/srv/irclogs.ubuntu.com/2011/03/08/#ubuntu-x.txt

RAOFkees: Looks like that might be fixed in the most recent drm-fixes pull request? (So, not yet in any kernel that you'll have access to, but still… ☺)00:01
bryceh_kees, I can repro it but haven't had time to poke around with it yet00:13
bryceh_RAOF, is the drm-fixes branch built by the kernel team?00:13
bryceh_http://kernel.ubuntu.com/~kernel-ppa/mainline/00:14
bryceh_looks like not, but they have builds of drm-next and drm-intel-next, - think the fix is in one of those?00:14
RAOFI don't think so, but let's play chase-the-git-commit!00:17
RAOFbryceh_: No, not in drm-next, but is in drm-intel-next.00:26
bryceh_RAOF, what's the commit id(s)?00:30
RAOFbryceh_: 467cffba8 (ooh!  gen2 *and early gen3*), a1656b, and 0ee537abbd00:31
bryceh_thanks00:31
bryceh_heh, I just finished saying my 945 never freezes, and then guess what00:32
RAOF:)00:32
RAOFDing!00:32
bryceh_so maybe for the betterment of our testing, we just need to get a lot more brash and braggy about how bugfree our hw is?00:33
RAOFMy GM45 *never* hangs!00:33
trinikronovesa never hangs!00:37
RAOF(Because it's never used ☺)00:37
bryceh_feh, of course now I can't seem to reproduce it00:43
bryceh_kees, assuming you've still been seeing the bug with the current kernel, please load the drm-intel-next kernel and run that for a day or two and see if you can verify the issue is missing there00:44
keesbryceh_: I'll try, I'm in the middle of debugging some other ubuntu-specific kernel changes :P00:54
bryceh_yeah I'm in the middle of patch pilot work but can bang on it tomorrow or wednesday probably00:56
bryceh_(if I can ever reproduce it reliably again)00:56
RAOFYou know one of the biggest things I miss in git? from bzr  Being able to reference tools by their unique prefixes (bzr st, etc)02:26
bryceh_RAOF, yeah02:40
bryceh_been rockin' the bzr today with patch piloting02:40
RAOFI think mesa 7.10.1 is ready for upload, by the way.02:41
RAOFI also think that there are some cherry-picks that will fix bug #721881 (or at least turn it into a non-segfault)02:41
ubot4`Launchpad bug 721881 in xserver-xorg-video-intel (Ubuntu) "compiz crashed with SIGSEGV in intel_region_data() (affects: 1) (heat: 246)" [Medium,New] https://launchpad.net/bugs/72188102:41
bryceh_RAOF, how much testing has it gotten so far?02:42
bryceh_i.e. is it ready to go now, or should I test it on my own hw first (which will have to wait 'til tomorrow)02:43
RAOFIt's had a couple of day's testing on my hardware.  It wouldn't hurt to test on yours, too.02:45
bryceh_alright; better safe than sorry, I'll todo it for tomorrow02:45
RAOFI'll ensure it's in cooperteam.net/Packages along with binaries.02:47
bryceh_ok coo02:47
bryceh_wow, we're getting wayland bug reports - bug #72961402:48
ubot4`Launchpad bug 729614 in wayland (Ubuntu) "wayland-compositor crashed with SIGSEGV in wl_event_loop_dispatch() (affects: 2) (dups: 1) (heat: 18)" [Medium,New] https://launchpad.net/bugs/72961402:48
bryceh_someone has actually used it :-D02:48
* bryceh_ wanders off02:49
bjsniderbryceh_, now...you have to fix the bug02:50
bryceh_heh02:51
bryceh_maybe if I get bored between now and release day02:51
bryceh_actually it's probably already fixed upstream, there's probably a patch that can be backported02:51
bryceh_hrm, Mr. Spikey McSpike has returned to the graph...  http://www.bryceharrington.org/X/Reports/ubuntu-x-swat/totals-natty-workqueue.svg02:52
bryceh_meh, just dupes of known gpu lockups.  sure wish we could get those solved02:53
* bryceh_ really wanders off02:56
LLStarkshmmm. does sandy bridge optimus mean that the framebuffer for the discrete gpu would be on the cpu?04:56
RAOFNo, because the CPU doesn't have that kind of memory on it.05:10
RAOFUnless you feel like taking up many megabytes worth of L3 cache with your framebuffer that you'll need to copy out to system memory anyway :)05:10
tjaaltonbryceh_: evtouch doesn't even build, and should be removed from the archive imo :)06:21
tjaalton(re: the recently added patch)06:21
tjaaltonoh goodness, new unity that should fix the crashing decorator every time I send an email from tb06:23
bryceh_tjaalton, ah, I knew it was obsolete.  Yeah we should get rid if it if it'll never build again07:09
tjaaltonwell it's patchable, but it's gone from debian too, and no upstream..07:10
LLStarksisn't deprecation wonderful?07:24
tjaaltonyep07:24
tjaaltonbryceh_, RAOF: thoughts about merging xorg? I haven't looked at it too closely, but at least dexconf would get removed, finally07:26
bryceh_tjaalton, would be a good idea, not sure how we've missed it07:43
tjaaltonbryceh_: I'll prepare it today07:46
tjaaltonbeen a while..07:46
bryceh_tjaalton, last week I happened to clean up most of the bugs filed against xorg - https://bugs.edge.launchpad.net/ubuntu/+source/xorg07:47
bryceh_pretty much all the ones with an importance set seem to be things relevant to the xorg package now07:48
tjaaltonyeah I noticed, good stuf07:49
tjaaltonf07:49
tjaaltonbryceh_: the next UDS we might discuss about splitting failsafe/apport from xorg, so they could be maintained separately, to minimize the churn (and keep them on lp if appropriate)?07:50
tjaaltoniirc that has been planned at some point07:51
lagCan any of you tell me what a CMAP (colour map) is please?07:51
bryceh_tjaalton, in fact we already have a blueprint on it.  I'd planned to work on it this cycle but just didn't get to it (work on wayland took a bit more time than expected).  I can make it a focus next cycle.07:53
tjaaltonbryceh_: ah good07:54
bryceh_in fact there already is a package with these files in universe - 'xdiagnose'07:54
bryceh_so mostly is just a matter of setting up the appropriate rules and so on to transition us, and do plenty of testing07:55
tjaaltonyeah07:56
RAOFbryceh_: I don't have strong opinions on merging xorg.  If tjaalton wants to do it, great :)08:40
RAOFbryceh_, tjaalton: If you're interested in testing mesa 7.10.1 it's in http://cooperteam.net/Packages (i386 and amd64)08:50
tjaaltonRAOF: ok, I hear there are lot of intel changes, so I'll load it on my laptop..08:51
RAOFYup.08:51
RAOFIt's almost all intel fixes (mainly for sandybridge)08:51
RAOFBut also a bunch of glsl, etc.08:52
RAOFOh, and I pulled in a patch to make nvaf not kill X. 08:53
lagMorning all09:50
lagI just conducted a dist-upgrade to Natty09:50
lagWhich removed my Nvidia driver09:50
lagWhen I try to install it again, apt attempts to uninstall 23 packages - what's the latest?09:51
tjaaltonnvidia-current is at 270.2909:52
lagIt was nvidia-current I tried to install09:53
lagBut it wants to remove btrfd-tools, libclutter, python things, and a load of libs09:53
tjaaltonnot because of nvidia09:54
lagThe following packages have unmet dependencies: 09:55
tjaaltonplease use pastebin09:56
lagxserver-xorg-core: Breaks: xserver-xorg-video-8 which is a virtual package09:56
lagI can't, I'm not typing any more out :)09:56
lagit's on a console, no copy-paste09:56
tjaaltoninstall pastebinit :)09:56
tjaaltonanyway09:57
tjaaltonsure about not having some ppa's around?09:57
tjaaltoncould be a stale mirror too09:57
lagOddly, my sources file still says Maverick?10:01
tjaaltondidn't use do-release-upgrade?10:01
lagI didn't initiate it10:01
tjaaltonoh, who did?-)10:02
lagThe Update Manager10:02
tjaaltonyou said "dist-upgraded"10:02
tjaaltonso I thought you ran apt-get by hand10:02
lagYes, it did a dist upgrate10:02
lagupgrade*10:02
tjaaltonwell the upgrader probably failed and it fell back on maverick?10:03
lagIt said it has competed successfully 10:03
tjaaltonrun do-release-upgrade -ed10:03
tjaaltonuh, -d10:03
lagI've just initiated another full upgrade10:04
tjaaltonso sources.list is full of maverick, or both m and n?10:04
lagThere were no m10:04
lags/m/n10:04
tjaaltonprobably ask mvo how that could've happened10:05
mvolag: could you mail me the logs in /var/log/dist-upgrade/* (or file a bug) please?10:18
lagmvo: I'll take a look when my (second) upgrade has finished10:20
mvook10:20
lagtjaalton: mvo: Same deal10:30
tjaaltonRAOF: can't connect to your server10:43
tjaaltonlag: file a bug against update-manager then (ubuntu-bug update-manager)10:43
lagWoooooooooooo10:44
lagI'm in 10:44
mvoto diagnose this, we need the logs10:44
lagWhat a nuisance 10:44
RAOFtjaalton: Try again?10:45
lagmvo: One step at a time10:45
tjaaltonRAOF: yeah, works now10:45
lagmvo: Okay, I still have issues10:59
lagmvo: What logs are going to be of most help to you?10:59
lagdmsg, gdm10:59
mvolag: if you used update-manager or do-release-upgrade for the upgrade /var/log/dist-upgrade/* will contain what is needed11:02
* mvo will be away for lunch for a couple of minutes now11:02
asacRAOF: hey ... do you know if "00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c)" gallium is now working in natty or xorg-edgers?11:22
* asac would like to have that ;)11:22
RAOFYou want gallium on that?  I don't believe 965g has seen more than a handful of commits since last I tested it, and then it immediately hung the GPU pretty much regardless of what I threw at it.11:23
RAOFNow, if you had a slightly older GPU, i915g is getting to the point where we might want to ship it in dri-experimental :)11:23
lagmvo: You are now in possession of my logs11:24
RAOFasac: Why do you particularly want gallium support for that?  egl_dri2 should give (complete?) GL|ES1 and GL|ES2 support with the classic driver.11:25
RAOFBut not OpenVG, and I guess there may be differing levels of extension support.11:25
asacRAOF: hmm. are there packages for maverick? guess not right?11:25
RAOFUUuuuum… dunno :)11:25
asacRAOF: i want GLES2 accell ... not more11:25
RAOFWorks in Natty, then :)11:25
asacRAOF: if there are other ways to achieve that i ma happy11:26
RAOF(And should in xorg-edgers)11:26
asacRAOF: maverick in xorg-edgers11:26
asac'11:26
asac?11:26
RAOFJust checking.11:26
RAOFYup, I think so.  xorg-edgers has a shiny trunk snapshot of mesa, and should build egl_dri2.11:27
asacok what was the script that helps downgrading to archive version for that stuff again?11:27
* asac wants to be save11:27
RAOFppa-purge.11:27
RAOFAvailable in xorg-egders and also in Natty universe :)11:28
asacthanks ... i will try ;)11:28
brobostigoni was talking to someone last about trying to confirm 730099, and trying a non-ubuntu kernel, howeer it reminded me, that only a couple of weeks ago, i had debian sid on my eeepc, and no gpu lockup problems, so it confirms what wnted to test, because of my memory failiure.11:33
lagmvo: Are you back from lunch yet?12:42
lagmvo: I'm guessing this may have something to do with my issue: Xorg:1258 conflicting memory types c0000000-c0960000 write-combining<->uncached-minus12:42
tjaaltonumm, a failure to dist-upgrade from 10.10 has little to do with X12:45
mvolag: hello, the logs have some uncommon pattern, I wonder what was triggering the conffile prompt for /etc/lsb-release ? did you modify this file manually? it seems the actual problem is a postinst failure from seamonkey-browser, but the error message is very odd (/var/lib/dpkg/info/seamonkey-browser.postrm): No such file or directory)12:49
lagI haven't modified lsb-release, only read it (to make sure the update was successful 12:50
lag)12:50
mvolag: I can not say much about why gdm does not start anymore, but the error look not good :/12:50
lagI did manual edit /etc/apt/sources.list12:50
lagI s/maverick/natty12:51
tjaaltonmvo: a case of the "top of the list" dpkg bug? (the one with gazillion dupes)12:51
mvolag: is ther a /etc/lsb-release.dpkg-old  or something like this still left? if so, could you pastebin it please?12:51
lagBut that was to 'fix' the already occurred problem12:51
mvotjaalton: well, the seamonkey one is the origin of the failure cascade, but it is a odd failure12:51
lagShould I try to uninstall it and do another upgrade?12:52
lagWhere did you see the failure?12:52
mvolag: I was looking at apt-term.log12:53
lagtjaalton: It must have something to do with it12:53
lagtjaalton: I upgrade, x stops working12:53
mvohm, actually this looks like a old log 12:53
lagOkay12:53
tjaaltonlag: but because of the incomplete upgrade, not due to some drive bug12:55
lagtjaalton: Well I've never seen it before12:55
lagtjaalton: The kernel has been upgraded too12:55
tjaaltonof course12:56
lagMore timestamps and less DOS EOLs needed in the logs me thinks12:57
lagOh I see, they're not EOLs they're backspaces12:58
cndlag, the device.event file you sent me is empty13:49
cndat least, that's what thunderbird tells me13:49
cndI don't know whether to believe it or not :)13:49
lagcnd: It is on my desktop too13:51
cndlag, that's odd13:51
lagcnd: I can't create another one just yet, as I'm having major dramas with my Desktop13:51
cndlag, when you go to create it, chvt to a console13:52
lagcnd: I don't know why it would be empty though, I followed the instructions very carefully 13:52
cndsome X drivers, synaptics in particular, grab the event device node13:52
cndoh wait, that's in the instructions13:52
cndhmm13:52
cndlag, you did touch the screen to generate events, right :)13:52
lagNo?13:52
cndoh, I guess I need to make that clear :)13:52
lagThat'll be it then :)13:52
lagI guess you did :)13:53
cndyou need to generate events for the recording13:53
lagI'll do it again, when I can log back in to my Desktop13:53
lagmvo: Any advice on that?13:53
cndlag, ok, thansk!13:53
tseliotSarvatt: when are we going to have xserver 1.10 RC3 in Natty? (just to know when I should upload nvidia)14:51
Sarvatttseliot: could do it today if you want to sponsor ~50 packages :P xorg-edgers is up to date14:56
tseliotSarvatt: 50 might be a bit excessive, given my todo list ;) Good to know that it's in xorg-edgers14:57
Sarvatti'm not sure though, need to ask RAOF and bryce and see if they have any plans yet, I've had my head in the sand doing other stuff14:59
Sarvattit's a pretty painless transition though, 2 xserver patches upstream already and one that needed an easy refresh, everything else just needs rebuilds15:01
tseliotah, that's good15:06
Sarvatthaven't been able to reproduce any kind of text corruption on gen3 intel darnit15:54
tjaaltonRAOF: the mesa seems to work fine, no gpu hangs on 965GM unlike with 7.1016:33
tjaaltonthe _new_ mesa, that is16:36
tjaaltonbryceh_: btw, you haven't pushed the latest xorg changes to git16:41
tjaaltonwhee, waltop tablets work just fine on natty, need to switch 50-wacom.conf to match it17:24
bryceh_tjaalton, pushed18:23
Sarvattany objections to doing something like this for ati to fix jcastro's bug? http://sarvatt.com/downloads/patches/palm_disable_pageflip.patch18:59
Sarvatthttps://bugs.launchpad.net/xserver-xorg-driver-ati/+bug/71533019:00
jcastroSarvatt: it has some issues resuming from suspend, but I haven't investigated19:04
cndbryceh_, do you know when the next xorg-server upload will be?19:37
Sarvattdo we need a FFe for it?19:42
bryceh_Sarvatt, yeah probably 19:52
bryceh_cnd, within the next week or so probably19:52
cndbryceh_, ok, I've pushed some bug fixes and am working on one last issue right now19:53
cndthey don't have to go in asap, so it can wait19:53
bryceh_cnd, oh thought you meant when were we going to do the xserver abi bump19:53
cndI mean package upload for xorg-server19:54
cndwhether or not it includes any abi bumps19:54
=== ajmitch_ is now known as ajmitch
bryceh_cnd, ok thanks for letting me know. 19:55
Sarvattcnd: looks like we're on the same flight to/from budapest along with the rest of the world19:56
cndSarvatt, cool :)19:56
cndlet me see what seat I am19:56
SarvattI dont have a seat yet, just booking it now19:58
cndSarvatt, I'm at 25G both ways19:59
cndSarvatt, though I doubt you're on the same flight as me going to budapest19:59
cndI'm going for the design sprint the week before too :(19:59
Sarvattoh whoops, didn't look at the date19:59
cndyeah, tricky tricky :)20:00
Sarvattwell shoot, I accepted it because I thought you and vanhoof would be on it but you're both going early, whoops :P20:01
Sarvattseat in front of you has power but you don't, boo20:01
cndreally?20:03
Sarvattcnd: at least it isn't 1 week each in shanghai then portland then 2 weeks in budapest like vanhoof has to do :D20:03
Sarvatthttp://www.seatguru.com/airlines/American_Airlines/American_Airlines_Boeing_767-300_B.php20:04
cndSarvatt, oh well... I don't usually try to do work on these flights20:04
cndtoo hard20:04
cndand I should be sleeping on the flight over20:05
bryceh_cnd, xserver uploaded21:37
cndbryceh_, oh, thanks!21:38
LLStarksis -nv gone from the default archive now?21:47
tjaaltonLLStarks: not yet21:51
tjaaltonbryceh_: thanks!21:51
RAOFSarvatt: Yeah, that disable-pagefilp patch looks like a good idea.21:56
bryceh_so, I've lost track a bit, but is the latest -nvidia support the latest xserver abi bump?22:34
bryceh_if that's the case, then should we think about updating to the xserver with this abi now?22:34
RAOFIf it does, then yeah.  We should move to 1.10 final.22:34
bryceh_RAOF, do you want to go ahead and prep the package upload?22:35
RAOFYup.22:35
bryceh_I'll find out what the -nvidia situation is22:35
bryceh_I think we're good, but just to doublecheck22:36
tjaaltonthat should then wait for tseliot to prep nvidia?22:36
tjaaltonhe was asking for the schedule this afternoon22:36
bryceh_tjaalton, ahh, ok22:36
bryceh_tjaalton, yeah maybe we should set up a staging ppa for this22:36
bryceh_then we can put all the bits and pieces together in there and do test upgrades22:37
tjaaltonor prepare the packages and we'll upload them tomorrow when we're around22:37
bryceh_right.  my earlier statement was ambiguous, I was meaning the thinking should be done now, not the upload ;-)22:38
tjaaltonyeah, sure22:38
RAOFWe should certainly coordinate with tseliot.  After all, that's why we haven't updated already :)22:39
bryceh_ok, looks to me like everything is good to go on the proprietary driver front, but I'll shoot off another coordination email with tseliot and everyone22:56
RAOFAnd I'll stop trying to debug this unity deadlock and get on the packaging.\22:57
bjsniderRAOF, the 270.30 driver works with 1.10 final, and tseliot is waiting for that to be uploaded before he send the blob in. 23:20
bryceh_alright, I've emailed tseliot and cc'd raof and others23:20
bryceh_RAOF, I think we can just reuse https://edge.launchpad.net/~ubuntu-x-swat/+archive/ppa for staging purposes23:21
bryceh_this is pretty much what was in mind when that was set up originally anyway23:21
tjaaltoncan the packages get pulled from the ppa to the archive, or would it need another round of uploads?23:22
=== yofel_ is now known as yofel
bryceh_tjaalton, I think we'll have to do another set of uploads23:23
tjaaltonsince if it's the latter, I'd just shove them in the archive. the diff to the current stack is small23:23
bryceh_it's more the abi change I'm concerned with than actual bugs23:23
tjaaltonrebuilds take care of that, and versioned depends23:24
tjaaltonversioned build-depends23:24
bryceh_yeah, but in theory that should have worked last time too23:24
tjaaltonwell there are packages we are not in charge of23:24
tjaaltonbut -video-all doesn't depend on them (like vbox)23:25
tjaaltonor maybe I'm missing something :)23:25
bryceh_tjaalton, RAOF and I found that when we bumped the ABI going from 1.9 to 1.10 that even though we both tested the individual packages and .debs thoroughly, there were bugs we missed because we installed the .debs individually, resolving dependencies by hand.23:27
bryceh_so we felt it would be a better practice and not that much additional effort, to stage the packages in a ppa first, so we can easily do test upgrades23:28
bryceh_tjaalton, it seems that as ubuntu has evolved, the tolerance for cowboying in xserver changes has decreased a lot since a couple years ago ;-)23:28
tjaaltonmeh :)23:29
bjsnidergee, i wonder why?23:32
tjaaltonthose bastards breaking blobs again?-)23:33

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