| Sarvatt | damn, forgot about the x11-apps bustage before FF | 00:21 |
|---|---|---|
| Sarvatt | i just just distro patch the current one is the way to go now? | 00:21 |
| Sarvatt | there's a bug where there was a typo in the recommends preventing upgrades | 00:21 |
| Sarvatt | i updated to new versions for packages in debian git but its not released, was going to sync it but it doesn't look important enough for a freeze exception | 00:23 |
| Sarvatt | wow, http://www.bryceharrington.org/Arsenal/Reports/ubuntu-x-swat/totals-precise-workqueue.svg | 00:25 |
| Sarvatt | pretty sure https://bugs.launchpad.net/ubuntu/+source/unity/+bug/926379 is hiz fail that might be fixed in mesa 8.0.1, getting a lot of dupes | 00:26 |
| RAOF | Sarvatt: Yeah. Some of those xorg-server bugs shouldn't be on the workqueue; they're protocol requests. | 00:27 |
| RAOF | But xorg-server's been gathering no moss :( | 00:27 |
| Sarvatt | i'll ask the debian guys if its ok to push 8.0.1 to git in the morning and merge that over, 8.0+ needs some /etc/drirc installation that i dont see in the debian branch yet | 00:29 |
| Sarvatt | or else unigine will be busted and our QA runs those unigine benchmarks | 00:29 |
| Sarvatt | http://cgit.freedesktop.org/mesa/mesa/commit/?h=8.0&id=b5efe0881ed385de36fa2f4889e1dd962990420a | 00:31 |
| RAOF | cnd: Valgrind on the server doesn't appear to point to any smoking-gun; there's some read-after-free in the intel driver, some jump-based-on-uninitialised-values that appear harmlessish, and a whole bunch of ioctl noise, but nothing that would point to memory corruption. | 02:00 |
| === lool- is now known as lool | ||
| Daviey | Hola, anyone noticing oddities with multi-touch trackpads? Selecting text, and left click becoming unusable ? | 09:41 |
| bryceh | yeah | 09:43 |
| bryceh | Daviey, guessing it's due to the new input stack. cnd might know more. | 09:44 |
| Daviey | bryceh: it's making me cry. | 09:44 |
| Daviey | cnd: please help clear my tears.. :) | 09:45 |
| bryceh | Daviey, on my laptops when it happens, then generally desktop swapping or rebooting seems to help | 09:45 |
| Daviey | bryceh: i find leaving it for 30 seconds resolves it, but it's making my workspeed suck :) | 09:45 |
| bryceh | Daviey, mm | 09:45 |
| bryceh | Daviey, tell cnd your LP bug # | 09:45 |
| Daviey | compiz is using 35% CPU :) | 09:46 |
| === yofel_ is now known as yofel | ||
| bryceh | Daviey, well compiz cpu usage != X issue | 09:46 |
| Daviey | bryceh: I'm yet to report a bug, as currently it would be "multitouch mouse does weird things".. not even sure what package to report it against. | 09:46 |
| Daviey | bryceh: right, i know /that/ | 09:46 |
| bryceh | Daviey, yes "weird" is an anti-report word, but if it seems specific to MT, then may well be something cnd cares about. | 09:48 |
| bryceh | if in doubt, report against 'xorg' | 09:48 |
| bryceh | at least that way it'll include the log files an so on we care about. We're not at all shy about reassigning if we don't think it's an X bug. ;-) | 09:49 |
| bryceh | Daviey, a really useful bit of info would be if you can think back to when the behavior first _started_ | 09:50 |
| bryceh | coupled with your /var/log/dpkg.log (which 'ubuntu-bug xorg' typically includes), that can help in isolating what package regressed. | 09:50 |
| cnd | Daviey, what is your trackpad? | 09:51 |
| Daviey | bryceh: sadly, i didn't update granular enough, but certainly within the last 8 days of uploads :) | 09:51 |
| bryceh | although I suspect in this case it's going to finger the X server update (which has an even newer input stack) | 09:51 |
| Daviey | that certainly was in the apt-get update session of updates.. | 09:51 |
| bryceh | Daviey, even 8 days can at least bracket it; we don't update X *so* often. | 09:52 |
| Daviey | bryceh: xorg was certainly in the session of updates i performed | 09:52 |
| Daviey | bryceh / cnd: bug 937654 | 09:55 |
| bryceh | Daviey, thanks | 09:55 |
| cnd | Daviey, my guess is you have entered "locked drag" mode | 09:56 |
| cnd | Daviey, do you have tap to click enabled? | 09:56 |
| Daviey | cnd: hmm | 09:56 |
| Daviey | cnd: forgive me, but how would i check? | 09:56 |
| cnd | Daviey, mouse settings | 09:57 |
| cnd | but you should know :) | 09:57 |
| cnd | if you tap on the trackpad to perform a click | 09:57 |
| Daviey | cnd: i think i tried to enable it last year, but not touched mouse settings since | 09:57 |
| bryceh | Daviey, also, you have some obsolete settings in your xorg.conf; try moving aside your /etc/X11/xorg.conf, or at least commenting out the Driver "kbd" and Driver "mouse" bits. | 09:58 |
| cnd | Daviey, the magic trackpad is very sensitive | 09:58 |
| cnd | it might be that you are accidentally tapping | 09:58 |
| cnd | and when you tap in precise, it locks the mouse button down | 09:58 |
| cnd | you have to tap again to release | 09:58 |
| cnd | this allows for dragging longer than the size of your trackpad | 09:59 |
| Daviey | cnd: yeah, that is the feature i wanted... but not how it was previously.. has it 'improved'? | 10:01 |
| Daviey | bryceh: going to move it out of the way | 10:01 |
| Daviey | brb | 10:01 |
| cnd | Daviey, yeah, I guess it's improved :) | 10:02 |
| bryceh | heya cnd | 10:02 |
| cnd | bryceh, hey | 10:02 |
| bryceh | cnd, glad to be done with the Oregon winter rain? | 10:02 |
| Daviey | cnd: right, seems gently tapping left does break out of the mode.. but as soon as i scroll using two fingers, it re-enters the mode. | 10:05 |
| cnd | RAOF, did you hit the crash while under valgrind? | 11:44 |
| cnd | I have seen an invalid write out of bounds under one valgrind run | 11:45 |
| cnd | however, I didn't have the dbg symbols and I've lost the trace and stuff | 11:45 |
| cnd | so I'm trying to reproduce, but I haven't been able to | 11:45 |
| === yofel_ is now known as yofel | ||
| cnd | bryceh, I have a fix for the random crashers and crashers on disconnect of a magic trackpad | 20:18 |
| cnd | it's a clean fix, but it hasn't been exhaustively tested, and I'll be on flights coming home tomorrow | 20:19 |
| cnd | I would suggest that I upload it, and if you could watch for any breakage and revert that would be good. | 20:19 |
| cnd | what do you think? | 20:19 |
| bryceh | cnd, sure sounds good | 20:20 |
| cnd | k | 20:20 |
| cnd | bryceh, btw, it's a fix in synaptics | 20:21 |
| bryceh | cnd, alright, will keep an eye out | 20:25 |
| RAOF | cnd: No, I haven't hit the crash under valgrind. | 21:54 |
| RAOF | cnd: Oh, I see you've found and fixed it. Superb! | 21:54 |
| sforshee | would it be possible to get someone to take a look at bug 933710? At a fundamental level the problem is caused by the x video drivers not supporting the randr ConnectorType property for outputs, and I'd like to know whether it's likely for this to get fixed in x soon or if the fallback heuristic used by gnome need to be patched. | 23:14 |
| RAOF | sforshee: I'll have a look shortly. | 23:15 |
| sforshee | RAOF, thanks | 23:16 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
