| jcristau | s/maybe/definitely/ | 13:45 |
| jcristau | "if somebody said so on the intarwebs it must be a good idea" | 13:46 |
| tjaalton | and macbooks suck | 13:48 |
| mlankhorst | indeed :( | 13:50 |
| mlankhorst | need to finish up on my quirking for it | 13:50 |
| === Quintasan_ is now known as Quintasan | ||
| mlankhorst | sent \o/ | 14:23 |
| tjaalton | mlankhorst: oh you got a macbook now?-) | 14:39 |
| mlankhorst | tjaalton: for testing prime stuff | 14:41 |
| mlankhorst | debugging that it lost interrupts after a switcheroo cycle was fun :/ | 14:41 |
| mlankhorst | bryce: hm what would be useful to have is one of those versions check, but then for all precise .*-lts-quantal package versions compared against what's available in quantal | 15:15 |
| shadeslayer | seems like a minor issue | 15:41 |
| shadeslayer | erm | 15:41 |
| shadeslayer | s/minor/temporary/ | 15:41 |
| shadeslayer | gone after fully upgrading and rebooting | 15:41 |
| mlankhorst | ok started uploading the raring stack to my ppa as well, hopefully I was right and upgrading between stacks will work succesfully too :-) | 15:46 |
| === chrisccoulson_ is now known as chrisccoulson | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!