| Sarvatt | Kangarooo: crashed with the PPA one yet? :) | 00:50 |
|---|---|---|
| Kangarooo | Sarvatt: no not jet. im tryng. i havent restarted pc jet. and had FF open with 100 gif tabs and i just came back and FF crashed but crash didnt had ram enough (that was the apport error). so if it crashes ill tell. also im tryng to make crash and not working crash. | 00:59 |
| === Kangarooo is now known as Guest17630 | ||
| === Guest17630 is now known as KaNGaR | ||
| KaNGaR | Sarvatt: just happened crash again :) so i make backtrace again? or just ubuntu-bug ? i also restarted pc maybe patch didnt made effect until restarted? | 01:08 |
| === KaNGaR is now known as Kangarooo | ||
| Kangarooo | Sarvatt: cant i in (gdb) make command with && ? like attach (pid) && cont ? so system doesnt freeze? | 01:17 |
| Kangarooo | no not && and not & working (gdb) attach 769 && cont Illegal process-id: 769 && cont. (gdb) attach 769 & cont Illegal process-id: 769 & cont. | 01:20 |
| RAOF | Kangarooo: You're not running gdb from a VT? Your X will still stop, but you won't be trying to interact with it, so it should be ok. | 01:24 |
| Kangarooo | whats VT? im was running gdb from pc witch is not so close thrue ssh conecting to crashed comp. and so that i dont need to run to working comp im in VNC to working comp :) | 01:25 |
| Kangarooo | RAOF: ive just tryd running this in TTY and when freezing happens (where i need to continue running with command cont) then TTY still works. But then i Ctrl+alt+f7 and again all froze. | 01:26 |
| RAOF | There are a couple of signals you need to not stop on. If you change the keyboard to raw mode (alt+sysrq+r) you should be able to switch back to your original TTY and continue. | 01:27 |
| Kangarooo | it would be very good if on (gdb) attach (pid) xorg doesnt stop | 01:27 |
| RAOF | Those signals should be on the X debugging wiki. | 01:28 |
| Kangarooo | whts sysrq? yesterday i was asking and nobody knew | 01:28 |
| maco | its a key on your keyboard | 01:28 |
| RAOF | The sysrq button; generally found on the same key as prtsc | 01:28 |
| Kangarooo | hehe no? | 01:28 |
| maco | usually on prtsc or ins or del | 01:28 |
| maco | mine's on the del key | 01:29 |
| Kangarooo | whata.. alt+del removed one of my workspaces | 01:29 |
| Kangarooo | that actually deleted a workspace | 01:30 |
| Kangarooo | wow. yes i have Sysrq on Prt Scr. wow never used so never noticed :) | 01:30 |
| Kangarooo | im only getting a screenshot. so when a freezing happens then (alt+sysrq+r) will work? | 01:31 |
| RAOF | To trigger sysrq you probably need to hit shift, too. | 01:32 |
| Kangarooo | ok and last buton is key "r" ? | 01:32 |
| RAOF | Yup. | 01:33 |
| Kangarooo | ok so that all will help me only to get to TTY ? but not to xmode ? | 01:33 |
| Kangarooo | couse it will still be frozen until in gdb continue | 01:34 |
| RAOF | Right | 01:40 |
| RAOF | That will allow you to switch terminals, because it makes the kernel handle the ctrl+alt+f keys itself without letting X trap them in its frozen jaws. | 01:41 |
| Kangarooo | ok so in TTY6 i made gdb and when made attach (pid) TTY6 still responding and keyb also. but then i changed to Xorg mode and keyb not responding and nothing moves. so i got in raw mode and that made only keyboard raw mode but not screen. keyboard responded to Numlock and to Scroll Lock but not caps lock. | 01:42 |
| Kangarooo | and ctrl+alt+f6 i could get back to tty6 | 01:42 |
| Kangarooo | not to any tty1-6 | 01:43 |
| Kangarooo | RAOF: so raw mode doesnt give me possibility for me to get back to TTY1-6 | 02:02 |
| Kangarooo1 | is gdb still running or it stopped? http://pastebin.com/8Scimr30 what command should i do now? again attach (pid) or cont ? | 02:10 |
| RAOF | That hasn't worked because X is already running. gdb is running, but the X process you tried to start isn't. | 02:14 |
| RAOF | Ok. Git question: how do I take uncommitted changes over to a different branch? | 02:19 |
| Kangarooo | no i cant get in TTY back. i even did whats in subsection of https://wiki.ubuntu.com/X/Backtracing#Backtrace%20with%20gdb section gdb problems. theres writen inside gdm, start up Xorg: (gdb) run -keeptty -dumbSched. so i did start with attach (pid) and then that run -blabla -bla and thats the log i got. couse after that command it asked to restart? on yes x restarted to login screen. on no it crashed when going to tty7 and still cant get bac | 02:21 |
| Kangarooo | also the link for gdb problems has nothing to do with the command posted in that section | 02:22 |
| Bernardo | good morning | 06:03 |
| RAOF | Hm. DisplayLink doesn't really work. | 07:55 |
| Sarvatt | is it bad i'm starting to actually look forward to intel being broken in git? i swear i learn so much each time trying to figure it out :) | 09:22 |
| Sarvatt | RAOF: drivers in the archive? | 09:22 |
| RAOF | Sarvatt: Yeah. | 09:23 |
| Sarvatt | werent those really old? | 09:23 |
| RAOF | kernel fb module in the kernel, xserver-xorg-video-displaylink as the DDX. | 09:23 |
| RAOF | I'm not sure whether there's anything newer. | 09:23 |
| Sarvatt | dont think i've even seen any displaylink stuff since... xserver 1.6 just came out | 09:24 |
| Sarvatt | on the lists | 09:24 |
| Sarvatt | where's upstream? | 09:24 |
| RAOF | Well, bryceh gave me a displaylink box at UDS. It's easy enough to get a framebuffer up, once you manage to switch it from USB mass storage mode. | 09:25 |
| Sarvatt | needs a udev rule? | 09:25 |
| RAOF | Upstream? http://libdlo.freedesktop.org/wiki/ | 09:25 |
| RAOF | Yeah. A udev rule would do it. | 09:25 |
| RAOF | At least for _this_ hardware. No guarantees that other hardware isn't crazy in different ways :) | 09:26 |
| Sarvatt | nothing like spending 8 hours or so bisecting and having the fix already commited :D | 09:34 |
| Sarvatt | why is it called libdlo? | 09:36 |
| * Sarvatt is confused | 09:36 | |
| Sarvatt | ah i see it talks to libusb, where's the X driver? | 09:38 |
| Sarvatt | oh you linked a wiki too! ugh, sorry, i've been going through xtrace logs all night and text is kind of blurring together | 09:39 |
| Sarvatt | oh so basically you just use fbdev? | 09:41 |
| Sarvatt | Now i will focus on patching the intel driver to add a virtual crtc for every displaylink device in the system. This will allow to use the GPU to draw (maybe accelerated) contents and send the output via the usb,and to use a single big screen (the old Xinerama/MergedFB way). | 09:44 |
| Sarvatt | (then releases stop a year ago) :) | 09:44 |
| RAOF | Right. | 09:54 |
| RAOF | It did seem to be fairly dead. | 09:55 |
| RAOF | I can get a screen with a movable cursor up, but there's not a lot of rendering happening. | 09:55 |
| Sarvatt | no real point signing up for things on blueprints if you're not canonical right? | 10:04 |
| RAOF | Hm. Kinda. | 10:06 |
| Sarvatt | i haven't touched anything on them because i'd rather one of you guys get marks on your performance reviews :) | 10:07 |
| RAOF | I'm not entirely sure why you'd want to sign up for work items; I guess it's a nice tracker. | 10:07 |
| Sarvatt | because i'm doing or have done the work, i wasn't sure if they really were just canonical or ubuntu too | 10:10 |
| RAOF | Hm. You might want to ask Rick. I don't see any particular reason why you couldn't put work items on the list if you wanted to track things. Or, rather, the only reason I can think of is that one of us needs to be ultimately responsible for ensuring it's delivered. | 10:13 |
| Sarvatt | you a core-dev yet? :) | 10:43 |
| RAOF | Not yet :) | 10:48 |
| Sarvatt | I'm not sure, seems like a lot of the problem is on the AGP side | 10:57 |
| apw | RAOF, got a link to the bug from which that attachment attaches? | 10:58 |
| RAOF | apw: http://bugs.freedesktop.org/show_bug.cgi?id=27187 | 10:59 |
| ubot4 | Freedesktop bug 27187 in Driver/intel "[i855GM] gtt chipset flush is not cache coherent" [Normal,New] | 10:59 |
| RAOF | Linked from this memorable launchpad bug: https://bugs.edge.launchpad.net/ubuntu/lucid/+source/xserver-xorg-video-intel/+bug/541511?comments=all | 11:00 |
| apw | bug #541511 | 11:01 |
| * apw notes ubot4 is useless | 11:01 | |
| RAOF | I'm off to the movies now; it's 8pm. | 11:01 |
| ubot4 | RAOF: Error: Bug #541511 is private. | 11:01 |
| ubot4 | Launchpad bug 541511 in xserver-xorg-video-intel (Ubuntu Lucid) (and 3 other projects) "MASTER: [i855] GPU lockup (apport-crash) (affects: 118) (dups: 29) (heat: 814)" [High,Triaged] https://launchpad.net/bugs/541511 | 11:01 |
| RAOF | Hah. | 11:02 |
| Sarvatt | there's been a surprising amount of 8xx stuff done post 2.11 though and one person mentioned they were 100% stable on 855 earlier | 11:02 |
| RAOF | I wouldn't get my hopes up about _that_. A subset of users have always reported they were 100% stable on 855. | 11:03 |
| Sarvatt | true! | 11:03 |
| * RAOF really off. | 11:03 | |
| Sarvatt | apw: do you know if there are there any plans to enable DMAR? | 11:10 |
| apw | Sarvatt, DMAR ? | 11:11 |
| Sarvatt | just a heads up you can expect a large amount of bug reports about broken graphics if you do :) | 11:11 |
| Sarvatt | yeah some people want it on for virtualization but we've avoided all of the problems because we dont have it enabled :) | 11:13 |
| Kangarooo | hello Sarvatt did u read all i wrote after u asked Sarvatt: Kangarooo: crashed with the PPA one yet? :) | 13:00 |
| Kangarooo | i was tryng to get somehow so Xdoesnt stops with attach (pid) handle all nostop handle all pass cont but still xorg was stopped when crash happened. ouh and yeah crash still happened. ill make a new dbg now thrue ssh again | 13:02 |
| mvo | superm1: hey! I merged xserver-xorg-video-nv in lp:ubuntu/maverick/xserver-xorg-video-nv but I ran into the issue that nv_list.csv seems to need updating. is there a automatic way of doing this? | 14:07 |
| mvo | superm1: all the other bits are in bzr, but that part is currently missing | 14:07 |
| Kangarooo | Sarvatt: i got again crash but crashed all comp not just screen. also ssh wasnt responding. but also i could quit ssh and also couldnt open new ssh | 14:09 |
| Sarvatt | Kangarooo: did you attach that backtrace you got to the bug? I looked last night but didnt see it | 14:11 |
| Sarvatt | mvo: can't merge X drivers from debian at the moment until xserver is updated.. | 14:11 |
| Sarvatt | well you can but they wont build :D | 14:11 |
| mvo | Sarvatt: can merge, just can't build :) | 14:11 |
| Kangarooo | no couse didnt get response is it needed. ill post it up now. and then im again tryng to make crash with backtrace with your ppa | 14:12 |
| mvo | Sarvatt: its just the depends generation that will not work, so it can be tested to a certain extend :) | 14:12 |
| Sarvatt | ohhh interesting, never saw superm1's patch before | 14:14 |
| lucazade | Bernardo: http://meego.gitorious.org/~xinyunliu/meego-os-base/xylius-kernel-source | 14:14 |
| lucazade | Initial PowerVR core gfx driver support for IVI based on build 1616 | 14:14 |
| Sarvatt | it shouldn't really be needed anymore, should it? | 14:14 |
| mvo | Sarvatt: hints welcome (feel free to take over if you want, my stuff is in bzr) | 14:14 |
| mvo | Sarvatt: no? what did change that makes it no longer needed? | 14:15 |
| Sarvatt | nv no longer loads if KMS is in use so it'd never claim it, KMS is in use unless forced off for nouveau | 14:15 |
| Sarvatt | i think there was more, looking | 14:15 |
| Sarvatt | what i'm thinking of might not have made an official release yet | 14:16 |
| Sarvatt | yeah that csv generation would be a PITA to maintain lol | 14:17 |
| Sarvatt | well then again nv is abandoned | 14:17 |
| mvo | PITA> my feeling as well | 14:17 |
| Sarvatt | not like they're going to make GPU's that work with nv anymore to update it after this last one | 14:17 |
| Sarvatt | i'll fix it up | 14:18 |
| mvo | cool, thanks :) | 14:18 |
| lucazade | Sarvatt: hi! seems there is a new IVI driver with poulsbo support | 14:22 |
| lucazade | in meego git repo | 14:23 |
| Kangarooo | ok Sarvatt added. now with your ppa doing the same commands made system and ssh unresponsive.. :( | 14:23 |
| Kangarooo | *tryng backtracing with same commands. | 14:23 |
| Sarvatt | lucazade: found any images yet? | 14:23 |
| Sarvatt | i've got a meego chroot but theres still nothing as of this morning | 14:24 |
| lucazade | not yet | 14:24 |
| Kangarooo | Sarvatt: bug 587710 | 14:24 |
| ubot4 | Launchpad bug 587710 in nvidia-graphics-drivers (Ubuntu) "if FF wit lot Gif & overloading CPU then X-boom to login (affects: 1) (heat: 6)" [Undecided,Confirmed] https://launchpad.net/bugs/587710 | 14:24 |
| lucazade | Sarvatt: found this repo reading meego forum http://forum.meego.com/showthread.php?t=38 | 14:25 |
| Sarvatt | i'm in the middle of doing 3 things at once right now, will look in a bit | 14:25 |
| Kangarooo | Sarvatt: iim tryng to get system not lock up so i dont need to cont on other comp. (gdb) attach 787 -keeptty -dumbSched gives Illegal process-id: 787 -keeptty -dumbSched. also tryd just handle all nostop and handle all pass then cont (without handle 11 stop) and still system froze | 14:26 |
| Sarvatt | what the heck, superm1 made this -nv change last september? no clue how I didnt see it all this time | 14:28 |
| bjsnider | i can't make sense of that bug report subject line | 14:28 |
| Kangarooo | bjsnider: mine bug report? | 14:29 |
| Sarvatt | superm1: ah theres a little problem with this patch | 14:36 |
| Sarvatt | superm1: nvidia GPU's that are natively PCIE but have an AGP bridge are not in the list | 14:37 |
| Sarvatt | so you blocked out all of those users with this change | 14:37 |
| Sarvatt | the GPU has the pci id of the agp bridge and not the actual GPU | 14:38 |
| Sarvatt | and adding the agp bridge pci ids to the list would screw it up too, i went through this with aplattner a year ago | 14:38 |
| Sarvatt | it was a problem when we had the pci.id's files, that wasn't picking up those GPU's either | 14:39 |
| Sarvatt | dont tell me i dont have the patch i used to get around that anymore.. :( I added all of the bridge devices in an #if 0 block so they didn't do anything but they still got parsed | 14:40 |
| Sarvatt | there it is - http://launchpadlibrarian.net/27885379/101_agp_bridge_pci_id.patch | 14:56 |
| Sarvatt | almost done updating the csv with the new chips so the script can be run | 15:04 |
| superm1 | mvo, there is a script in there yes to do it that came with the patch | 15:07 |
| superm1 | Sarvatt, the patch was supposed to be something to get us by until upstream X got support to find the devices sooner | 15:07 |
| superm1 | Sarvatt, er not find them sooner, i think it was so that it could fail earlier and not fail so hard | 15:07 |
| superm1 | Sarvatt, since we're nouveau by default, i'd say the patch is pretty moot now though | 15:08 |
| Sarvatt | yeah but some devices relied on that broken behavior and the maintainer didnt want to have anything to do with changing it, we had a lot of bugs about non working AGP cards :( | 15:09 |
| superm1 | if i remember right slangasek raised something like that | 15:15 |
| superm1 | but since they were kicked to vesa (and were probably going to install -nvidia anyway) this patch made most sense | 15:16 |
| superm1 | it certainly fixed a bunch of hardware I had that *didn't* work with nv, but worked with vesa (and thankfully works with nouveau now) | 15:17 |
| Sarvatt | I stopped messing with it mvo, I really agree dropping the 2 patches is the way to go :) | 15:31 |
| Sarvatt | Kangarooo: hmm your backtrace is completely different now | 15:38 |
| mvo | Sarvatt, superm1: so what is the best way forward? dropping -nv entirely ? or just dropping the patch and upload the merged -nv? | 15:39 |
| Sarvatt | dropping the 2 patches yeah, not sure that the panel fitting is worth merging over just syncing it either. we could just special case the ion board that had a problem, the problem superm1 was having with multiple gpu's where one wasnt supported by nv but the other was shouldnt be around anymore either | 15:43 |
| lucazade | Sarvatt: MeeGo 1.0 In-Vehicule Preview ntel eMenlow (US15W) http://vimeo.com/11779730 | 15:44 |
| Sarvatt | lucazade: got an img? :) | 15:44 |
| Sarvatt | or a .ks? | 15:44 |
| lucazade | http://repo.meego.com/MeeGo/builds/0.9/0.9.80.2.20100512.1/ivi/images/meego-preview-ivi-noemgd-ia32/ | 15:44 |
| lucazade | maybe.. | 15:44 |
| lucazade | bingo? | 15:45 |
| lucazade | :) | 15:45 |
| Sarvatt | nope thats not it | 15:45 |
| Sarvatt | i have that one | 15:45 |
| lucazade | in the vimeo description says us15w | 15:45 |
| Sarvatt | This image will boot to XFCE utilizing the Linux VESA graphics driver and has been tested on Intel eMenlow (US15W) and Tunnel Creek platforms. A Beta release of the Intel EMGD 3D graphics driver for MeeGo on these platforms is expected in June 2010. | 15:46 |
| Sarvatt | there's nothing psb/img/mrst related in the repos either besides psb-kernel-headers, just checked | 15:48 |
| lucazade | ok thanks | 15:49 |
| Sarvatt | none of the kernel patches on there, no modules for it | 15:49 |
| Sarvatt | none of the stuff in the kernel-headers package either :( | 15:49 |
| lucazade | on #meego vgrade said: "no the its not the same patch as the moorsetown one, it looks quite similar to a traditional SGX driver." | 15:51 |
| Sarvatt | yeah they have a straight up powervr binary image, i've seem them talking about it in bug reports | 15:52 |
| lucazade | http://bugs.meego.com/show_bug.cgi?id=2205 | 15:53 |
| ubot4 | bugs.meego.com bug 2205 in Graphics subsystem "request integration of EMGD into ivi/embedded distribution" [Major,Reopened] | 15:53 |
| Sarvatt | so far i've seen 4 different ones, pvr mrst img and psb | 15:55 |
| Sarvatt | + } else if ((dev->device_id == 0x8108) || (dev->device_id == 0x8109) || (dev->device_id == 0x4102)) { | 15:56 |
| Sarvatt | +driverList[0] = "psb"; | 15:56 |
| Sarvatt | +driverList[1] = "pvr"; | 15:56 |
| Sarvatt | pvr supporting psb! | 15:56 |
| lucazade | :) | 15:56 |
| Sarvatt | unless they are just lazy/sloppy :) | 15:56 |
| lucazade | unbelievable | 15:57 |
| Sarvatt | oh | 15:57 |
| Sarvatt | yeah, i remember now, one of the bug reports was saying there were graphics problems with the pvr internal images, | 15:58 |
| Sarvatt | they patched around the problems in mesa and i found the bug number from one of the patches, let me find it | 15:58 |
| lucazade | ok | 15:58 |
| Sarvatt | Kangarooo: try sudo nvidia-xconfig --no-composite | 15:58 |
| Sarvatt | http://bugs.meego.com/show_bug.cgi?id=352 | 16:03 |
| ubot4 | bugs.meego.com bug 352 in Graphics subsystem "[Qt demos] opengl/Pixel Buffers failed due to not support pbuffers" [Normal,Verified: fixed] | 16:03 |
| Sarvatt | Pvr driver: 5.3.0.0002-1.1 | 16:04 |
| lucazade | yes | 16:04 |
| Sarvatt | the mrst kernel module has a similar version | 16:07 |
| Sarvatt | hmm should events and such from include/GL/glxext.h be included in xserver's dix/protocol.txt and xtrace? | 16:22 |
| Sarvatt | i noticed some events from there is in xtrace that aren't defined in glproto | 16:23 |
| lucazade | i don't know, sincerly.. i know just a bit of xorg | 16:26 |
| Sarvatt | CONFIG_IMG_DOES_NOT_SUPPORT_MENLOW=y | 16:27 |
| Sarvatt | in the menlow kernel config | 16:28 |
| Sarvatt | ahh ok theres img the pvr one, then theres mrst the poulsbo one | 16:29 |
| lucazade | i trust you! | 16:29 |
| Sarvatt | err, ok, i'll shut up. the poulsbo kernel config has all the pvr stuff enabled | 16:29 |
| Sarvatt | need to look at the source and see whats up, darn gitorious doesnt let you browse large stuff | 16:30 |
| Sarvatt | http://sarvatt.com/git/cgit.cgi/xylius-kernel-source/ | 16:31 |
| Sarvatt | so the mrst one is the open source one and thats the closed source one | 16:32 |
| lucazade | ok | 16:33 |
| Sarvatt | man trying to load that latest commit on this atom cpu is insane | 16:35 |
| Sarvatt | going on 3 minutes now and i still cant scroll down | 16:36 |
| lucazade | :) everything related to psb look insane to me | 16:36 |
| Sarvatt | yeah mrst the one they had before has that its going to try to go upstream, this one says upstream: never | 16:36 |
| Sarvatt | perfect timing though i got a psb netbook coming to me in a few days :) | 16:37 |
| lucazade | great.. i can test if you need | 16:38 |
| lucazade | if i'm able | 16:38 |
| Sarvatt | i guess iged changed names to emgd | 16:38 |
| Sarvatt | or something.. | 16:39 |
| lucazade | ah ok... | 16:39 |
| Dr_Jakob | oh so they have released that driver as well. | 16:39 |
| * Sarvatt waits to find Dr_Jakob's name in a copyright :) | 16:41 | |
| Dr_Jakob | Probably not | 16:41 |
| Dr_Jakob | I only touched the modesetting code and stuff inside the 3D driver. | 16:42 |
| Dr_Jakob | Oh and one or two things inside the 2D driver.. | 16:42 |
| Dr_Jakob | but that was for the none iged driver. | 16:43 |
| Dr_Jakob | Sarvatt: what parts of the new vmwgfx svga driver is in the gallium ppa? | 16:55 |
| Dr_Jakob | vmware 10.0.1, mesa master? | 16:55 |
| Sarvatt | 1:11.0.1+git20100423.423d8a06-0ubuntu0sarvatt but no svga in mesa yet, hit a snag because the arm people need opengles headers to build stuff and there's a lot of packaging things to work out | 16:56 |
| Sarvatt | not to mention its completely different on master than 7.8.1 where RAOF is packaging the stuff up at | 16:57 |
| Sarvatt | the gallium stuff not being in /usr/lib/dri really stinks, --with-dri-searchpath isnt really an option outside of testing | 17:00 |
| bryceh | Sarvatt, RAOF, there is a -displaylink X driver package that's in universe | 17:01 |
| Sarvatt | it'd be nice if all the gallium drivers had different names and it would check the g suffix one if KMS was in use or something :) | 17:01 |
| Sarvatt | err try to load the r300g_dri.so first and fall back to r300 if it can't I mean, I don't know | 17:02 |
| Sarvatt | heck, i can just enable svga and it'll be installed with /usr/lib/dri until its packaged up, lets try that :) | 17:09 |
| Sarvatt | oh yeah the _drv.. need to build and see where it goes | 17:09 |
| Sarvatt | still can't download IEGD from intel in any browser, just downloads the windows version no matter what I pick | 17:19 |
| Dr_Jakob | Sarvatt: what about kernel drivers? Does the ppa include new kernel drivers as well? | 17:21 |
| lucazade | if i remember well inside iegd windows executable there is linux stuff | 17:21 |
| Sarvatt | yeah its got maverick's kernel in it but you have to install it manually, its linux-image-2.6.34-5-generic right now | 17:22 |
| Sarvatt | linux-libc-dev which has the drm headers in it from the kernel automatically upgrades though | 17:22 |
| Sarvatt | libdrm has libkms built as well and the vmwgfx header | 17:22 |
| Dr_Jakob | ok | 17:22 |
| Dr_Jakob | have you tested it all in a VM? | 17:23 |
| Dr_Jakob | I don't know if the driver still barfs at QEMU/KVM. | 17:24 |
| Sarvatt | yeah it works, didnt test 3D or anything though | 17:25 |
| Sarvatt | (dont know if thats what you meant) | 17:25 |
| Dr_Jakob | Ok cool | 17:26 |
| Dr_Jakob | you sure you didn't get the old driver? | 17:26 |
| Sarvatt | i'm sure I did | 17:26 |
| Dr_Jakob | ok | 17:27 |
| Sarvatt | never build mesa with svga enabled, doing that now | 17:27 |
| Dr_Jakob | ok | 17:28 |
| Dr_Jakob | as long as you install vmware 10.0.1 it should auto pick the correct driver. | 17:28 |
| Dr_Jakob | err | 17:28 |
| Dr_Jakob | 11.0.1 | 17:28 |
| Dr_Jakob | but yeah you had that. | 17:29 |
| Sarvatt | all i need to know is where the _drv is when adding xorg to the state trackers list and building it and i'll upload it to edgers | 17:29 |
| Sarvatt | just because of the weird places debianized mesa puts things | 17:30 |
| Kangarooo | Sarvatt: that backtrace i added is still without your PPA. icant get backtrace with your ppa since comp not only freezes desktop but also freezes ssh | 18:16 |
| Sarvatt | well at least thats different :) | 18:17 |
| Sarvatt | did you try sudo nvidia-xconfig --no-composite ? | 18:18 |
| Kangarooo | should i try that sudo nvidia-xconfig --no-composite with your ppa installed or remove them and then with original files try that? | 18:18 |
| Sarvatt | yeah get rid of the PPA one | 18:19 |
| Sarvatt | your backtraces look different every time | 18:20 |
| Sarvatt | http://launchpadlibrarian.net/49499123/gdb-Xorgwithtail-fXorg.c.txt http://launchpadlibrarian.net/49412494/GdmLog1.txt http://launchpadlibrarian.net/49412495/GdmLog2.txt | 18:21 |
| Sarvatt | doh, mesa didn't check for xorg-server.pc before starting the compile with xorg state tracker? | 18:22 |
| Sarvatt | silly me forgot it in the chroot | 18:24 |
| Kangarooo | so ppa i can remove from synaptic only? so when i uncheck and use aptitude reload && dist-upgrade yes? | 18:25 |
| Sarvatt | ok easy enough - dh_install: usr/lib/glx/pkgconfig/gl.pc exists in debian/tmp but is not installed to anywhere | 18:28 |
| Sarvatt | dh_install: usr/lib/glx/xorg/modules/drivers/radeong_drv.so exists in debian/tmp but is not installed to anywhere | 18:28 |
| Sarvatt | dh_install: usr/lib/glx/xorg/modules/drivers/modesetting_drv.so exists in debian/tmp but is not installed to anywhere | 18:28 |
| Sarvatt | dh_install: usr/lib/glx/xorg/modules/drivers/vmwgfx_drv.so exists in debian/tmp but is not installed to anywhere | 18:28 |
| Sarvatt | dh_install: usr/lib/glx/xorg/modules/drivers/i965g_drv.so exists in debian/tmp but is not installed to anywhere | 18:28 |
| Sarvatt | Kangarooo: wget https://edge.launchpad.net/~xorg-edgers/+archive/ppa/+files/ppa-purge_0.2.6_all.deb && sudo dpkg -i ppa-purge_0.2.6_all.deb && sudo ppa-purge -p bugsbugsbugs sarvatt | 18:32 |
| Sarvatt | its a little script that removes a ppa and downgrades everything back to the archives | 18:34 |
| Kangarooo | ok that is done. also some upgrades i put. so now i without restart can sudo nvidia-xconfig --no-composite ? | 18:41 |
| Kangarooo | so now ill also try reproducint with backtrace gdb runnin yes? | 18:44 |
| Sarvatt | yeah without restart, it just adds a line to your xorg.conf | 18:44 |
| Sarvatt | nah | 18:44 |
| Sarvatt | just see if you can crash it | 18:44 |
| Sarvatt | Dr_Jakob: i'll probably have it up on xorg-edgers in the next hour, almost done with it | 18:51 |
| Sarvatt | is there anything else in mesa that would be worth packaging up? | 18:53 |
| Sarvatt | i really want to wrap my head around all of this EGL stuff but it's crazy complicated in mesa | 18:54 |
| Kangarooo | Sarvatt: nope that didnt help. again boom goes the x | 19:36 |
| Sarvatt | is there anything special about your machine? | 19:40 |
| Sarvatt | you seem to be having problems with that one in a ton of different packages, just seems like it might be a bigger problem but thats reaching for straws :) | 19:42 |
| Kangarooo | Sarvatt: nothing special. i tink lspci can be seen in bug report no? just that hardware and clean xubuntu 10.04 (of course with latest updates) but with ur ppa it wasnt crashing. ill maybe try again your ppa and without dbg try to crash it. couse with gdb tryng crash all froze (even ssh) | 19:56 |
| === Kangarooo is now known as Kangaroo | ||
| === Kangarooo1 is now known as Kangarooo | ||
| Kangarooo | Sarvatt: u removed ur ppa? i get an error when using update after adding ur ppa again | 20:12 |
| Kangarooo | ah now its working- i added ppa without /ubuntu and got error. rm thoyuse sources and added ppa again with /ubuntu now its working | 20:16 |
| Sarvatt | sorry, trying to get this vmwgfx packaged up because it finally finished building. why are you activating the ppa again? | 20:17 |
| Kangarooo | couse maybe that will be working? couse ppa i tryd crashing with gdb working also. maybe without gdb system wont freeze totaly? | 20:18 |
| Sarvatt | naah that patch I added really was a longshot and probably did make things worse, it sounds like its a bigger problem and its over my head | 20:20 |
| Kangarooo | ah ok. Sarvatt so i should undo that nvidia command also? ill buy a replacement for this pc after 2 weeks if for somebody it can be needed i can put vnc and leave everything open in it. | 20:23 |
| Kangarooo | in this crashing pc | 20:23 |
| Sarvatt | i'd try with all 3 options disabled | 20:37 |
| Sarvatt | no-composite no-render and no-render-accel? (not sure on the last one) | 20:38 |
| Sarvatt | those might not be right, check nvidia-xconfig --help | 20:38 |
| Sarvatt | Dr_Jakob: sorry it took so long, uploading it now :) | 20:52 |
| Sarvatt | now to find a vmware release to play with this on :) | 21:16 |
| Sarvatt | i dont think the vmwgfx kernel module works with qemu vmware graphics mode? remember airlied having a hacked up patch for it | 21:17 |
| === Kangarooo1 is now known as Kangarooo | ||
| Sarvatt | that guys post about harvest the other day on planet ubuntu really made me want to automate xorg-edgers more via something like that :) | 21:40 |
| bryceh | totally, go for it | 21:58 |
| bryceh | it was always my hope too that more of xorg-edgers could be automated | 21:58 |
| bryceh | well you know how much of an automation geek I am ;-) | 21:59 |
| Sarvatt | darn, its java! yuck | 22:04 |
| === schmidtm_ is now known as schmidtm | ||
| Sarvatt | i guess i could just use ppa-update, never actually tried it | 22:46 |
| Sarvatt | still lots of little hidden env variables i dont know about in auto-xorg-git, i just do things like make the package for other releases manually every time | 22:47 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!