/srv/irclogs.ubuntu.com/2011/08/17/#ubuntu-x.txt

Sarvattlooks like i just hit the upgrade problem too on my machine updated this morning00:31
Sarvattlightdm fails to start00:31
brycehSarvatt, exit code 127?00:31
Sarvattgnome-session: symbol lookup error: /usr/lib/gio/modules/libgiozeitgeist.so: undefined symbol: g_desktop_app_info_launch_handler_get_type00:32
=== yofel_ is now known as yofel
SarvattAmaranth: darnit, you convinced me to buy fusion while i915 doesn't work with these new macbook airs :)04:01
AmaranthSarvatt: hehe04:01
Amaranthit works surprisingly well04:01
Amaranthsupposedly it'll even show you battery status04:01
AmaranthI use it because wifi doesn't work in ubuntu on the latest MBP04:02
Prf_JakobSarvatt: I'm sure I can set you up with a license.04:12
Sarvattalready bought it, appreciate it though man!04:12
Prf_JakobSarvatt: ok, let me know if some ubuntu dev needs any license's, as long as its for vmw related stuff I have been able to get a license issued.04:14
Prf_Jakobbefore*04:14
Prf_JakobNow is time for sleep04:15
SarvattAmaranth: quite a lot of commits enabling the HT PHY broadcoms for b43 in 3.1, it may actually work by 3.204:52
Amaranthyay04:52
Sarvatthttp://www.amazon.com/AirLink101-AWLL5088-Wireless-Ultra-Adapter/dp/B003X26PMO/ref=sr_1_2?ie=UTF8&qid=1313556752&sr=8-2 works great if you dont mind dkms realtek drivers btw :P04:54
AmaranthSarvatt: I have that exact one04:58
AmaranthBut the hassle of using up a USB port plus some oddness with the intel driver and various other things drove me away for now04:59
=== Amaranthus is now known as Amaranth
=== seb128_ is now known as seb128
tseliottjaalton, RAOF: is it expected to have X stuck in a loop (and eventually backtracing) because of evdev in Oneiric?10:58
RAOFtseliot: No, that's a new one on me.10:59
tseliothmm...10:59
tjaaltonnot seen that either11:01
tjaaltonand it's likely a hung gpu anyway11:01
tjaaltonso.. depends on the driver i guess ;)11:01
RAOFRight.  Generally MI: EQ overflow is a hung gpu.11:02
tseliotRAOF: it seems weird to me as I'm using the same NVIDIA driver I was using yesterday and I only updated my system11:03
RAOFIt's entirely possible that something else is broken.11:04
RAOFJust, generally, EQ overflows are GPU hangs.  Because GPUs are the most fragile part of X :)11:04
tseliotI'm switching to nouveau to see if I can start X11:06
RAOFtseliot: You're not being hit by the libzeitgeist-gio craziness?11:10
RAOFIf you've got that installed, it'll break almost everything.11:10
tseliotRAOF: yes, I think I've see that error too when upgrading11:13
tseliotoh, I think it I messed up my system yesterday11:15
tseliots/it//11:15
tseliotRAOF, tjaalton: oh, maybe it's "Desktop failing to start today due to glib2.0 update" (see the ubuntu-devel mailing list)11:24
tselioti.e. https://bugs.launchpad.net/ubuntu/+source/glib2.0/+bug/82775311:24
ubot4Launchpad bug 827753 in glib2.0 (Ubuntu) "libglib 2.29.16-0ubuntu1 breaks desktop session - downgrade fixes (affects: 2) (heat: 16)" [Critical,Fix released]11:24
tseliotoh, same issue11:25
RAOFtseliot: Yup ;)11:25
tseliotRAOF: also, did you add an alternative (with ld.so.conf) on libGLES?11:27
RAOFtseliot: No.11:27
RAOFMy thinking was that libGLES/libOpenVG are going to be tied to their associated libEGL implementations.11:28
RAOFSo the egl alternative suffices.11:28
RAOFIf there *is* any time when you might want to use nvidia's libGLES with mesa's libEGL, then I'd need to rethink that.11:29
RAOFSo what you should do is ship libGLES/OpenVG in the same directory as your libEGL; then the libEGL alternate will take care of them.11:29
tseliotRAOF: I was just thinking of the driver for Tegra on ARM which right now conflicts with mesa-utils, etc.11:30
tseliotyes, I guess that's how it works now11:30
RAOFtseliot: I've dumped some thoughts on https://wiki.ubuntu.com/X/EGLDriverPackagingHOWTO but they need tidying up.11:31
tseliotRAOF: nice, thanks, I'll have a look at it11:33
mterryThere haven't been any updates recently that would cause an intel driver to not start in oneiric, has there?15:30
mterryAlternatively, how do I force use of vga and not the intel driver in X?15:33
jcristauboot with nomodeset15:33
* mterry tries15:34
mterryWell, that worked and got me the VESA driver.  But x still won't start...  /me digs deeper15:38
Duke`wow new fresh xorg-intel driver, with no more image corruptions \o/ has SNA been disabled? (i945)15:40
tseliotmterry: are you sure it's not bug #827753 ?15:51
ubot4Launchpad bug 827753 in glib2.0 (Ubuntu) "libglib 2.29.16-0ubuntu1 breaks desktop session - downgrade fixes (affects: 3) (heat: 20)" [Critical,Fix released] https://launchpad.net/bugs/82775315:51
mterrytseliot, it is, but in a different place.  The known issue was about zeitgeist-gio, but I found a similar problem with wncksyncdaemon using the dropped symbol15:53
tseliotoh15:54
mterrysorry, should have followed up in this chat room  :)15:54
lamalexhey guys, in O my external monitor is no longer found at all18:43
lamalexmakes testing unity/compiz multi-monitor bugs... hard18:43
tjaaltonlamalex: file a bug with 'ubuntu-bug xorg'19:06
SarvattAmaranth: are you using oneiric in fusion?19:06
SarvattAmaranth, Prf_Jakob: http://sarvatt.com/downloads/vmwgfx1.png http://sarvatt.com/downloads/vmwgfx2.png does not like 3D enabled :)19:09
lamalextjaalton, https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/82828619:10
ubot4Launchpad bug 828286 in xorg (Ubuntu) "External monitor not detected on macbook pro 7.1 (affects: 1) (heat: 6)" [Undecided,New]19:10
tjaaltonlamalex: ah, nvidia. you need to configure it with nvidia-settings, or use nouveau instead19:13
Sarvattnote to self: dont enable 2DAccel in xorg.conf for vmwgfx :)19:18
tjaaltonI'd be happy if vmware workstation worked on an oneiric host, but the kernel modules won't compile19:19
tjaaltonwhich is hardly surprising, but means i can't test it19:21
brycehis there a simple command line tool that does an XUngrabPointer action, so if an app has an invalid grab on the cursor you can force it to let go?19:28
tjaaltonhow would you run that?-)19:29
Duke`Sarvatt, did you disable SNA in latest xorg-intel driver?19:29
tjaalton(from a vt, yes)19:30
SarvattDuke`: yep!19:30
brycehtjaalton, like via ssh with DISPLAy set?19:30
Duke`Sarvatt, that works better now :D19:30
SarvattDuke`: was trying to find you on IRC to tell you, ya beat me to it :)19:31
Duke`hehe19:31
Duke`unfortunatly, I haven't found a lot of work for the graphic corruptions19:32
Duke`s/work/activity19:32
Duke`+for some weeks19:32
Duke`x_x19:32
tjaaltonbryceh: yeah, and usually keyboard works so that you can change the vt too19:32
Sarvattits odd I never could reproduce them but I had so many complaints19:32
Sarvattone person was even using the same exact machine I use most of the time19:33
Duke`Sarvatt, I could reproduce 100% some of them. Eg. moving a file on the desktop from one place to another: the icon is totaly corrupted19:33
Duke`note that I'm using a 64-bit install19:33
SarvattDuke`: natty or oneiric?19:34
Duke`natty19:34
Sarvattwhat kernel?19:34
Duke`3.019:34
Duke`wut I had the same problem with 2.6.3919:35
Duke`and 2.6.38 I think19:35
Sarvattso like almost all the complaints are in natty now that i'm looking19:35
Duke`ah !19:36
Sarvatti wonder if it isnt something that was fixed in unity/compiz in oneiric as to why i didnt have problems, no natty machines19:36
Duke`but what would differ on the graphic side from natty to oneiric?19:36
Duke`ah maybe!19:36
Duke`It does not happen with compiz disabled/metacity19:36
Duke`hum yeah there's probably something with compiz19:37
SarvattDuke`: how are you using edgers on natty? do ya not use wine at all?19:37
Sarvatti'm just curious if you have multiarch set up19:37
Duke`I don't use wine currently (haven't used for a year maybe)19:38
Sarvattheck, now that i think about it theres absolutely no way multiarch works in natty, need to multiarch all those mesa depends in there19:38
Duke`mainly desktop work, and some old dosbox games + some games requiering low hardware19:38
Sarvattit doesnt even work in oneiric without libpciaccess multiarched still19:38
Sarvattdo you use x64 flash or something?19:39
Duke`I'll check19:39
Sarvattbecause 32 bit looks for ia32-libs mesa thats totally busted19:39
Duke`seems to be the 32-bits version, since the installed package requires ia32-libs19:40
Duke`and I experiments white graphic corruptions on some flash animations19:40
Duke`s/experiments/have experimented until latest edgers update/19:41
Duke`I'm happy today, no graphic corruption anymore, flash works fine, shadows under windows/menus/panel are back on my desktop \o/19:42
Sarvattyeah thats due to not having any kind of mesa available for 32 bit, ya want to grab x64 flash for sure. i'm not sure that i'm going to fork the packaging for natty to disable multiarch, that'll end up with me not updating natty at all when it breaks most likely (mesa builds still break and need too much time to fix up 2x weekly already)19:42
Duke`dunno how is flash64 stable today, did you test it?19:44
Sarvattyep its good, it'll be released soon as the stable 11 version19:46
SarvattDuke`: what was the unity/SNA bug # for your bug again?20:04
Sarvatton fdo20:04
Sarvattlet me try that again, what the was bug number on fdo for the unity/SNA problem you had?20:05
Duke`#3873220:33
Duke`https://bugs.freedesktop.org/show_bug.cgi?id=3873220:33
ubot4Freedesktop bug 38732 in Driver/intel "[sna unity] windows have white borders from time to time." [Normal,New]20:33
Duke`of course, if it's not reproductible with a recent compiz/unity, xorg developers may not show a lot of activity on this bug...20:34
Duke`gonna sleep, bye20:52
brycehSarvatt, btw in the call with Intel Max sounded interested in getting a list of SNA/unity bugs that are showstoppers for us; would you be able to compile a list of bug#'s?20:53
Sarvattbryceh: sure thing, https://bugs.freedesktop.org/show_bug.cgi?id=38732 is the only fdo one I know off the top of my head but I'll dig up some of the others and ask these people emailing me to file bugs because peope were saying like the whole right click context menu popup was white so you couldn't interact with it instead of just shadows, and other people saying icons and random graphics were corrupted like Duke`21:00
ubot4Freedesktop bug 38732 in Driver/intel "[sna unity] windows have white borders from time to time." [Normal,New]21:00
Sarvatthttps://bugs.freedesktop.org/show_bug.cgi?id=3904421:01
ubot4Freedesktop bug 39044 in Driver/intel "[SNA unity] Desktop is messed up" [Normal,New]21:01
Sarvatthttps://bugs.freedesktop.org/show_bug.cgi?id=3873321:01
ubot4Freedesktop bug 38733 in Driver/intel "[sna unity] drag & drop desktop icon contains uninitialised junk" [Normal,New]21:01
Sarvattbryceh: they actually have them all tagged on fdo21:04
=== yofel_ is now known as yofel
brycehSarvatt, ah good21:07
Sarvattthere are a few duplicates of those 3 but those look like the major ones21:08
SarvattKDE wont be happy when its enabled from the looks of it https://bugs.freedesktop.org/show_bug.cgi?id=3932121:09
ubot4Freedesktop bug 39321 in Driver/intel "[sna kde] swrast on common paths?" [Major,New]21:09
* bryceh forwards21:12
ScottKHow is Intel GMA 3150 support in Natty/Oneiric?22:19
RAOFAs far as I'm aware, pretty stable.22:20
RAOFMight be slightly better in oneiric than natty.22:21
SarvattScottK: probably the most robust of all the intels, outside of the external montior fun generation 3 entails22:21
ScottKDon't care about external monitor for this particular use case, so sounds good.22:21
ScottKThanks.22:21
ScottKWhatever SNA is, it sounds like I don't want it.22:22
brycehyeah don't think I've ever seen a pineview bug report22:22
brycehScottK, actually the performance numbers I've seen with SNA are very exciting22:22
brycehjust seems it has some bugs still to be worked out22:22
ScottKRight, but the breaks KDE part I don't like.22:22
brycehI'm glad Intel is holding off on  it, rather than pushing it hard like we saw with EXA and UXA22:23
brycehScottK, true but who uses KDE?22:23
ScottKI know a few people.22:23
bjsniderthere are 6 people in the world that are confirmed to use kde23:00

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!