| dbro | nouveau has similar - but milder - behavior. Memory usage increases by 0.2MB when I minimize and restore a system monitor window. | 00:10 |
|---|---|---|
| dbro | virtual memory increases by a bit more than 2MB | 00:10 |
| dbro | can anyone else see this same behavior? | 00:15 |
| dbro | is there more information I can provide? eg xrestop or something like that? | 00:29 |
| RAOF | dbro: And does that continue to increase indefinately? | 01:59 |
| RAOF | dbro: Other interesting datapoints include dmesg & Xorg.0.log. | 01:59 |
| dhillon-v10 | bryceh: can you have a look at this bug: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/429384 it is reported against karmic alpha 5 so does it make sense to ask the reporter to check if this problem exists in lucid devel. release and mark the bug incomplete or should I just leave this one alone | 02:17 |
| ubottu | Ubuntu bug 429384 in xserver-xorg-video-intel "[i965] [karmic i386 alpha 5] X hangs (intel 965gm)" [Undecided,New] | 02:17 |
| bryceh | dhillon-v10, yes, we probably will only care about lucid bugs here on out | 02:20 |
| dhillon-v10 | bryceh: alright so can I'll mark others with karmic tag incomplete :) | 02:21 |
| bryceh | dhillon-v10, thanks, that'd be a big help | 02:22 |
| bryceh | dhillon-v10, btw once the original reporter has tested it on lucid, tag it 'lucid' | 02:23 |
| bryceh | dhillon-v10, when I go through X bugs, I only look at ones tagged 'lucid' now | 02:23 |
| dhillon-v10 | bryceh: np :) oh yeah I'll keep that in mind :) | 02:23 |
| bryceh | dhillon-v10, generally I don't count it as tested against lucid if some other random person tested it, since they usually haven't provided the full collection of needed info that the original reporter gave | 02:24 |
| bryceh | and plus, who knows if the person *actually* had the same bug, or just something that seemed similar to them | 02:24 |
| bryceh | but use your judgment there | 02:25 |
| bryceh | usually we want the other person to report their issue as a separate bug report. We have to do that anyway if we upstream the bug. | 02:25 |
| dhillon-v10 | bryceh: alright, thanks for the advice | 02:28 |
| dbro | RAOF: it increases only when I perform those actions. The increment happens when I trigger it, not simply due to passing time | 02:31 |
| dbro | RAOF: what should I look for in dmesg? I can't see anything since bootup | 02:32 |
| dbro | RAOF: and what should I look for in Xorg.0.log? No obvious errors apparent there either | 02:33 |
| RAOF | dbro: Various forms of nouveau kernel debug spew; also it would allow me to determine a couple of things about your setup such as your card generation, whether it needs firmware, whether it *has* firmware, etc. | 02:33 |
| RAOF | dbro: Just pastebinning them would be good; I can easily check through them for items of interest. | 02:34 |
| dbro | ok, should I put it in a pastebin page or something like that? | 02:34 |
| dbro | ok | 02:34 |
| dbro | RAOF: re Xorg.0.log - how many lines should I include? It seems to go back to before the current boot | 02:38 |
| RAOF | dbro: All of them will do. If you've got “pastebinit” installed, just go “dmesg | pastebinit” and “pastebinit /var/log/Xorg.0.log”. | 02:39 |
| dbro | RAOF: ok. http://www.pastebin.com/m39087e9d | 02:40 |
| dbro | and the dmesg is here: http://www.pastebin.com/m43f96c9f | 02:41 |
| dbro | and my xorg.conf is http://www.pastebin.com/m764f40e5 | 02:43 |
| dbro | (it's got a lot of commented lines... sorry about that) | 02:43 |
| RAOF | THat pastebin seems to be empty? | 02:45 |
| dbro | uh oh... | 02:46 |
| dbro | first one? | 02:46 |
| RAOF | The last 2. | 02:46 |
| dbro | sorry- I should't have included the "www." in those 2. | 02:47 |
| dbro | so they are http://pastebin.com/m43f96c9f | 02:47 |
| dbro | and http://pastebin.com/m764f40e5 | 02:47 |
| RAOF | Ta. I'm off for a while, but I'll have a look at them when I get back. | 02:47 |
| dbro | thanks, I'll keep an eye on this window | 02:48 |
| dhillon-v10 | bryceh: mission accomplished :) the only bugs left in new state are in lucid | 03:05 |
| dhillon-v10 | bryceh: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bugs?search=Search&field.status=New | 03:06 |
| bryceh | dhillon-v10, nice | 03:08 |
| dhillon-v10 | bryceh: yeah around 33 left, now time to fix them :) | 03:08 |
| bryceh | actually we'll probably need to have them test against the 2.6.33 drm; quite likely their bugs are fixed there | 03:09 |
| bryceh | so if we end up bringing that in, we'll be able to close the bugs | 03:09 |
| RAOF | dbro: Hm. Interesting. How much does your graphics actually work again? :). You've got the nvidia binary driver loading as well, which is unlikely to improve matters :) | 04:46 |
| RAOF | dbro: Also, you've got quite an old nouveau kernel module which isn't using KMS. | 04:47 |
| RAOF | That's not necessarily a problem, but it might be fixed in a newer version. Also, you've got the same chipset as me, so I'll see if I can reproduce it. | 04:50 |
| RAOF | dbro: I can't reproduce your memleak: I've minimised & restored a system monitor window about 20 times now; memory usage has oscilated between 32MiB and 31.9MiB. Virtual size has remained static at 154.7MiB. | 04:57 |
| RAOF | dbro: It might be interesting for you to try a Lucid alpha 3 livecd; that should have nouveau by default so would be a good environment to see if your bug has been fixed. | 04:57 |
| dbro | ok. is that available now? | 05:14 |
| dbro | how can I get the extra nvidia driver to not load? | 05:16 |
| RAOF | Alpha 3 is not yet out; you could either blacklist the nvidia driver or remove the nvidia-source package (I'm not sure what Ubuntu version you're running, so I'm not sure what the exact package name will be). | 05:19 |
| dbro | RAOF: thanks for the suggestions. I removed the nvidia source package, then rebooted. there was no mention of the nvidia driver in dmesg, only nouveau was loaded. and now I see no change in Memory usage by x when minimizing and restoring the system monitor window. However, I do see an increase in xorg's virtual memory usage when minimizing and restoring the system monitor window. | 06:24 |
| dbro | about 2MB per minimize,restore cycle | 06:24 |
| dbro | RAOF: should I try to use a more recent version of nouveau, eg by following these instructions? https://edge.launchpad.net/~xorg-edgers/+archive/nouveau | 07:04 |
| RAOF | dbro: You could give it a whirl. Of course, I'd be more interested in your Lucid experience (once Alpha 3 has landed), since that's what we're going to be supporting :) | 07:07 |
| dbro | upgrading to the lucid versions of the nouveau packages seems to involve a lot of other package removals and upgrades- is there a way to avoid that? | 07:08 |
| RAOF | You don't want to do that - I was thinking more a livecd situation. | 07:09 |
| dbro | ok | 07:09 |
| dbro | I will try the live cd -- expected to be available feb 25, right? | 07:09 |
| * RAOF rechecks the release schedule. Why isn't it in my google calendar? | 07:10 | |
| RAOF | Yup. 25th, all things going smoothly. | 07:11 |
| dbro | thanks again. good night! | 07:11 |
| tjaalton | bryceh: you need to take nvidia-graphics-drivers a separate entry on the bug chart, it's starting to make the "Remainder" look bad/worse :) | 09:43 |
| tjaalton | serial wacom support for xserver now in git | 11:01 |
| === wgrant_ is now known as wgrant | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!