[06:17] <mlankhorst> ah
[07:45] <tjaalton> sigh, fed up with kubuntu users filing useless bugs because it doesn't have xdiagnose installed
[07:47] <mlankhorst> make it part of kubuntu-desktop then :p
[07:48] <tjaalton> xdiagnose depends on gtk
[07:49] <tjaalton> so that's not an option before it's split
[07:50] <mlankhorst> add 'write qtdiagnose' for todo in r then :-)
[07:51]  * mlankhorst wonders if he should
[07:51] <tjaalton> split the frontend and write a new one, yes
[07:52] <tjaalton> doesn't ubuntu-desktop pull qt anyway?-)
[07:52] <mlankhorst> well I like qt, I guess that means I'll volunteer
[07:52] <tjaalton> great!
[07:52] <mlankhorst> nah but it should be possible to split it up in a qt and gtk frontend
[07:53] <tjaalton> right
[07:55] <xnox_ac100> hello, I get desktop lock-ups ~5-15 minutes - can't do anything even switch to tty. I filed bug 1061441 and I am willing to do anything to get a working desktop.
[07:56] <mlankhorst>  dmi.chassis.asset.tag: To Be Filled By O.E.M.
[07:56] <mlankhorst> always the greatest vendor!
[07:57] <tjaalton>  UpgradeStatus: Upgraded to quantal on 2012-01-04 (273 days ago)
[07:57] <tjaalton> whoa
[07:57] <mlankhorst> that just means their system clock is off :p
[07:58] <mlankhorst> mine would travel years into the future if i repeatedly suspended/resumed it
[07:58] <tjaalton> xnox_ac100: when did it start?
[07:58] <tjaalton> smells like bad hw
[07:58] <mlankhorst> xnox_ac100: does a precise livecd work?
[08:00] <xnox_ac100> tjaalton: i work on foundations team on ubiquity. I needed python3 bits of various software very early on. But I did upgrade *after* UDS not before =))))
[08:00] <mlankhorst> tjaalton: argh all bugs are going down except intel :((
[08:01] <xnox_ac100> tjaalton: it started.... after yesterdays reboot.... how many dist-upgrades that was, I am not sure.
[08:01]  * xnox_ac100 can check the logs I suppose
[08:01] <tjaalton> there should be no updates that result in this
[08:02] <tjaalton> mlankhorst: let's sort it out then
[08:04] <xnox_ac100> So I had uptime since 16:00 UTC 1st of October. reboot at 20:00 3rd of October and it all went to downhill from that.
[08:05] <mlankhorst> :/
[08:05] <mlankhorst> did you upgrade before that?
[08:05] <tjaalton> right, no x related updates there
[08:06] <mlankhorst> some linux kernel though
[08:06] <xnox_ac100> yes, full dist-upgrade just before the reboots. I did have rebooted for kernel update.
[08:06] <xnox_ac100> mlankhorst: oh...
[08:07] <tjaalton> well I bet 3.6 would still work
[08:07] <tjaalton> I've no issues with sandybridge here
[08:07] <tjaalton> even with 3.5
[08:07] <mlankhorst> mine chronically overheats :p
[08:08]  * xnox_ac100 is running 3.6-final currently => locks up. rc4 was good, let me reinstall that just in case I lock up again.
[08:08] <tjaalton> ok then, try an earlier 3.5 kernel
[08:08] <tjaalton> you should have plenty installed
[08:09] <tjaalton> so you probably got 3.5.0-16 installed, try -15
[08:10] <xnox_ac100> well actually i have only 4, as I only have 200MB /boot with full disk encryption
[08:10] <tjaalton> disk is cheap :)
[08:10] <xnox_ac100> also i did mark myself as affected by bug 1041790
[08:11] <mlankhorst> xnox_ac100: full disk encryption ftw
[08:11] <mlankhorst> even though it limits my ssd to 400mb/s instead of more XD
[08:11] <xnox_ac100> tjaalton: yeah... i have plenty in the lvm, just not much left for /boot, and now i want to split a chunk of it for UEFI boot partition....
[08:11] <xnox_ac100> i did recently create an unencrypted lvm for sbuild
[08:11] <mlankhorst> hehe, I cheated by rolling my own kernels :P
[08:11] <mlankhorst> I only need 1 initramfs
[08:12] <xnox_ac100> lol =) nice
[08:12] <mlankhorst> but all my test systems boot off nfs anyway
[08:12] <xnox_ac100> so i have -15, -16 & mainline rc4 and final. Will try -15 next I guess....
[08:14] <tjaalton> if there's a regression I'd think it's reverted/fixed in the upcoming -17
[08:14] <tjaalton> since it merged 3.5.5
[08:15] <xnox> well if you see xnox drop out, yet xnox_ac100 stays, then you know.... i locked up again.
[08:15] <mlankhorst> I ignore join/quit messages since its too much spam
[08:16]  * Hanmac is sorry that i spam you :'(
[08:16] <mlankhorst> I no longer see it :p
[08:28] <tjaalton> mlankhorst: so how's mesa doing?
[08:28] <mlankhorst> it builds and i pushed it back
[08:28] <tjaalton> ok, I'll test it briefly and upload
[08:29] <mlankhorst> we might want to see if we can drop some of the build patches before doing the final 9.0 upload though :)
[08:30] <tjaalton> oh right there was patch 14
[08:36] <mlankhorst> yeah it caused a conflict
[08:36] <tjaalton> huh?
[08:37] <Hanmac> mlankhorst the last time i SAW the output crashing and i had enough time to make an dmesg output ...
[08:39] <mlankhorst> tjaalton: well the upstream patch for 118 conflicted with the other patches we had in that tree
[08:39] <tjaalton> wth? laptop drops to failsafe after upgrade
[08:39] <mlankhorst> O_O
[08:39] <tjaalton> mlankhorst: oh I see
[08:41] <tjaalton> ah, using 3.6.. no wonder
[08:42] <Hanmac> http://pastebin.com/JL03Ef0E << mlankhorst that is my dmesg before its totaly crash ... shortly before the scene got's green, and the tty1 gots purple
[08:45] <mlankhorst> Hanmac: dupe it for now I suppose, not much that can be done until upstream responds
[08:52] <Hanmac> ps: the linux-image-3.5.0-17 packages are coming (maybe) today ... (the meta packages are pointing to it) but i dont think it solv my problem :'(
[08:52] <tjaalton> we know, and haven't suggested it would
[08:54] <mlankhorst> Hanmac: I don't mean to be rude but we're aware of the problem and affected too, if you keep focusing our attention to it we get distracted from finding a solution and our other work on solving problems for others
[09:00] <mlankhorst> tjaalton: any idea when the 3.5.5 linux image comes?
[09:00] <tjaalton> mlankhorst: uploaded already
[09:00] <tjaalton> https://lists.ubuntu.com/archives/quantal-changes/2012-October/010845.html
[09:00] <tjaalton> armhf failed to build it seems
[09:00] <tjaalton> others are in NEW
[09:01] <mlankhorst> ah :)
[09:01] <Hanmac> hm it seems the failing "GPU softreset" & "[drm:r600_resume] *ERROR* r600 startup failed on resume" only destroys the X-Output (so it does not change anymore) but tty does still work
[09:01] <Hanmac> "irq/18-radeon Tainted" are the bad guy that destroy tty1 too 
[09:01] <Hanmac> is this information helpful?
[09:02] <mlankhorst> Hanmac: no we already collected all that information from your log, nothing to add
[09:02] <mlankhorst> and please use the bug report instead of here
[09:06] <tjaalton> building mesa
[09:07] <tjaalton> mlankhorst: added a changelog entry for the patch ;)
[09:07]  * Hanmac is joyful because of new packages
[09:16] <mlankhorst> tjaalton: yeah I'm just marking a whole bunch of xorg-server bugs incomplete now when they were filed at x < 1.13 or one of the rc's
[09:27] <tjaalton> mlankhorst: cool
[09:27]  * xnox_ac100 locked up. booting into 3.5 -15
[09:28] <tjaalton> mlankhorst: the build patches are now in 9.0
[09:28] <mlankhorst> \o/
[09:28] <mlankhorst> thought so
[09:57]  * xnox_ac100 ponders if it's actually compiz locking me up
[09:57] <tjaalton> can you ssh in?
[09:57] <xnox_ac100> no. I should set up this netbook for ssh in.
[09:57] <xnox_ac100> tjaalton: what do you want me to check on the next lock up?
[09:58] <tjaalton> xnox_ac100: well, if you can log in with ssh it means it's not a full system lockup
[09:58] <tjaalton> also, does mouse work?
[09:58] <tjaalton> might just be some app taking the input grab
[09:59] <tjaalton> you can try chvt
[09:59] <tjaalton> via ssh
[10:00] <Hanmac> xnox_ac100 i maybe need more information, what is your xserver version, whats your graphic card, what do you mean with "locking up"?
[10:01] <tjaalton> Hanmac: please, don't
[10:01] <Hanmac> okay :(
[10:02] <xnox> Hanmac: check irclogs from today, if you are interested =)
[10:03]  * xnox setup working ssh now.
[10:23] <xnox_ac1001> hmm... actually it's a full system lock down. it doesn't even reply to pings.
[10:24] <Hanmac> so it puts finger in its ears and ignores you? that is very mean
[10:26] <tjaalton> xnox_ac1001: ok so it happens with -15 too?
[10:28] <xnox_ac1001> tjaalton: -15, 3.6rc4, 3.6final now going to -16 again (can't remember now) with or without unity/compiz
[10:28] <tjaalton> ok, dodgy hw then
[10:28] <tjaalton> check your ram
[10:29] <xnox_ac1001> hmmm....
[10:29] <xnox_ac1001> memcheck or is there anything else?
[10:29] <tjaalton> selectable from the grub menu
[10:29] <tjaalton> memcheck yes
[10:29] <xnox_ac1001> ok
[10:29]  * xnox_ac1001 doesn't deal with hw that often.
[10:29] <xnox_ac1001> it works without a hitch usually =)
[10:29] <tjaalton> that's all I can think of
[10:30] <xnox_ac1001> sounds plausible. since it's out of the blue in dependant of everything.
[10:30] <tjaalton> you can check dpkg.log for the upgrade diff, but don't think you'll find anything suspicious there, if the kernel update has already been ruled out
[10:31] <xnox_ac1001> and compiz....
[10:31] <xnox_ac1001> ruled out.
[10:31] <xnox_ac1001> and unity. no updates to X as you said....
[11:06] <tjaalton> mesa uploaded
[11:06] <mlankhorst> yay
[11:08]  * Hanmac says yay too
[11:20]  * xnox_ac1001 doing memtest now
[11:21] <Hanmac> tjaalton where do you upload the mesa? is there an link where i can look?
[11:23] <mlankhorst> quantal-proposed presumably
[11:23] <mlankhorst> or simply quantal
[11:25] <tjaalton> quantal
[11:25] <Hanmac> hm is this the right package for looking? https://launchpad.net/ubuntu/+source/mesa
[11:25] <tjaalton> it's not accepted yet
[11:25] <xnox_ac1001> you can fetch it from unapproved queue
[11:26] <tjaalton> right
[11:30] <mlankhorst> but it won't do you any good since your problem is in the kernel..
[11:31] <Hanmac> but i could still dream that its going better ...
[11:35] <xnox_ac1001> hmmm... http://forums.linuxmint.com/viewtopic.php?f=59&t=113043 recommends i915.i915_enable_rc6=0
[11:36] <tjaalton> it shouldn't break like that
[11:37] <xnox_ac1001> ok
[11:41] <tjaalton> I mean, break all of a sudden
[11:41] <tjaalton> if it worked fine before
[11:43] <xnox_ac1001> bug 993187 sounds like a winner
[11:43] <xnox_ac1001> including the comment #493 about buggy ethernet driver locking the system up
[11:44] <xnox_ac1001> but i am pondering about trying comment #321 boot param settings
[11:45] <tjaalton> that bug is the worst bug in the history of mankind :)
[11:45] <tjaalton> bug report
[11:45] <xnox_ac1001> tjaalton: nah the worst bug was brother printers don't print on tuesdays.
[11:45] <mlankhorst> :>
[11:46] <xnox_ac1001> with reporter confirming the bug every tuesday & bug triangers unvalidating it every wednesday
[11:46] <tjaalton> people are still posting new comments there even though it was closed months ago
[11:46] <tjaalton> as impossible to follow
[11:46] <mlankhorst> haha
[11:46]  * xnox_ac1001 the reporter had weekly reports to print on tuesdays....
[11:46] <xnox_ac1001> it took a while for the pattern to be noticed.
[11:47] <tjaalton> heh
[11:47] <tjaalton> guess it would be helpful to mark this bug as private..
[11:47] <tjaalton> should help with the hand-waving
[11:49] <xnox_ac1001> tjaalton: please don't =)
[11:51] <xnox_ac1001> google indexed each comment and i would never find out about those two comments and learn that "sandybridge is disappointment of the year 2011@
[11:53] <tjaalton> ok then, I won't touch it :)
[11:58] <mlankhorst> tjaalton: you forgot to push the xserver-xorg-video-nouveau commit after pushing?
[11:58] <tjaalton> mlankhorst: no, should be there
[11:59] <mlankhorst> oh right, i forgot to fetch
[12:02] <mlankhorst> tjaalton: I added a small exa fix for nouveau, can you upload it?
[12:02] <tjaalton> sure
[12:15] <tjaalton> mesa accepted
[12:16] <tjaalton> mlankhorst: btw, if you have nothing else to add to a package, you can finalize the changelog entry with dch -r
[12:16] <mlankhorst> tjaalton: I know but I can't upload :-)
[12:17] <tjaalton> you can commit the change
[12:17] <tjaalton> to git
[12:17] <tjaalton> it doesn't matter who is listed on the changelog entry
[12:17] <tjaalton> as the signer
[12:18] <tjaalton> what matters is who has signed the package files
[12:19] <tjaalton> hmm, didn't the patch fix any of the lp bugs?
[12:20] <tjaalton> uploaded anyway
[12:21] <Hanmac> yeah the build begins
[12:25] <mlankhorst> most nouveau bugs might be in mesa :/
[12:25] <mlankhorst> or kernel
[12:26]  * Hanmac wonder behind with door my zonk is hiding ...
[12:32] <tjaalton> amd64 build now failed twice
[12:33] <mlankhorst> on what?
[12:34] <tjaalton> mesa
[12:34] <tjaalton> some bug in parallel build
[12:34] <tjaalton> builds fine here
[12:34] <Hanmac> the error was  "SIRegisterInfo.td:2:1: error: Unknown token when parsing a value " ... hm it uses llvm maybe this was the problem?
[12:35] <tjaalton> no, there's a known bug in the parallel build, still not fixed upstream
[12:35] <tjaalton> it doesn't fail every time
[12:35]  * mlankhorst tries -j32
[12:40] <mlankhorst> seems xatracker needs -ldrm and -pthread still and glapi is missing symbols too :/
[12:40] <tjaalton> details.. :)
[12:43] <mlankhorst> tjaalton: seems debuild -b -j256 works here?
[12:44] <tjaalton> how many times did you run it?
[12:44] <tjaalton> also, since you don't have 256 cores I doubt it'll hit the race :)
[12:45] <mlankhorst> fine I'll let it loop a bunch of times :P
[12:45] <tjaalton> I'm running with -j8
[12:45] <mlankhorst> 4 cores?
[12:45] <tjaalton> 4+4
[12:45] <tjaalton> HT
[12:45] <mlankhorst> ah
[12:45] <mlankhorst> yeah i have the i5 no ht :(
[12:45] <mlankhorst> next system!!
[12:47] <tjaalton> it seems to be failing on gallium/radeon though, generating *.td isn't finished before llvm-tblgen is run
[12:47] <tjaalton> I mean all the failures have been there now
[12:50] <mlankhorst> tjaalton: yeah repeatedly doing it on that directory fails
[12:51] <mlankhorst> those incs seem like they should be added though by make depend
[12:55] <tjaalton> the lp build passed this time
[13:12] <mlankhorst> xorg-server seems to love to hang on the xvfb randomly, didn't manage to reproduce it :/
[13:54] <mlankhorst> but that's at least radeon nailed down, talk about thinkos :)
[14:53] <Hanmac> hm nearly all packages for 3.5.0-17.26 are finish ... only the armhf takes so long :'( ... hm launchpad should have some "notic me when the build is finish" option ...
[14:57] <tjaalton> you shouldn't care about that
[14:57] <tjaalton> it's not going to help
[14:58] <mlankhorst> you mean xorg-server?
[15:00] <Hanmac> i meand https://launchpad.net/ubuntu/+source/linux/3.5.0-17.26 ... i wait for the packages so i could upgrade my linux and linux-lowlatency packages ... 
[15:00] <Hanmac> i know it does not solv my problem :( but i am happy about new packages ... its like x-mas :D
[17:36] <Sarvatt> RAOF: seen https://bugs.launchpad.net/unity/+bug/1057000 ?
[17:37] <Sarvatt> that got uploaded to quantal today, whoops :(
[17:40] <tjaalton> sigh
[17:51] <mlankhorst> argh, my keyboard is detected as a mouse..
[20:01] <Hanmac> yeah the 3.5.0-17.26 packages are finish ... they took > 8 hours ... i could read a book in this time ...