| bjsnider | Sarvatt, ping | 00:27 |
|---|---|---|
| yofel | hm, what's upstream for nvidia-settings? (I'm wondering where to send the patch on bug 604525 to) | 00:31 |
| ubot4 | Launchpad bug 604525 in nvidia-settings (Ubuntu) (and 1 other project) "nvidia-settings "Configure Display Device" menu should be drop-down list instead of dialog box (affects: 1) (heat: 8)" [Wishlist,New] https://launchpad.net/bugs/604525 | 00:31 |
| === yofel_ is now known as yofel | ||
| hyperair | Sarvatt: [ 13004.621] (II) intel(0): Kernel page flipping support detected, enabling | 06:24 |
| hyperair | thankfully the hangs seem minor enough that i can get back in by using ssh | 06:24 |
| RAOF | Yeah, Sarvatt said he was going to drop that patch to get closer to git. | 06:28 |
| RAOF | Damn, what has happened to all the amd64 PPA buildds? | 06:30 |
| hyperair | RAOF: take a look at https://launchpad.net/builders | 06:30 |
| RAOF | Hah. The kernel thwarts me again! | 06:31 |
| hyperair | lol | 06:31 |
| RAOF | Have you tried the intel-legacy DDX? | 06:31 |
| hyperair | what's that? | 06:31 |
| ricotz | RAOF, hello | 06:32 |
| RAOF | hyperair: https://edge.launchpad.net/~raof/+archive/aubergine/+packages | 06:32 |
| RAOF | ricotz: Hello. | 06:32 |
| ricotz | are there some currently known issues with nouveau in maverick? | 06:32 |
| RAOF | hyperair: It's essentially a gem-less DDX grafted onto 2.12 for those not using KMS. | 06:33 |
| RAOF | ricotz: Not known by me particularly. Are you going to tell me of one? :) | 06:33 |
| hyperair | RAOF: ah i see. but why would i not want to use KMS? | 06:33 |
| RAOF | hyperair: Because you've got an i8xx card, and GEM has been a one-way ticket to random GPU hangs. | 06:33 |
| RAOF | That's not why _you'd_ want it, that's why someone would want it. | 06:34 |
| hyperair | RAOF: no, i've got an i965 card. | 06:34 |
| hyperair | ah i see. | 06:34 |
| ricotz | i cant start a normal x session, xorg.0.log say "no screens found", is there some race condition between kms and gdm or something? | 06:34 |
| RAOF | And _I'd_ like you to test it, so it can go upstream and we can have rainbow-dancing unicorns. | 06:34 |
| RAOF | ricotz: Care to pastebin dmesg & Xorg.0.log? | 06:34 |
| ricotz | it seems to work after a xserver reinstall so ureadahead it triggered, i think | 06:35 |
| hyperair | RAOF: test how? just install it and leave things be? | 06:35 |
| RAOF | hyperair: Yup. | 06:35 |
| hyperair | or do i have to disable KMS? | 06:35 |
| RAOF | Well, you can try disabling KMS. | 06:35 |
| RAOF | And checking that it doesn't explode horribly. | 06:35 |
| hyperair | heh okay. | 06:36 |
| hyperair | i'll give it a go. | 06:36 |
| RAOF | But even just installing it and finding the KMS works fine is useful. | 06:36 |
| RAOF | s/the/that/g | 06:36 |
| hyperair | i'll give it a go once your amd64 builds are done. | 06:36 |
| ricotz | RAOF, http://paste.pocoo.org/show/238186/ - http://paste.pocoo.org/show/238187/ | 06:37 |
| ricotz | RAOF, it happens with maverick packages and edgers | 06:38 |
| RAOF | So, the very first thing I see is that you're using xorg-edgers ;) | 06:38 |
| ricotz | yes, but i am getting the same error without edgers | 06:39 |
| RAOF | Hm. | 06:41 |
| RAOF | As they say in the movies: let me check my notes. | 06:42 |
| ricotz | RAOF, ok | 06:43 |
| ricotz | RAOF, xorg.failsafe.log of x which is working right now - http://paste.pocoo.org/show/238191/ | 06:48 |
| dupondje | nouveau is broken atm in maverick ? | 06:50 |
| RAOF | Apparently the answer is “yes” | 06:55 |
| RAOF | ricotz: Just for my own amusement, can you get a non-edgers Xorg.0.log? | 06:55 |
| ricotz | RAOF, can you confirm this problem? ok, this will take some time | 06:56 |
| dupondje | https://launchpad.net/ubuntu/+source/libdrm/2.4.21-1ubuntu1/+build/1874547 | 06:58 |
| dupondje | could this be the reason ? | 06:58 |
| RAOF | No, that's not going to be the reason. | 06:58 |
| RAOF | I'll update my nouveau laptop & see if this happens locally. | 07:01 |
| ricotz | RAOF, mhh, now i got it with my intel laptop | 07:03 |
| dupondje | its broken since yesterday evening somewhere .. | 07:03 |
| dupondje | dunno what got upgraded since then | 07:03 |
| RAOF | dupondje: Learn to love the Ubuntu Software Centre. | 07:04 |
| RAOF | Fire it up, and hit “history” | 07:04 |
| RAOF | Tada! An answer to the question: what upgraded since last night :) | 07:05 |
| dupondje | i'm not on my broken laptop atm :) can't check | 07:05 |
| ricotz_ | RAOF, for me it started before yesterday | 07:06 |
| dupondje | I updated @ around 16:00 and rebooted, everything was fine. | 07:07 |
| RAOF | ricotz_: You've got the same problem with an intel laptop? | 07:08 |
| RAOF | I'm *definitely* not seeing it, then. | 07:08 |
| ricotz_ | i am just getting the nouveau log | 07:08 |
| ricotz_ | RAOF, ok, first boot after "downgrade" http://paste.plurk.com/show/280757/ | 07:11 |
| ricotz_ | second boot http://paste.plurk.com/show/280758/ | 07:12 |
| dupondje | second looks quite the same error like I got ... | 07:13 |
| ricotz_ | RAOF, with intel http://paste.plurk.com/show/280759/ | 07:13 |
| RAOF | Ok. So it looks a lot like libdrm is broken for you. | 07:14 |
| RAOF | Or, possibly, the kernel. | 07:14 |
| dupondje | I rebooted after kernel upgrade, and then it still worked ... | 07:15 |
| RAOF | Does this only happen with 2.6.35-8-generic? | 07:15 |
| dupondje | no | 07:15 |
| RAOF | Hah, too slow. | 07:15 |
| ricotz_ | it also happened with -7 | 07:15 |
| ricotz_ | on intel it seemed to start with -8 | 07:16 |
| RAOF | Ok, I've got a newer libdrm 'cause I'm testing the legacy DDX. I'll downgrade that and see. | 07:17 |
| RAOF | Nope, still works fine here. | 07:25 |
| RAOF | What versions of libdrm, libdrm-intel, libdrm-nouveau, kernel, DDX, and xserver are you using? | 07:26 |
| ricotz | the current maverick versions | 07:27 |
| dupondje | same here ... | 07:27 |
| RAOF | So, 2.4.20-1ubuntu2, 2.6.35-8-generic, intel 2.11.0, and 1.8.2RC2? | 07:28 |
| dupondje | can't check atm :( | 07:28 |
| RAOF | Got an xorg.conf? | 07:31 |
| ricotz | libdrm 2.4.20-2ubuntu1 - kernel 35-8 - intel 2:2.11.0-1ubuntu2 - xserver-xorg 2:1.8.1.902-0ubuntu2 | 07:31 |
| ricotz | no | 07:31 |
| RAOF | I need to run some errands - I'll ponder this while I do so, and while the nouveau lappy is updating. | 07:34 |
| ricotz | ok, np | 07:35 |
| dupondje | ricotz: amd64 or i386 ? | 07:35 |
| ricotz | both amd64 | 07:35 |
| dupondje | you also RAOF ? | 07:38 |
| RAOF | dupondje: Yup. amd64 | 07:53 |
| ricotz | RAOF, any progress? using nomodeset gets it to work | 08:37 |
| RAOF | ricotz: For values of “work” equal to “uses vesa instead”, I presume :) | 08:37 |
| RAOF | It doesn't help that I can't reproduce this locally :/ | 08:38 |
| ricotz | oh you are right this disables nouveau | 08:38 |
| ricotz | so it might be a hardware specific problem | 08:39 |
| RAOF | It shouldn't be. | 08:40 |
| ricotz | i tested kernel 35.7-11 which works on the first boot, and freezes on the second | 08:41 |
| RAOF | It's not entirely obvious why drmGetBusid is returning "" rather than “pci:0000:01:00.0”, though. | 08:41 |
| RAOF | Works on the first boot, fails on the second? | 08:41 |
| RAOF | I don't suppose that if you log off and log in again it works? | 08:42 |
| ricotz | no i need to reboot, like i said, it seems to be related to some trigger which is called after a new kernel/x installation | 08:43 |
| ricotz | like ureadahead which delays some things | 08:43 |
| dupondje | after kernel upgrade it was still working here ... tho only 1 reboot, so might have been lucky ;) | 08:44 |
| RAOF | Can you confirm that? Delete /var/lib/ureadahead/pack and it'll re-profile for you. | 08:44 |
| ricotz | RAOF, will try that | 08:46 |
| RAOF | You know what? I need to get a big, beefy desktop machine to build stuff on. | 08:56 |
| ricotz_ | RAOF, i have restarted now couple times, deleting this ureadahead pack file seems to resolve it | 08:57 |
| RAOF | Can you give me two dmesg logs — one that has worked and one that hasn't? | 08:59 |
| ricotz_ | is the "var.run.pack" introduced recently? | 08:59 |
| RAOF | You get one pack per filesystem. | 09:00 |
| RAOF | The root filesystem gets “pack”; the others get “boot.pack”, “usr.pack”, etc. | 09:00 |
| ricotz_ | ok, then it is fine | 09:01 |
| ricotz | RAOF, FAIL: http://paste.pocoo.org/show/238215/ - GOOD: http://paste.pocoo.org/show/238216/ | 09:05 |
| ricotz | i hope these are the right ones (after rebooting this much) | 09:06 |
| RAOF | Wait - the *second* one is good? | 09:07 |
| ricotz | this is the current boot | 09:08 |
| RAOF | Oh, of course. ureadahead. *That's* why udev would be starting some 12 seconds earlier in the second trace. | 09:09 |
| RAOF | So, I'd guess that if you removed “splash” from your kernel command line, this would work. | 09:10 |
| ricotz | it seems to work already | 09:12 |
| RAOF | Even without deleting the pack? | 09:13 |
| RAOF | You had a problem before - are you suggesting that this has magically fixed itself? :) | 09:13 |
| ricotz | yes, i have deleted it twice, but not the third time | 09:14 |
| ricotz | i will reboot again to be sure | 09:14 |
| ricotz_ | ok, there is it again :( | 09:16 |
| dupondje | https://launchpad.net/ubuntu/+source/libdrm/2.4.21-1ubuntu1/+build/1874547 is build ... maby install ? :)- | 09:16 |
| RAOF | I'd be surprised if that fixed it. | 09:17 |
| steveire | Is there any other useful information I can put in this bug report? https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/606001 | 09:18 |
| ubot4 | Launchpad bug 606001 in xorg (Ubuntu) "Dual screen setup does not work after installation (affects: 1) (heat: 6)" [Undecided,New] | 09:18 |
| ricotz_ | RAOF, removing "splash" doesnt work | 09:20 |
| RAOF | Well that's put a crimp in an otherwise damn-fine theory :) | 09:20 |
| RAOF | Dump a file containing “FRAMEBUFFER=y” into /usr/share/initramfs-tools/conf.d ? That'll get the drm module added to the initramfs, which should make loading it (a) faster and (b) independent of ureadahead. | 09:21 |
| ricotz_ | ok | 09:22 |
| RAOF | steveire: What have you done to try and get a dual-head display? | 09:24 |
| steveire | RAOF: I have opened system settings and gone to the 'monitors' section, which tells me I do not have 2 monitors. | 09:25 |
| RAOF | steveire: Also, running “apport-collect 606001” will attach all sorts of tasty information. | 09:25 |
| steveire | "This module is only for configuring systems with a single desktop spread across multiple monitors. You do not appear to have this configuration." | 09:26 |
| RAOF | Ok. That would be useful information to have on the bug :) | 09:28 |
| ricotz_ | RAOF, patching initramfs works | 09:28 |
| RAOF | Consistently? | 09:29 |
| ricotz_ | i will reboot a third time, but fsck might do some things here | 09:29 |
| ricotz_ | i need to reset the fscheck count | 09:30 |
| steveire | RAOF: Is that apport thing interactive? Can I see the info it sends back before it does? | 09:31 |
| steveire | http://dpaste.com/218731/ | 09:32 |
| RAOF | steveire: I think so, from memory. | 09:32 |
| RAOF | You want to give it access - it'll upload xorg.0.log, dmesg, etc. | 09:33 |
| steveire | It implies it will upload more than that. unspecified 'private data' | 09:33 |
| steveire | And it doesn't tell me if I can review it or not. | 09:34 |
| RAOF | No, that means that it will be able to write to bugs you've marked as “private” | 09:34 |
| ricotz_ | RAOF, ok, two working reboots | 09:36 |
| steveire | RAOF: https://bugs.launchpad.net/ubuntu/+source/apport/+bug/606175 I can't run apport-collect yet | 09:40 |
| ubot4 | Launchpad bug 606175 in apport (Ubuntu) "apport-collect does not tell me what it will do (affects: 1) (heat: 6)" [Undecided,New] | 09:40 |
| RAOF | steveire: I've just checked - it will show you what it'll send. | 09:42 |
| steveire | RAOF: This post http://blog.grossmeier.net/2009/03/02/apport-collect-just-what-you-wanted/ says it works without any other interaction with the reporter | 09:45 |
| steveire | The man page is also not very clear on it http://man.he.net/man1/apport-collect | 09:47 |
| RAOF | Well, I've just checked with a report on staging, and it pops up the “here's the stuff we'll be reporting” page. | 09:52 |
| RAOF | Possibly if you run it without X it doesn't do that, but it's generally pretty careful to ensure you can easily find out what data is being reported. | 09:52 |
| ricotz | RAOF, so it looks like a race condition, something has gotten slower or faster so this problem appears | 09:56 |
| RAOF | Right. | 09:56 |
| ricotz | now nouveau is loaded 1 second ealier, but initrd is 16MB now | 09:57 |
| RAOF | Because it's got all the drm modules in it, yeah. | 09:58 |
| RAOF | Plus plymouth | 09:58 |
| steveire | I don't see why it needs 'write access' to read dmesg... | 09:58 |
| RAOF | It needs write access to your *launchpad* account. | 09:59 |
| RAOF | Otherwise it can't add anything to the bug, obviously! | 09:59 |
| RAOF | Is that not obvious from the messages - both in the terminal, and the launchpad message? | 10:00 |
| RAOF | steveire: Well, I guess it's clearly not obvious, or you wouldn't be asking the question. If you can think of a better wording, feel free to add it to the apport bug you just submitted. | 10:03 |
| steveire | RAOF: Ok, well I ran the tool | 10:09 |
| RAOF | steveire: Ok, now I'm confused. Only one of your shiny, fancy dell monitors is showing an image? | 10:23 |
| RAOF | steveire: As far as X is concerned, they should be both lit up and running. | 10:24 |
| jcristau | means the kernel is confused, afaict | 10:24 |
| RAOF | I guess so. | 10:25 |
| * RAOF heads to dinner. | 10:25 | |
| steveire | That's correct. Only one of them is useful. The other is completely blank | 10:26 |
| RAOF | steveire: I don't suppose running “xrandr --output DVI-I-2 --right-of DVI-1-1” does anything? | 11:08 |
| ara | hello all! | 11:20 |
| ara | in the middle of a big X upload? | 11:20 |
| ara | I think I upgraded and rebooted at the wrong time :) | 11:20 |
| ara | my X won't boot anymore | 11:20 |
| steveire | stephen@chimera:~$ xrandr --output DVI-I-2 --right-of DVI-1-1 | 11:21 |
| steveire | xrandr: cannot find output "DVI-1-1" | 11:21 |
| jcristau | steveire: DVI-I-1, not DVI-1-1 | 11:21 |
| steveire_ | That command borked my workstation once I'd fixed the 1 to an I | 11:23 |
| steveire_ | All my windows disappeared, and I had fragments of plasmoids in different places and nothing worked. I had to hard-reset | 11:25 |
| steveire_ | And now it's having a hard time booting. | 11:25 |
| RAOF | Well, at least it had some impact, even if it wasn't particularly good. :/ | 11:27 |
| RAOF | ara: No - the next flurry of X will be xserver 1.9, which I won't upload until it stops crashing with clutter apps. | 11:28 |
| RAOF | Gah, too slow. | 11:28 |
| jcristau | maybe mark it as blocking 1.9 in bugzilla? | 11:29 |
| jcristau | where "it" is the bug you filed | 11:29 |
| RAOF | I wasn't sure if I should do that - I guess I could. | 11:29 |
| steveire | I'm wondering if I should try again. You know that xkcd comic with the lightning machine? | 11:30 |
| RAOF | Heh. | 11:31 |
| RAOF | “Hm. I wonder if it does that *every* time” | 11:31 |
| steveire | What is the command supposed to do? | 11:31 |
| tseliot | steveire_: I found kde 4 to act in a weird way when using multiple screens. None of this can be reproduce in gnome, at least here | 11:31 |
| RAOF | That's meant to make one of your displays left of the other - IE: de-clone them | 11:31 |
| RAOF | I certainly found that KDE4 didn't deal particularly well with multiple heads. It didn't go totally bananas, though. | 11:32 |
| RAOF | jcristau: Incidentally, I'm thinking of attending XDS, and I was wondering what you thought. | 11:32 |
| steveire | Maybe I'll install ubuntu-desktop and see what happens. Even the boot splash only appears on one screen though. | 11:36 |
| RAOF | ara: No - the next flurry of X will be xserver 1.9, which I won't upload until it stops crashing with clutter apps. | 11:36 |
| RAOF | steveire: Which is strange, given that it worked on the livecd. | 11:36 |
| RAOF | jcristau: Bug marked as blocking. | 11:37 |
| jcristau | i think this xds is the first time i might be able to attend | 11:37 |
| steveire | Indeed. | 11:39 |
| dupondje | is the x-issue fixed already ? or bugreport / workaround ? That I can boot my computer again this evening :) | 11:58 |
| jcristau | whatever "the x-issue" means... | 12:00 |
| RAOF | dupondje: If you add the drm module to your initramfs you'll avoid the race condition. | 12:07 |
| RAOF | ricotz: Did you file a bug, incidentally? | 12:08 |
| ricotz | RAOF, no, i havent | 12:08 |
| RAOF | Please do so - I can't fix that race off the top of my head, so a bug will be needed. | 12:11 |
| ricotz | RAOF, will do, as soon i can | 12:16 |
| ricotz | RAOF, what about this bug report https://bugs.edge.launchpad.net/ubuntu/+source/xorg-server/+bug/459639 | 12:33 |
| ubot4 | Launchpad bug 459639 in xorg-server (Ubuntu) "[Karmic] X server starts randomly in failsafe when starting from cold boot (affects: 6) (heat: 42)" [High,Confirmed] | 12:33 |
| jcristau | karmic? | 12:34 |
| dupondje | and not only from cold boot imo :) | 12:38 |
| ricotz | yeah, i will file a new one | 12:38 |
| RAOF | And a huge muddle of multiple “X doesn't start” bugs. | 12:40 |
| RAOF | Handy hint: the binary blob not starting is *almost always* a different bug to any free driver not starting :) | 12:40 |
| seb128 | http://launchpadlibrarian.net/52029105/buildlog_ubuntu-maverick-i386.clutter-gtk-0.10_0.10.4-0ubuntu3_FAILEDTOBUILD.txt.gz | 12:46 |
| seb128 | is that libgl installability issue known? | 12:47 |
| ricotz | RAOF, https://bugs.edge.launchpad.net/ubuntu/+source/xorg-server/+bug/606244 | 12:52 |
| ubot4 | Launchpad bug 606244 in xorg-server (Ubuntu) "X doesn't find a screen and is not starting due a race condition (affects: 1) (heat: 6)" [Undecided,New] | 12:52 |
| RAOF | seb128: clutter-gtk builds fine locally. I guess it could be an i386 issue? | 13:15 |
| seb128 | I figure it out now | 13:15 |
| seb128 | I newed libkms1 to universe yesterday | 13:16 |
| seb128 | I've promoted it now, should be fixed after the next publisher run | 13:16 |
| seb128 | would be nice to document new libraries in the changelog though | 13:16 |
| seb128 | so we know what to do when newing them | 13:16 |
| seb128 | Sarvatt, ^ | 13:16 |
| RAOF | Aaah, the newer libdrm. Right. | 13:16 |
| yofel | short question, what's upstream for nvidia-settings? (where to send a patch..) | 13:16 |
| seb128 | or to use -v to list debian changes | 13:17 |
| RAOF | That obviously hasn't hit my amd64 mirror yet. | 13:17 |
| RAOF | Well, back to packing. | 13:19 |
| tseliot | yofel: you can contact Aaron plattner | 13:20 |
| yofel | where? | 13:20 |
| tseliot | you can have a look at whatever commit in git an see his address: http://cgit.freedesktop.org/~aplattner/nvidia-settings/commit/?id=655fb95354fdfb7ffe5f44f0341c132165b0c983 | 13:22 |
| tseliot | what kind of change is that? | 13:22 |
| yofel | bug 604525 | 13:23 |
| ubot4 | Launchpad bug 604525 in nvidia-settings (Ubuntu) (and 1 other project) "nvidia-settings "Configure Display Device" menu should be drop-down list instead of dialog box (affects: 1) (heat: 10)" [Wishlist,Confirmed] https://launchpad.net/bugs/604525 | 13:23 |
| tseliot | aah, that one | 13:23 |
| tseliot | yes, feel free to send Aaron a patch | 13:23 |
| yofel | was asking as launchapd doesn't tell me where to find upstream.. | 13:23 |
| yofel | thanks | 13:23 |
| tseliot | seb128: I didn't notice that there were new libraries when I tested/sponsor the upload. So it's my bad too | 13:24 |
| seb128 | tseliot, that's ok, I figure what the issue was now ;-) | 13:25 |
| tseliot | good | 13:25 |
| dupondje | RAOF: so its a kernel bug? But why is it broken for -7 and -8 since today ? and it worked without issues for days on -7 ? :) | 13:35 |
| dupondje | somebody else hitting the https://bugs.edge.launchpad.net/ubuntu/+source/xorg-server/+bug/606244 bug ? | 13:43 |
| ubot4 | Launchpad bug 606244 in linux (Ubuntu) "X doesn't find a screen and is not starting due a race condition (affects: 2) (heat: 12)" [Undecided,New] | 13:43 |
| tseliot | maybe you should try the latest libdrm | 13:53 |
| dupondje | yea indeed :) | 13:53 |
| dupondje | i'm not @ home, so I can't test | 13:53 |
| dupondje | but I commented the bug | 13:53 |
| tseliot | ok | 13:57 |
| dupondje | https://bugs.launchpad.net/ubuntu/+source/libdrm/+bug/606244 | 14:36 |
| ubot4 | Launchpad bug 606244 in libdrm (Ubuntu) "X doesn't find a screen and is not starting due a race condition (affects: 2) (heat: 12)" [Undecided,Fix released] | 14:36 |
| dupondje | like I tough | 14:36 |
| dupondje | libdrm and its fixed :) | 14:36 |
| tseliot | very good | 14:41 |
| tseliot | thanks for testing | 14:41 |
| ricotz | dupondje, the problem is i was hitting this bug also with edgers ppa | 14:49 |
| ricotz | where libdrm was up2date for some time | 14:50 |
| dupondje | ricotz: so you still have it with newest libdrm ?! | 16:36 |
| dupondje | feel free to re-open then btw :) | 16:36 |
| Sarvatt | the stuff in the section of the changelog you quoted was already in the previous libdrm dupondje | 16:40 |
| dupondje | mmmm weird it started working then after libdrm update here .. | 16:41 |
| dupondje | Sarvatt: previous version was 2.4.20-2ubuntu1, now its 2.4.21-1ubuntu1 | 16:43 |
| dupondje | the changelog section is in 2.4.20-3 .. | 16:44 |
| Sarvatt | yeah I know, but I did the abi bump stuff from 2.4.20-3 in 2.4.20-2ubuntu1 before they did it in debian, thats why we didn't bother merging 2.4.20-3 | 16:44 |
| dupondje | ah but 2.4.20-2ubuntu1 gives + 03_revert_abi_change.diff - Obsolete | 16:45 |
| dupondje | anyway it was broken here, and libdrm update resolved it ... so something must be changed ;) | 16:45 |
| Sarvatt | your initrd got rebuilt with the FRAMEBUFFER=Y maybe? :) | 16:46 |
| dupondje | well no :s | 16:46 |
| dupondje | it didn't even regenerated my initrd :) | 16:46 |
| dupondje | was getting completely the same error as ricotz | 16:49 |
| Sarvatt | i dont see anything in libdrm that would have affected that | 16:50 |
| dupondje | [ 51.454] drmOpenByBusid: drmGetBusid reports => when it was broken | 16:53 |
| dupondje | [ 25.539] drmOpenByBusid: drmGetBusid reports pci:0000:01:00.0 | 16:54 |
| dupondje | now | 16:54 |
| dupondje | :) | 16:54 |
| dupondje | ricotz: did you try libdrm_2.4.21-1 ? | 16:55 |
| Sarvatt | that happens when the kernel drm isn't ready when the ddx is trying to start | 16:56 |
| dupondje | look, I gonne try to reboot some times, to see if it still works :) | 16:59 |
| Sarvatt | do you guys have agp cards by any chance? | 17:01 |
| dupondje | Seems like i've been very lucky, that the first boot after libdrm upgrade worked :p | 17:06 |
| dupondje | https://bugs.launchpad.net/ubuntu/+source/libdrm/+bug/606244 | 17:07 |
| dupondje | so not fixed ... :) | 17:07 |
| ubot4 | Launchpad bug 606244 in libdrm (Ubuntu) "X doesn't find a screen and is not starting due a race condition (affects: 2) (heat: 12)" [Undecided,New] | 17:07 |
| Sarvatt | do you guys have agp cards by any chance? | 17:08 |
| dupondje | in my old computer yes :) | 17:08 |
| ricotz | dupondje, i havent tried it with the maverick version, i am using edgers which gave me the same error | 17:09 |
| ricotz | Sarvatt, it is a PCIs card | 17:09 |
| ricotz | PCIe | 17:09 |
| dupondje | its laptop here, so euh pcie bus also I bet | 17:10 |
| ricotz | dupondje, so you are still getting the error? | 17:11 |
| ricotz | dupondje, nm ;) | 17:11 |
| dupondje | yep :s | 17:11 |
| dupondje | been very lucky that just after the libdrm reboot it worked :p | 17:12 |
| dupondje | the reboot after I upgraded libdrm .. :) | 17:12 |
| ricotz | yes, like i said before the first boot after some kernel/x changes it works | 17:12 |
| dupondje | I also see kernel bug in boot: [ 33.758638] BUG: unable to handle kernel NULL pointer dereference at 00000000000003c0 | 17:13 |
| dupondje | :p | 17:13 |
| dupondje | [ 33.758809] Process plymouthd (pid: 375, threadinfo ffff880037e28000, task ffff880037d38000) | 17:13 |
| ricotz | brb | 17:13 |
| hyperair | Sarvatt: i've observed something rather interesting.once i'm past the first GPU hang, (ssh in and forcefully kill X and restart gdm), it doesn't hang any more. | 17:44 |
| hyperair | Sarvatt: also, the first hang happens very quickly after logging in, within the first 5 hours or so. | 17:45 |
| hyperair | my current uptime is 15h. Xorg's uptime is 12h | 17:46 |
| hyperair | er wait. 11h 23m | 17:46 |
| bryce2 | Sarvatt, tseliot, if lp #567007 sounds like a useful launchpad feature, mind clicking 'Affects me too'? The more people marked as it affects them, the more likely it'll be to get on the Launchpad team's planning list when work starts on search improvement | 17:49 |
| ubot4 | Launchpad bug 567007 in malone "User-specific default tag search parameters (affects: 1) (heat: 3)" [Low,Triaged] https://launchpad.net/bugs/567007 | 17:49 |
| Sarvatt | done, that would be a nice feature :) | 17:52 |
| Darxus | There's a fix for bug #541492. It involves 5 source packages (one of which is linux). What needs to be done to get them in Maverick? | 18:02 |
| ubot4 | Launchpad bug 541492 in xserver-xorg-video-intel (Ubuntu Lucid) (and 2 other projects) "MASTER: [i845] GPU lockup (apport-crash) (Should KMS be blacklisted?) (affects: 77) (dups: 30) (heat: 523)" [High,Triaged] https://launchpad.net/bugs/541492 | 18:02 |
| dupondje | everything seem to crash now :s | 18:07 |
| dupondje | ricotz: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/606393 you have this also ? | 18:37 |
| ubot4 | Launchpad bug 606393 in linux (Ubuntu) "BUG: unable to handle kernel NULL pointer dereference at 00000000000003c0 (affects: 1) (heat: 6)" [Undecided,New] | 18:37 |
| ricotz | dupondje, yes, this is plymouthd trying | 18:58 |
| ricotz | dupondje, should be the same problem which causes X to fail | 18:58 |
| bjsnider | Sarvatt, remember that issue with nvidia-current where it's creating two modprobe.conf files? | 19:51 |
| Sarvatt | bjsnider: ya want tseliot, but yeah I remember it :) | 20:00 |
| bjsnider | what i don't understand is, the alternatives are putting one of the files by a link into the /etc/modprobe.d directory, the one without the alias, and yet the system behaves as though it knows about the alias | 20:01 |
| Sarvatt | look at the dkms.conf | 20:05 |
| Sarvatt | DEST_MODULE_NAME[0]="nvidia-current" | 20:05 |
| Sarvatt | guessing thats why | 20:05 |
| bjsnider | yeah, the module is built as nvidia-current | 20:06 |
| bjsnider | but the xorg.conf says nvidia | 20:06 |
| bjsnider | without the alias the system should fail to load that module because it doesn't exist | 20:06 |
| bjsnider | and it intermittently does so | 20:07 |
| bjsnider | but for some reason it works most of the time. i can't see why. | 20:07 |
| Sarvatt | the xorg.conf doesn't have anything to do with it though, the ddx module is still nvidia_drv | 20:08 |
| bjsnider | then what is the alias used for? | 20:08 |
| Sarvatt | it only loads if it wasn't loaded by udev already at startup but yeah i thought it might run into the same problem because thats hardcoded to load nvidia.ko and dkms is just renaming it to nvidia-current.ko instead of aliasing nvidia to nvidia-current | 20:09 |
| Sarvatt | it's still built as nvidia.ko by dkms just installed as nvidia-current.ko after | 20:10 |
| bjsnider | without the alias i have tried modprobe nvidia, and the response is no such module exists | 20:10 |
| Sarvatt | you'd need to change that DEST_MODULE_NAME to nvidia and alias nvidia nvidia-current to do that | 20:11 |
| bjsnider | $ lsmod|grep nvidia | 20:12 |
| bjsnider | nvidia 11105770 38 | 20:12 |
| Sarvatt | the module is nvidia-current.ko though, its confusing :( | 20:14 |
| bjsnider | i think what you have here is the dx module is being renamed so that each nvidia module can be built and installed at the same time | 20:14 |
| bjsnider | i'm fairly sure putting driver "nvidia" in the xorg.conf file is the same as saying modprobe nvidia | 20:15 |
| Sarvatt | the xorg.conf driver you pick specifies which _drv to load bjsnider, nothing to do with kernel modules | 20:48 |
| Sarvatt | the X driver can load kernel modules though | 20:49 |
| Sarvatt | those nvidia_drv.so's are swapped around with the /usr/lib/xorg/extra-modules alternative pointing to the different driver's directories | 20:51 |
| Sarvatt | xf86LoadKernelModule() does it, i see that in nvidia_drv.so. it'll be trying to load nvidia that doesn't exist probably if it hasn't already been loaded by udev before | 20:55 |
| bryce2 | tjaalton, BEL? | 22:01 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!