| billybigrigger | can someone take a look at this error? | 00:06 |
|---|---|---|
| billybigrigger | http://pastebin.ca/1492112 | 00:06 |
| Sarvatt | its a problem with kernel-package with nvidia-common thats been around for a looong time billybigrigger | 00:30 |
| Sarvatt | were there changes to usb in 2.6.31 that need updated userland support or something? been having alot of problems with usb devices after 2.6.30 | 00:33 |
| billybigrigger | my camera, ipod, and videocamera all work good | 00:34 |
| billybigrigger | its just the webcam | 00:34 |
| billybigrigger | but that could probably because they're all recognized as mass storage | 00:34 |
| Sarvatt | i see usbfs moved over to CONFIG_EMBEDDED, dont know if we use that before | 00:34 |
| Sarvatt | my webcam and android phone stopped working | 00:36 |
| billybigrigger | Sarvatt, no dice | 00:39 |
| Sarvatt | thanks for checking | 00:40 |
| billybigrigger | its there | 00:40 |
| billybigrigger | but webcam not working | 00:40 |
| billybigrigger | billybigrigger@cabo:/proc/bus/usb$ ls | 00:41 |
| billybigrigger | billybigrigger@cabo:/proc/bus/usb$ | 00:41 |
| billybigrigger | empty, but its there | 00:41 |
| Sarvatt | oh so usbfs was used in .30 where it was working before? | 00:43 |
| Sarvatt | hmm | 00:43 |
| billybigrigger | funny thing, i have sound in flash on this kernel :P | 00:44 |
| johanbr | my Android phone works fine with 2.6.31 | 00:49 |
| johanbr | my webcam behaves a bit funny though, but I'm not sure if that's the kernel's fault | 00:49 |
| Sarvatt | i get a flood of error messages in dmesg while its plugging in without enabling mass storage on the phone (like when i'm just charging it) | 00:50 |
| johanbr | seems to work for me | 00:51 |
| johanbr | the webcam is definitely screwed up, though: http://nullinfinity.org/tmp/screenshot6.png | 00:55 |
| johanbr | (I am actually not a smurf) | 00:55 |
| billybigrigger | johanbr, nvidia user? | 00:55 |
| johanbr | yep | 00:55 |
| billybigrigger | looks like the nvidia hue -1000 bug | 00:55 |
| billybigrigger | :) | 00:55 |
| johanbr | oh | 00:56 |
| johanbr | do you have a launchpad link? | 00:56 |
| billybigrigger | yup lemme dig it up | 00:56 |
| Sarvatt | yeah change the saturation under 0 in nvidia-settings then save, then back up to 0 and save | 00:56 |
| billybigrigger | Sarvatt, funny thing was, i was affected by it, and my hue was at 0 | 00:56 |
| billybigrigger | just -1000 in vlc | 00:56 |
| Sarvatt | they changed the xv saturation ranges in the driver and our old nvidia-settings doesnt copy with it | 00:56 |
| Sarvatt | yeah because it isnt really at 0 | 00:56 |
| billybigrigger | so it might be on an app to app basis | 00:56 |
| johanbr | billybigrigger, you're absolutely right | 00:56 |
| johanbr | thanks | 00:56 |
| billybigrigger | ooooooh | 00:56 |
| Sarvatt | it thinks its at 0 but its really at 4096 from the old ranges | 00:57 |
| johanbr | I hadn't noticed that regular video had funny colours too | 00:57 |
| billybigrigger | https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-180/+bug/395476 | 00:57 |
| ubot3 | Malone bug 395476 in totem "nvidia sets HUE to -1000" [Low,Triaged] | 00:57 |
| Sarvatt | things that use their own saturation settings arent affected but things that use nvidia-settings ones are | 00:57 |
| billybigrigger | bugabundo started that one, but im pretty sure it should be against nvidia-glx-180 and not totem | 00:57 |
| billybigrigger | Sarvatt, that makes sense | 00:58 |
| billybigrigger | Sarvatt, will i be able to file bugs under my custom kernel? | 01:27 |
| billybigrigger | or will they have to be done upstream to the package maintainers | 01:27 |
| Sarvatt | just file the bug on the ubuntu kernel since its the same | 01:34 |
| billybigrigger | ok | 01:34 |
| === bjf is now known as bjf_afk | ||
| === Omegamoon is now known as Omegamoon|away | ||
| === Sarvatt__ is now known as Sarvatt_ | ||
| thrice` | I have a question; any reason 2.6.30.1 hasn't shown up in the mainline PPA ? | 16:05 |
| billybigrigger | git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git | 18:13 |
| billybigrigger | Sarvatt, that is where i pulled the latest kernel source | 18:14 |
| billybigrigger | so when i want to update it... | 18:14 |
| billybigrigger | git pull from the same source right? | 18:14 |
| thrice` | "git pull" | 18:14 |
| billybigrigger | 10-4 | 18:14 |
| billybigrigger | there is life in here! :P | 18:14 |
| billybigrigger | thrice`, and that won't compile just the new stuff | 18:19 |
| billybigrigger | it will pull whatever is new, and recompile the whole shebang again right? | 18:19 |
| thrice` | yep | 18:19 |
| thrice` | compiling is not related to obtaining of the source | 18:19 |
| billybigrigger | no, thats separate | 18:20 |
| billybigrigger | fatal: Not a git repository (or any of the parent directories): .git | 18:22 |
| thrice` | are you inside the dir? | 18:23 |
| billybigrigger | no, was one above it | 18:23 |
| billybigrigger | ok now i get an error that i didn't tell git what branch i wanted to merge with | 18:24 |
| billybigrigger | http://pastebin.ca/1492879 | 18:24 |
| thrice` | just do "git pull" on its own | 18:25 |
| billybigrigger | oh | 18:25 |
| thrice` | ooh, 2.6.31.1 seems to break jfs booting from grub1 :( | 18:41 |
| thrice` | that's probably why it's not in the PPA :> | 18:42 |
| sn9 | thrice`: wrong | 18:44 |
| sn9 | that has been broken since at least hardy | 18:44 |
| sn9 | and probably gutsy | 18:45 |
| thrice` | huh? | 18:50 |
| thrice` | 2.6.29 boots my jfs just fine, as does 2.6.30 | 18:50 |
| sn9 | it is not a kernel issue | 18:51 |
| sn9 | grub1 is bitrotted | 18:51 |
| thrice` | of course :) | 18:51 |
| sn9 | if grub cannot read the fs, the kernel is never loaded, regardless of version | 18:51 |
| thrice` | i'm trying to say that 2.6.30.1 introduced a specific failure | 18:52 |
| sn9 | it did not; that failure has been there for some time | 18:52 |
| reto` | hey... I wanted to ask what I can do about the following problem: | 18:52 |
| thrice` | you know which I'm talking about sn9 ? | 18:52 |
| sn9 | it is why i have an ext3 /boot on this laptop | 18:52 |
| reto` | with the new kernel it seems to | 18:53 |
| reto` | support cpufreq for the via cpu... but in a strange | 18:53 |
| reto` | way... before it was always at max speed... now it | 18:53 |
| reto` | looks like it changes speed but I think it's not... I | 18:53 |
| reto` | can't even get it to run at full speed | 18:53 |
| reto` | anymore... compared the performace with the two | 18:53 |
| reto` | kernel versions | 18:53 |
| reto` | oops... sry for the line breaks | 18:53 |
| thrice` | sn9: I am specifically referring to: http://bugs.archlinux.org/task/15416 | 18:54 |
| sn9 | reto`: is the same governor used? | 18:54 |
| thrice` | git commit 206f0f05bdc291a9358ba59248e2bc44e8b3127d | 18:55 |
| reto` | sn9: uh... the old kernel doesn't support cpufreq | 18:55 |
| sn9 | reto`: which via? c7? | 18:55 |
| reto` | sn9: I can try any governor or freq setting with the new kernel | 18:55 |
| reto` | sn9: it doesn't change anything | 18:56 |
| reto` | sn9: c7-m | 18:56 |
| sn9 | thrice`: the bug report is wrong; jfs has been unreadable by grub for quite some time | 18:57 |
| thrice` | what are you talking about??? | 18:57 |
| thrice` | I boot jfs from 2.6.29.6 right now | 18:57 |
| sn9 | thrice`: how big is the fs? | 18:59 |
| reto` | sn9: should I report this somewhere or how is the situation with the VIA? | 18:59 |
| thrice` | size does not matter | 18:59 |
| sn9 | reto`: are you using the longhaul driver, or e_powersaver? | 19:04 |
| sn9 | thrice`: then why has jfs booting consistently failed for me on every machine after feisty? | 19:04 |
| thrice` | I don't know. I promise that it works with 2.6.29.x , and even 2.6.30 :) | 19:05 |
| thrice` | the bug I mentioned was a specific change to 2.6.30.1 regarding the creation of an initrd | 19:05 |
| reto` | sn9: hmm... I think the longhaul... it's displaying some message about that one though on boot | 19:05 |
| sn9 | the grub people told me this bug has been there since 2004 | 19:06 |
| sn9 | reto`: check through the /sys tree | 19:06 |
| thrice` | ok, i'm not really interrested in arguing :( My system booted just fine this morning with jfs on / | 19:06 |
| === MobileMyles6o7 is now known as TwoToneSpirit | ||
| sn9 | thrice`: the real solution would be for ubuntu to take the bold step of dumping grub1 as a default | 19:09 |
| thrice` | or lilo :> | 19:09 |
| reto` | sn9: it's using the longhaul... but I think it's displaying an error message on boot about that... just didn't know what it is... | 19:09 |
| sn9 | reto`: HTH | 19:10 |
| sn9 | thrice`: lilo would actually work | 19:10 |
| reto` | sn9: ok, let me fetch that | 19:11 |
| reto` | sn9: the message is: APIC detected. Lonhaul ist currently broken in this configuration. | 19:18 |
| sn9 | what about e_powersaver? | 19:19 |
| reto` | sn9: I would have to try... how can I disable the longhaul? | 19:21 |
| sn9 | is it a module? | 19:23 |
| reto` | sn9: how can I find out? | 19:25 |
| sn9 | sudo rmmod longhaul | 19:25 |
| sn9 | if there's no error, try loading e_powersaver | 19:26 |
| reto` | module isn't found | 19:26 |
| reto` | it's in the kernel then? | 19:26 |
| sn9 | must be | 19:26 |
| reto` | I can load e_powersaver with the old kernel though | 19:27 |
| sn9 | unless it just didn't load | 19:27 |
| sn9 | modinfo longhaul e_powersaver | 19:27 |
| reto` | it din't find the longhaul but e_ | 19:28 |
| sn9 | does it load? | 19:29 |
| reto` | no I get an error | 19:29 |
| reto` | longhaul seems to be loaded somehow but not working correctly | 19:30 |
| reto` | wonder why it is loaded when it knows "it is broken" | 19:30 |
| sn9 | bug? | 19:31 |
| reto` | looks like longhaul went into the kernel with .31 | 19:32 |
| sn9 | karmic fail | 19:33 |
| reto` | can I disable it somehow? | 19:33 |
| sn9 | not without another kernel build | 19:33 |
| sn9 | OTOH, there is "noapic"... | 19:33 |
| reto` | yeah... hmm... might try that one | 19:34 |
| reto` | hmm... no use... it still loads longhaul | 19:43 |
| reto` | ah well... at least the machine is not hollering at full speed | 19:43 |
| infinity | reto`: echo "blacklist longhaul" > /etc/modprobe.d/die_longhaul_die && update-initramfs -u && reboot | 19:49 |
| infinity | reto`: That should stop it from being loaded. | 19:49 |
| sn9 | infinity: no; it's compiled in, not a module, rendering that useless | 19:50 |
| infinity | sn9: Oh. Didn't notice that. Seriously, though | 19:51 |
| infinity | ? | 19:51 |
| infinity | We shoudln't be compiling any of that in. :/ | 19:51 |
| sn9 | as i said: | 19:51 |
| sn9 | [Sun 12 Jul 2009 11:33:01 AM PDT] <sn9> karmic fail | 19:51 |
| reto` | ok... I've checked it out... with e_powersaver the scaling is working... I'm not 100% sure but I think longhaul has been in the kernel even before .31 (.28) but with 31.2 there seems to be a bug where longhaul is loaded even when it is detected to be broken | 21:01 |
| reto` | I'm writing a bug report on launchpad for this problem. Any suggestions? | 21:31 |
| sn9 | just that it should not be a module | 21:33 |
| sn9 | i mean | 21:33 |
| sn9 | just that it should not be not a module | 21:33 |
| reto` | :) ... ok | 21:34 |
| billybigrigger_ | Sarvatt, ping | 21:54 |
| billybigrigger_ | what other optimizations can i make when compiling kernel? | 21:54 |
| billybigrigger_ | besides not compiling for intel | 21:54 |
| billybigrigger_ | i seen there was some more v4l2 stuff changed today | 21:58 |
| billybigrigger_ | thought i'd give it another go | 21:59 |
| billybigrigger_ | anyone built today's kernel? | 22:30 |
| billybigrigger_ | i get this error | 22:30 |
| billybigrigger_ | http://pastebin.ca/1493055 | 22:30 |
| billybigrigger_ | when trying to compile from git | 22:30 |
| billybigrigger_ | ill just remove wireless modules from menuconfig then :P | 22:45 |
| billybigrigger_ | thanks guys :P | 22:45 |
| billybigrigger_ | no need on a desktop anywho | 22:45 |
| billybigrigger_ | where can i find my make log from my last compile? do i have to specify one for it to be outputted? | 22:46 |
| hyperair | make-kpkg blah > log? | 22:46 |
| billybigrigger_ | CONCURRENCY_LEVEL=3 fakeroot make-kpkg --initrd -append-to-version=-billybigrigger kernel_image kernel_headers kernel_source > make.log | 22:48 |
| billybigrigger_ | does that look right? | 22:48 |
| Sarvatt_ | you probably want to call it billybigrigger2 because you already have a billybigrigger and use >whatever.log 2>&1 | 22:53 |
| Sarvatt_ | do it in screen and control+a d the tail -f whatever.log to watch it :) | 22:53 |
| billybigrigger_ | tailing it now | 22:54 |
| billybigrigger_ | and have it named billy-07.12 | 22:54 |
| billybigrigger_ | :P | 22:55 |
| billybigrigger_ | thanks though | 22:55 |
| * billybigrigger_ crosses fingers :P some more v4l2 stuff updated today | 22:55 | |
| billybigrigger_ | Sarvatt, you till convinced its udev and usb problems though? | 22:58 |
| billybigrigger_ | drivers/net/wireless/wl3501_cs.c:384: note: ‘tmp’ was declared here | 23:18 |
| billybigrigger_ | make[1]: *** [drivers] Error 2 | 23:19 |
| billybigrigger_ | :( still wireless driver errors, i didn't include any wirelss modules though | 23:19 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
