[00:16] Has anyone read "Understanding the Linux Kernel, 3e"? === rikai_ is now known as rikai [02:40] 144 minuts without response [02:40] *minutes [02:40] I have. [02:40] WEll, most of it. [02:41] useful? [02:41] Erm... that's a lie actually. Glancing that the index, I've probably read about half. [02:41] Very. [02:41] I highly recommend it. [02:41] *goes to amazon* [02:42] I'm in a "intro to data communications" class; the students all have to install Ubuntu to a live CD. I had to explain to the teacher why there's a "duck" (heron) on the desktop. OH, and she's running FC6 for a server o_o [03:06] heh [03:06] my uni LUG had an email asking for help with FC1 [05:39] /c [08:12] 2.6.27-1-generic is rebooting upon resume for me :( [08:13] i was running an upstream .27-rc4, 1941246dd98089dd637f44d3bd4f6cc1c61aa9e4, and it resumes fine [08:13] amd64, thinkpad x300 [08:13] any suggestions for how to collect logs would be welcome [11:06] hi, hardy-updates has virtualbox-ose-modules-2.6.24-20-generic, but not the associated kernel. is that only a glitch? [11:30] gaspa: VirtualBox maintainer error, i suppose. [11:30] I think that he'll fix it soon with .21 [11:31] abogani: i meant : there's no project to put in -updates linux-image 2.6.24-20... [11:31] if so, the solution is: just wait. [11:35] No .20 is supersede. We are just bumped to .22 into Hardy git archive... in any case wait seems the best option :-) [11:37] abogani: ok, but -updates repositories are broken. === thunderstruck is now known as gnomefreak === asac_ is now known as asac [14:16] could someone please triage bug #261590? i feel that it is rather critical, but don't want to triage my own bugs [14:24] siretart: you could try #ubuntu-laptop and their mailing list, but it's basically fallen apart [14:25] (there might be a thinkpad channel) [14:26] siretart: Thinkpads don't export fans via acpi [14:26] It's notabug [14:27] There's a thermal issue on Thinkpads that's been fixed upstream [14:27] alternatively, just spam keywords till mjg notices [14:30] mjg59: err, did this change from hardy? [14:31] siretart: No [14:31] mjg59: what is today's way of controlling fans on thinkpads? [14:31] thinkpad-acpi will do it if you ask to it really hard [14:32] so no way of controlling it somehow interactively? [14:33] Only using thinkpad-acpi [14:33] I thought of echo'ing stuff to /proc [14:34] mjg59: with 'asking it really hard' do you mean the option 'fan_control=1'? [14:41] siretart: Yeah [14:42] siretart: the fun happens in /proc/acpi/ibm [14:42] siretart: https://launchpad.net/tp-fan is handy too [14:43] Bear in mind that the X60 handles high temperatures by slowing down the CPU in firmware [14:43] That's what was broken at various points [14:44] mjg59: do you know what this message is about: [14:44] "thinkpad_acpi: setting the hotkey mask to 0x00ffffff is likely not the best way to go about it" [14:45] Yes, with recent thinkpad_acpis userspace shouldn't be setting hotkey_mask [14:48] mjg59: where could that be getting set? [14:50] /etc/modprobe.d somewhere? [14:59] mjg59: thanks for the information! I'll update the bug [14:59] <_MMA_> Does this deserve a bug report? http://mma.users.ubuntustudio.org/vbox.png It's using the latest on-disk image -generic kernel. [15:00] shenki: that would be bug LP #256887 [15:02] siretart: thanks,i just found it myself too [15:02] stgraber: heh, we added the same comment. you bet me by 49 seconds. [15:02] :) [15:05] shenki: it's hotkey-setup, I'll fix it [15:05] Remember that just removing the file from the package won't fix it, since it's a conffile [15:05] shenki: TBH, I'm a bit hesitant to add a /etc/modprobe.d/thinkpad_acpi.modprobe. what happens if 2 files define options for the same module? [15:06] siretart: mmm, hotkey setup would run after the modprobe, so it would over write modprobe's settings [15:06] mjg59: it's echo'ed during init, so just removing that part should do, right? [15:07] Yeah [15:07] Oh, of course, it's part of the init script [15:07] yep [15:08] the docs suggest setting it to /sys/devices/platform/thinkpad_acpi/hotkey_recommended_mask [15:08] cat /sys/devices/platform/thinkpad_acpi/hotkey_recommended_mask > /sys/devices/platform/thinkpad_acpi/hotkey_mask [15:09] indeed. [15:09] but that's something the module could it itself. if it doesn't do that already [15:10] later on, it talks about not messing with defaults for the backlight control when one has acpi video [15:10] "For Lenovo models *with* ACPI backlight control" [15:12] siretart: have you attempted resuming from suspend with .27? [15:12] shenki: suspend works on my X60s, I haven't tried hibernate yet, though. [15:12] resume btw as well [15:13] I wonder if that breaks my screen not resuming from powersave-mode [15:13] with .24 it's ok [15:13] my x300 was rebooting (instead of resuming) with .27-1-generic, but resumes fine with upstream ~rc4 [15:14] shenki: Known bug in ftrace [15:14] mjg59: thanks [15:17] ah, and there's the fix. ftrace: disable tracing for hibernation. cool [15:21] There's another one for suspend to ram [15:22] noticed that. figured i was talking to myself enough. don't need to correct myself too [15:39] siretart: I'v tried commenting out that stuff from the initscript, and now hotkey_mask is the same as hotkey_recommended_mask [15:40] ok, so my laptop screen really doesn't want to resume frome the powersaving mode [15:40] intel [15:40] need to reboot to get it back [15:41] tjaalton: so let's claim that setting the hotkey mask was an quirk that is not necessary anymore with current thinkpad_acpi, and we should remove it completely from hotkey-setup [15:41] siretart: indeed [15:43] tjaalton: does your ThinkVantage button do something useful? It doesn't for me since (and including) hardy [15:43] nope === BenC changed the topic of #ubuntu-kernel to: Ubuntu kernel development discussion ONLY | Kernel Wiki: https://wiki.ubuntu.com/KernelTeam | Latest kernel upload: 2.6.27-2.3 (based on 2.6.27-rc4-git) | Latest news: Moved to 2.6.27 kernel for Intrepid/8.10. | Next meeting: Aug 27, 16:00 UTC | Kernel Team machine: http://kernel.ubuntu.com === BenC changed the topic of #ubuntu-kernel to: Ubuntu kernel development discussion ONLY | Kernel Wiki: https://wiki.ubuntu.com/KernelTeam | Latest kernel upload: 2.6.27-2.3 (based on 2.6.27-rc4-git) | Latest news: Moved to 2.6.27 kernel for Intrepid/8.10. | Next meeting: Sept 2, 16:00 UTC | Kernel Team machine: http://kernel.ubuntu.com [18:08] 2.6.27 seems to have cleared up the intermittent usplash misrenderings for me [18:17] cjwatson: ogasawara said the bug showed others claiming the same [18:26] wl driver seems to work, though curiously b43 is still loaded [18:34] cjwatson: does jockey do anything to divert conflicting open/closed drivers? [18:39] no idea [18:39] wl is still closed from its point of view anyway [18:39] it's in linux-restricted-modules and listed in the jockey UI [18:47] cjwatson: I notice alsa-utils 1.0.17 was uploaded [18:47] that was fast [18:48] yeah, Luke squeezed it in under the freeze last night [18:48] sound still works for me at any rate [18:49] freeze? [19:22] .27 - I now have both ath5k and ath_pci loaded at the same time, which I think is bad - should I file a bug in lp? [19:26] CarlFK: Disable ath_pci by either uninstalling linux-restricted-modules or using "Hardware Drivers" control panel [19:28] * BenC => reboot === mkrufky is now known as mkrufky-away === mkrufky-away is now known as mkrufky === rikai_ is now known as rikai [21:57] is there some way to redirect kernel boot output to usb? (laptop has no serial ports) [22:03] booting .27 - it pauses at various places, like right now at "[3.11...] SCSI subsystem initialized" - been there for over 1 min. If I hit any key (spacebar) it will continue, then pause somewhere else. [22:03] this seems very odd. [22:05] didn't touch it - crash - [174...] stack dump. [22:25] CarlFK, Just a wild guess. Might try clocksource=jiffies (won't probably help with the crash, but there had been cases where it helped with pauses) [22:35] smb_tp: http://dev.personnelware.com/carl/a/pics/P1010005.JPG pic of screen with "BUG: soft lockup - CPU#1 stuck for 160s! [swapper:0]" [22:37] smb_tp: guessing you are the smb that gave me http://people.ubuntu.com/~smb/bug254668 (same box, .27 kernel) [22:37] CarlFK, Was that with the clocksource option or just the crash you mentrioned before [22:37] Yes, I am [22:37] hi [22:37] Hi [22:37] no clocksource [22:37] That one looks like somewhere in the usb code [22:38] here is it before the dump http://dev.personnelware.com/carl/a/pics/P1010004.JPG [22:39] http://dev.personnelware.com/carl/a/pics/P1010003.JPG same thing, better focus [22:41] Yeah somewhere setting up the usb controller. The crash place is a bit strange in delay(). Hm... [22:46] http://dev.personnelware.com/carl/a/pics/P1010006.JPG "CPU1 stuck for 98s! [udevd:2033]" [22:47] let me know if you want me to open a lp tickit and post these [22:48] CarlFK, Yes, I just have a quick look into the code. Could you probably try the clocksource option meanwhile [22:48] ? [22:48] sure [22:56] 18:48 yeah, Luke squeezed it in under the freeze last night [22:56] 18:49 freeze? [22:56] oops, Ben left already [22:56] was going to remind him of this little thing called feature freeze ;-) [22:58] cjwatson, Probably he just forgot it was so close. :) [23:01] smb_tp: clocksource=jiffies - now it is stuck on "setting the system clock" [23:02] space/enter/type - shows up on console, doesn't un-pause it. power button, nothing. [23:04] hit the "special dvd" button (hp laptop thing) - console shows "[11.8...] atkbd.c: unknown key released..." so it isn't froze just stuck [23:06] CarlFK, Very strange. Something somewhere in the timers region. I'd say file a bug for 2.6.27 and probably include dmidecode data so its clear what system this is. At the moment, does SYSRQ-t print something? [23:12] CarlFK, Since the oops happened in delay_tsc you might also try "notsc" to see whether that changes things. [23:14] k - I rebooted, and this time is slightly different: http://dev.personnelware.com/carl/a/pics/P1010009.JPG [23:14] hitting sysrq puts ^\ on the screen [23:15] what is SYSRQ-t ? (trying to cord it just does ^\t ) [23:16] CarlFK, Alt+Print+t [23:19] oh great - it fixed itself [23:20] CarlFK, Just by waiting or after another boot [23:21] another boot - with clocksource=jiffies - which would hang till I power cycled it [23:23] CarlFK, this really feels like timing/timer stuff. But I can't place it. [23:23] hmm - sudo apt-get upgrade "The following packages have been kept back: hal libsmbios-bin ubuntu-minimal" [23:24] should I install those? [23:25] CarlFK, I don't think they have some impact that early but it might be better to have a clean installation [23:26] CarlFK, well probably hal might [23:28] CarlFK, In any case, you should probably try the "notsc" option (instead of clocksource) and create a bug so this documented [23:30] CarlFK, I have to leave now, sorry [23:30] no prob [23:31] CarlFK, I uploaded another kernel to debug the other problem (it will still hang but prints more). But I did not have time to run it here, so you might want to wait or try it [23:32] smb_tp: bug report against .27 kernel ? [23:32] The crash yes [23:32] ill give it a shot - not like it will catch my box on fire :) [23:32] CarlFK, No rather more hanging :) [23:33] huh - hitting Alt causes a register dump [23:33] you better get going :) [23:34] CarlFK, Yeah, I better do... :) [23:34] * smb_tp runs