=== wgrant_ is now known as wgrant | ||
ebroder | I'd kind of like to try and get the fix for bug #340993 SRU'd into Hardy and Intrepid. Does this seem reasonable? (it's just an additional build-dep and dep) | 01:28 |
---|---|---|
ubottu | Launchpad bug 340993 in alpine "alpine's ./configure wants Build-Depends and Depends: aspell" [Undecided,Fix released] https://launchpad.net/bugs/340993 | 01:28 |
jdong | bryce_: looks like someone already filed my bug indeed (bug 329039) | 05:57 |
ubottu | Launchpad bug 329039 in linux "jaunty alpha 4 live cd hangs on macbook when starting gnome" [Undecided,New] https://launchpad.net/bugs/329039 | 05:57 |
jdong | the symptoms are frustratingly nondescript :) | 05:57 |
maco | jdong: are you able to do that bootchart hack that shows what's happening during gnome startup? | 06:03 |
jdong | I haven't tried; that's not really a priority right now | 06:04 |
bryce_ | jdong: that's a pretty poorly done bug report though; no logs provided, bisection, or other sort of analysis that a dev would need to look into it. | 06:17 |
jdong | bryce_: indeed; what kind(s) of logs would be useful? After a hard reset I don't really see anything worthwhile in /var/log... | 06:18 |
jdong | the closest to bisecting the problem is "it didn't happen in Intrepid"... | 06:18 |
kees | oh excellent, -Os still leaves fortify enabled. | 06:18 |
bryce_ | jdong: you really need to familiarize yourself with https://wiki.ubuntu.com/X/ ;-) | 06:20 |
bryce_ | lots of stuff there that'll help you analyze that bug better | 06:21 |
jdong | I'll see how close I can get by watching over SSH as it hangs | 06:24 |
jdong | bryce_: heh I can log in post-hang; from Xorg.0.log I see exaCopyDirty: Pending damage region empty! | 06:42 |
jdong | then a re-EDID probe | 06:42 |
jdong | is that EXA message suspicious or normal? | 06:42 |
bryce_ | I think it's normal | 06:42 |
jdong | ok, other than that I just see another EDID probe; nothing out of the ordinary | 06:43 |
jdong | Mar 15 06:39:16 ubuntu kernel: [ 413.257002] [drm:i915_getparam] *ERROR* Unknown parameter 6 | 06:43 |
jdong | is that normal? | 06:43 |
bryce_ | no | 06:43 |
jdong | it doesn't coincide with the hang though; just happens whn X.org starts | 06:44 |
jdong | the hang is triggered by compiz --replace | 06:44 |
bryce_ | if it doesn't coincide with the hang, it is probably not relevant | 06:45 |
jdong | ack this livecd booting is way too noisy; I'll do a quick install to disk and try to get a backtrace | 06:45 |
jdong | thanks for your time | 06:46 |
bryce_ | yup | 06:47 |
jdong | bryce_: aha, got a backgrace | 08:34 |
jdong | backtrace* | 08:34 |
jdong | looks identical to bug 327844 | 08:34 |
ubottu | Launchpad bug 327844 in xserver-xorg-video-intel "compiz freezes xorg on intel graphics card" [Undecided,Incomplete] https://launchpad.net/bugs/327844 | 08:35 |
jdong | http://paste.ubuntu.com/131446/ | 08:35 |
jdong | started compiz, first got a SIGPIPE, then continuing I got a hang on ioctl() / drm_intel_gem_bo_start_gtt_access | 08:35 |
* jdong wonders if his obscenely large fonts is a bug or feature | 08:59 | |
Tm_T | jdong: feature, font sizes are over 8 by default | 08:59 |
* Tm_T hides | 08:59 | |
jdong | ah, my DPI was at 112 for some reason. | 09:00 |
jdong | 96 looks "right" on my display | 09:00 |
Tm_T | thay all look wrong to me (: | 09:01 |
jdong | lol | 09:01 |
* Tm_T really hopes to get >160 dpi display some day | 09:01 | |
Tm_T | it's not fun that you have to count pixels when you choose your font settings, or watch those pixels anyway | 09:02 |
jdong | haha it also look like jaunty display brightness setting is in gray coding :) | 09:02 |
=== tkamppeter_ is now known as tkamppeter | ||
wgrant | jdong: Feature. 96DPI just looks reasonable because you're used to it. | 10:36 |
wgrant | LCDs tend to look a lot less bad with the default DPI setting when you use full hinting rather than slight. | 10:36 |
=== Zic_ is now known as Zic | ||
=== asac_ is now known as asac | ||
directhex | is there a "right way" to mount network drives on boot, given n-m, yet? | 12:18 |
* hyperair doesn't think so | 12:19 | |
hyperair | ifplugsomethingorother stuff is in /etc/network/if-*.d | 12:20 |
slytherin | pitti: is there any plan to update ekiga to 3.0.2? It adds settings/data migration from 2.x. | 13:03 |
yao_ziyuan | i want to use update-manager in kubuntu, | 13:40 |
yao_ziyuan | so i removed adept and kpackagekit, | 13:40 |
yao_ziyuan | and modified update-notifier-kde to launch update-manager when updates are available. | 13:40 |
yao_ziyuan | the remaining problem is how to set up a mechanism to periodically check apt for updates | 13:40 |
yao_ziyuan | since update-notifier-kde itself doesn't check for updates | 13:40 |
yao_ziyuan | solved. | 13:41 |
=== `6og is now known as Kamping_Kaiser | ||
=== azeem_ is now known as azeem | ||
joaopinto | hello | 15:24 |
joaopinto | I am trying to automate the console-setup install into a schroot, any tip on how to force it to install without asking for the kbd layout ? | 15:24 |
=== ssweeny_ is now known as ssweeny | ||
=== nick__ is now known as afv- | ||
=== j_ack_ is now known as j_ack | ||
=== nick__ is now known as afv- | ||
bryce_ | jdong, there's a page on the X wiki explaining about the changed font dpi situation | 21:29 |
fatal_ | bryce_: hi... I'm having problems with graphics corruptions since updating to Jaunty. I see you have updated packages in your PPA that failed to build.... are there built ones available somewhere for testing? | 21:44 |
bryce_ | fatal_: for -ati? latest is in the archive now. | 21:44 |
fatal_ | (xserver-xorg-video-intel) | 21:44 |
bryce_ | oh, no it needs a kernel patch first before it'll build. | 21:44 |
fatal_ | ok.. | 21:45 |
jdong | bryce_: any hints on that backtrace I got for my X hang? | 21:57 |
jdong | http://paste.ubuntu.com/131446/ | 21:58 |
jdong | #1 0x00007fd37cf383bd in drm_intel_gem_bo_start_gtt_access () | 21:58 |
bryce_ | jdong, when I was looking at the 2.6.2 cl there were a number of bo fixes, so perhaps that includes a fix for this issue | 22:00 |
bryce_ | jdong, are you using UXA or EXA? | 22:00 |
jdong | bryce_: whatever the default is on the GMA950; I made no changes to xorg.conf; assuming EXA | 22:00 |
jdong | I tried explicitly using UXA and that hung right at GDM | 22:01 |
bryce_ | hmm | 22:01 |
jdong | (a bunch of various kern.log oopses) | 22:01 |
bryce_ | jdong, unfortunately calls #0 #1 and #2 in your trace don't have symbols. Maybe you forgot to install the dbg package for libdrm? | 22:02 |
jdong | bryce_: could very well be; let me verify and augment the trace | 22:02 |
jdong | ah, -intel-dbg evaded me. | 22:03 |
jdong | bryce_: ok now you're gonna question my sanity. | 22:12 |
jdong | compiz loaded without crashing this time... | 22:12 |
jdong | albeit horrifically laggy | 22:12 |
jdong | *investigates* | 22:13 |
ScottK | jdong: I don't think anyone is confused about your sanity. | 22:14 |
jdong | :P | 22:14 |
jdong | I suspect a suspend/resume cycle did something. Let me reproduce. | 22:15 |
jdong | *cry* why is life so unfair; all I did was remove appletouch. | 22:18 |
jdong | and some USB autosuspending hacks. | 22:18 |
jdong | *sets off to reproduce bug* | 22:18 |
jdong | *sets off to reproduce bug* | 22:19 |
jdong | err | 22:19 |
jdong | ok only on freshboot blood. | 22:19 |
bryce_ | heisenbug | 22:19 |
jdong | maybe it only works when I DON'T have a full symbol gdb stack ready ;-) | 22:21 |
jdong | ok got the hang to happen | 22:23 |
jdong | it doesn't seem to be using the intel debugsyms though? | 22:24 |
jdong | still get No symbol table info available. | 22:24 |
jdong | on /usr/lib/libdrm_intel.so.1 with libdrm-intel1-dbg installed | 22:24 |
jdong | PEBKAC | 22:28 |
jdong | wait | 22:28 |
jdong | bryce_: any hints on not being able to get symbols on libdrm_intel.so.1 and i915_dri.so? I've got their -dbg packages installed | 22:29 |
bryce_ | hmm, that should do it | 22:31 |
bryce_ | jdong, I assume you restarted X? | 22:32 |
jdong | I am pretty sure I did | 22:33 |
jdong | let me reboot the hung system to be sure. | 22:33 |
bryce_ | I'm not seeing any errors in the libdrm rules script, I'm fairly sure this should all be there | 22:34 |
jdong | bryce_: getting closer, got a traceback but this one looks very different | 22:38 |
jdong | (I'll go grab drm symbols too, grumble) | 22:38 |
jdong | http://paste.ubuntu.com/131747/ | 22:39 |
bryce_ | jdong, maybe you should try seeing if Option "AccelMethod" "UXA" gives better stability for you | 22:40 |
jdong | bryce_: IIRC that gave me nasty kernel oops hangs | 22:41 |
jdong | I can try that next if you don't feel these backtraces are useful | 22:41 |
jdong | *tries UXA* | 22:41 |
bryce_ | jdong, well, without the libdrm symbols it's hard to troubleshoot. If it turns out to be an -intel EXA bug, upstream likely won't care about it, since they're moving on to UXA anyway | 22:44 |
jdong | bryce_: compiz didn't crash but I am getting one frame every 8-10s | 22:45 |
jdong | Mar 15 18:43:36 blackbook64 kernel: [ 108.203351] [drm:i915_getparam] *ERROR* Unknown parameter 6 | 22:45 |
jdong | got several of those again in dmesg | 22:45 |
jdong | (X does not seem to be CPU spinning at all) | 22:45 |
jdong | Xorg.0.log shows nothing of interest happening | 22:45 |
jdong | http://jdong.mit.edu/~jdong/broken_macbook_uxa.txt full log if that helps | 22:47 |
jdong | bryce_: SUCCESS -- turning off Sync to VBlank solves the performance issue | 23:10 |
bryce_ | jdong, excellent | 23:12 |
bryce_ | hmm, thought tjaalton had switched that off by default | 23:13 |
jdong | bryce_: hmm didn't seem to happen on my jaunty alpha6 stock install. | 23:13 |
jdong | bryce_: only bug is the display crashes on resume, I think I saw that bug filed | 23:13 |
geofft | directhex: (re 08:18) still around? | 23:35 |
directhex | hm? | 23:35 |
geofft | did you get your question about network manager on boot answered? | 23:37 |
directhex | no | 23:38 |
directhex | i reckon if-up.d/mountnfs is doing as it's supposed to though | 23:39 |
directhex | assuming the script is executed when nm brings up the interface (the script works, i haven't tested that it's triggered when it should be) | 23:39 |
geofft | I think that works | 23:40 |
directhex | unmount is another story | 23:40 |
directhex | cifs umounts after the network has disappeared involves a rather lengthy delay | 23:41 |
geofft | so, when I had a machine with wireless that needed to do networked login | 23:41 |
geofft | I wrote http://web.mit.edu/geofft/Public/start-network.py. It's a horrible hack but it sufficces | 23:41 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!