| bryce | wrote up https://wiki.ubuntu.com/X/Architecture | 01:56 |
|---|---|---|
| bryce | (corrections welcomed) | 01:56 |
| pwnguin | you know, I'd kind of expect a maintainer of inkscape to cook up a better diagram than that ;) | 03:43 |
| bryce | heh, actually I didn't draw that | 03:44 |
| bryce | I may redo it nicely in inkscape next time I have a chunk of time | 03:45 |
| bryce | but other than being ugly it's not a bad diagram | 03:45 |
| pwnguin | where does the kernel stop / start? | 03:46 |
| bryce | in the diagram, the kernel would be the box labeled "Graphics Hardware" | 03:46 |
| bryce | everything above that is userspace | 03:46 |
| tjaalton | whoever drew that forgot the "DRI driver" box :) | 05:22 |
| RAOF | bryce: You seem to have opened a bug task for DRI2 against nouveau. Any particular reason? :) The 3d component is (a) not included in Ubuntu, and (b) aggressively unsupported upstream. | 11:00 |
| tjaalton | RAOF: I think it's just a tracker | 11:01 |
| RAOF | Yeah. I just think it gives the wrong impression. | 11:02 |
| tjaalton | maybe importance could be lower | 11:02 |
| RAOF | I don't think it should be on that bug at all - nouveau in Ubuntu doesn't support 3d, so that bug doesn't apply, and by the time upstream stops asking me to not ship the 3d component in any packages whatsoever, it'll be supporting DRI2. | 11:05 |
| tjaalton | hmm well I marked it wishlist already | 11:06 |
| RAOF | Just as long as it doesn't give users the impression that nouveau does 3d at all. :) | 11:07 |
| RAOF | As it currently stands. | 11:07 |
| tjaalton | right | 11:08 |
| tjaalton | maybe mark it wontfix instead with a note that it can be reopened once there is something to offer? | 11:08 |
| RAOF | I'll do so, yes. | 11:09 |
| tjaalton | thanks | 11:09 |
| tjaalton | hm, dexconf is not run when using the alternate installer | 11:16 |
| tjaalton | meaning that xorg.conf is empty | 11:19 |
| tjaalton | laptop wedged with UXA.. time to switch back to EXA | 11:21 |
| tjaalton | "geez, no ctrl-alt-bspace!".. although the only "key" that works seems to be the rfkill-switch | 11:22 |
| === seb128_ is now known as seb128 | ||
| ion_ | I have an identical problem to http://launchpad.net/bugs/330126. I wonder how to debug? | 19:22 |
| ubottu | Ubuntu bug 330126 in xorg-server "Xorg crashed with SIGSEGV in xf86_wrap_crtc_notify()" [Medium,New] | 19:22 |
| tjaalton | http://wiki.ubuntu.com/X/BackTracing | 19:26 |
| tjaalton | sorry | 19:26 |
| tjaalton | http://wiki.ubuntu.com/X/Backtracing | 19:26 |
| ion_ | The backtrace is in the bug report. | 19:28 |
| ion_ | Mine was identical. | 19:29 |
| * bryce looks | 19:29 | |
| bryce | ion_: hmm, well the trace isn't showing that it's just a simple null pointer issue, so the next step for this bug is to step through the code in xf86_wrap_crtc_notify() and examine the values to see why it is segfaulting | 19:32 |
| bryce | possibly there's memory corruption of the config or scrn data structures | 19:32 |
| ion_ | I’ll try to do that. | 19:34 |
| tjaalton | isn't it a dupe of bug 319210? | 19:36 |
| ubottu | Launchpad bug 319210 in xorg-server "[jaunty] segfault during X startup with randr < 1.2 drivers" [High,In progress] https://launchpad.net/bugs/319210 | 19:36 |
| bryce | mm, maybe. ion_ it would be helpful if you could test the patch attached to that bug. It revises how xf86_wrap_crtc_notify works, which may prevent the segfault | 19:37 |
| ion_ | Will do. | 19:37 |
| ion_ | bryce: The patch (that is, the result of ‘git diff ea309e47457156b60aadbf113f04e5b6851029c8^ ea309e47457156b60aadbf113f04e5b6851029c8’ applied to the Ubuntu source package’s tree) fixed the problem. | 20:22 |
| ion_ | Thanks | 20:25 |
| bryce | ion_: sweet, thanks for testing | 20:31 |
| bryce | ion_: can you comment as such on the bug, and assign it to me, and I'll put the patch in when I get a chance | 20:31 |
| bryce | oh actually | 20:32 |
| ion_ | Someone already mentioned it fixes savage in the comments of 319210. | 20:32 |
| bryce | 319210 is already assigned to me, so comment your test findings there | 20:32 |
| bryce | I'll dupe your other bug to that | 20:32 |
| ion_ | I already duped it. (It was reported by someone else than me.) | 20:32 |
| bryce | ahh, great thanks | 20:33 |
| ion_ | Should i still add a comment? | 20:33 |
| bryce | sure | 20:52 |
| bryce | extra confirmations of a fix are always welcome | 20:53 |
| ion_ | Alright | 20:59 |
| tjaalton | bryce: it's already in 1.6, so we'll get it with the merge | 21:12 |
| tjaalton | hm, so there's a patch to wacom that adds WcmNewInputDeviceRequest() | 21:35 |
| tjaalton | might test that some time | 21:35 |
| bryce | tjaalton: oh right, I already mentioned that in the git changelog | 21:37 |
| tjaalton | hehe | 21:37 |
| bryce | I thought that looked familiar :-) | 21:37 |
| bryce | tjaalton: xserver-xorg-video-suncg6 can be sync'd too can't it? | 21:44 |
| tjaalton | bryce: I don't think it would build | 21:46 |
| tjaalton | 4ubuntu1 unincludes xf86Version.h | 21:47 |
| bryce | hmm ok | 21:48 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!