[03:30] mcgrof: ping [03:31] maybe on holiday given his client's idle time [03:31] heh [03:33] crimsun: same problem with rc1 [03:37] 3945? [03:38] yea [03:40] according to bc45a67079c916a9bd0a95b0b879cc0f259bac6e in wireless-testing.git, powersaving on it is broken [03:40] oh goodie [03:41] you could try looking in http://bugzilla.intellinuxwireless.org/ [03:42] how did you find that commit [03:42] (that one is http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2125) [03:42] bugzilla.intellinuxwireless.org bug 2125 in firmware error "Microcode SW error: SYSASSERT (#5) Line 947" [Major,Verified: fixed] [03:42] I read the commit logs for wireless-testing.git [03:45] shtylman: pong [03:46] mcgrof: we are investigating a wireless bug [03:46] intel 3945 does not connect to anything [03:47] crimsun: disabling power mode does not seem to work [03:47] I don't think I have that specific firmware problem... [03:47] shtylman: wireless isn't my area, hence my referral to mcgrof [03:47] gotcha :) [03:50] mcgrof: trying to see how best to report this wireless problem [03:50] or what to do about it in general [03:50] as currently lucid with intel 3945 is useless on wifi [03:52] I might have to try a wireless testing kernel [03:52] any idea on how often those get merged? [03:55] Um... my intel 3945 chip is happily connecting to my AP. [03:56] shtylman: http://wireless.kernel.org/en/users/Documentation/Reporting_bugs [03:56] RAOF: in lucid? [03:56] Yes, in Lucid. [03:57] RAOF: um... how... :) [03:57] By... not doing anything special? [03:57] hmm [03:57] I mean, it hasn't broken for a while :) [03:58] thats interesting... cause those bug reports would suggest it has [03:58] and I didn't do anything to mine [03:58] and it didn't work [03:58] what kernel version? [04:02] 2.6.32-9 [04:03] wtf [04:03] What firmware version is your driver loading? [04:04] I remember fiddling with that at one point in the past. [04:04] how can I tell? [04:04] dmesg | grep iwl will include a "firmware: requesting ..." line. [04:04] Mine's loading "iwlwifi-3945-2.ucode" [04:05] same [04:07] And your problem is that you can't connect to _anything_? I'm connecting to a wpa2 AP now, but I have recently connected to an open AP as well. [04:07] yep [04:07] nothin [04:08] I have a wpa2 ap .. [04:08] can't connect to it [04:08] and I couldn't connect to an open one recently either [04:08] wpa supplicant just spin [04:08] *spins [04:08] Sucks to be you apparently; mine hasn't had any trouble in Lucid at all. [04:09] wait, wpa_supplicant or network-manager*? [04:09] I kill network manager [04:09] and try supplicant manually [04:09] but neither one connectes [04:09] *connects [04:09] ok, and if you use wireless-tools directly with an open AP? [04:09] havn't tried that [04:10] oh wait [04:10] I think it worked with power off [04:11] my problem might have been that network manager wasn't fully dead... so with NM dead, I could use wpa supplicant manually [04:11] with power management off [04:13] so now... manual works... just no network manager... [04:13] at this point I guess it ceases to be a kernel problem :) [04:14] well, if it's ok with pm off, then you may be triggering that bug I mentioned earlier [04:15] right [04:15] but I never saw the fw error before turning the power management off [04:15] oh well [04:16] tragic that power management doesn't work for it [04:16] at least wireless works now [04:16] battery will have to suffer [04:19] suffer is probably a bit drastic [04:20] hehe [04:25] thanks for the help peoples === jk-_ is now known as jk- [07:21] Hi! I'm using Karmic and I've run into the "low memory corruption" bug (324894), but without having to suspend/resume. Does anyone know something about the background of this bug, or any fix for it? I've found this thread on LKML: http://osdir.com/ml/linux-kernel/2009-07/msg02383.html Based on that it seems to be a widespread problem. [12:26] * lamont finds it sad that the system doesn't want to let him strace pid 1 [12:27] why wouldn't it? [12:27] it seems to work for me [12:28] EPERM it says [12:29] I seem to have found an i915 regression in 2.6.32 (bug 492392). What is the proper upstream bug reporting place? [12:29] Malone bug 492392 in linux "[lucid, intel] After suspend, flickering screen and then blank screen." [Medium,Triaged] https://launchpad.net/bugs/492392 === fabbione is now known as fabbione_vac [13:35] maxb, normally one creates a bug in the kernel bugzilla [13:35] its likley helpful to report it in email to jessie barnes and dave aerlie [13:35] maxb, did you manage to bisect for it? [13:38] ahh in 2.6.32-rc1 [13:38] maxb have you tested the latest 2.6.32.2 kernel at all from the mainline builds? [13:54] can I poke apport into reporting a kernel oops that happened when apport wasn't running, but is logged? [13:55] I was rsyncing my laptop to a USB disk and some kind of minor kernel explosion killed X [14:18] apw: I have not yet, though I seem to be having some success with i915 from 2.6.31 backported against a 2.6.32 kernel, so localizing it to that module [14:20] coo [14:20] cool [14:22] I don't think I have an ironlake, but it's possible "drm/i915: Fix LVDS stability issue on Ironlake" might have some bearing on the problem even so [14:24] its a big patch changing generic i915 code, so could affect you [14:26] * maxb downloads [14:33] maxb, you able to make a kernel with that reverted? [14:33] Ng it is meant to do that, pick them up later i believe [14:33] oh, hrm [15:02] apw: I dunno if it would have noticed itself, but enabling apport and kerneloops and then poking kerneloops to read /var/log/syslog at least did enough to get me a .crash I could turn into bug #499858 :) [15:02] Malone bug 499858 in linux "kernel oops and X crash while copying data to USB disk" [Undecided,New] https://launchpad.net/bugs/499858 [15:07] Ng i don't see any oops can you look in the files in /var/log and see if one was captured ... [15:08] huh, I have the dmesg.txt from sshing in while it was in the crashed state, I'll attach that [15:09] apw: done [15:09] geat [15:11] are there any issues that could come out of disabling apic? [15:12] disabling apic, means we use differnet routing for interrupts, in theory not a disaster [15:12] you shouldn't have to tho [15:15] i'm considering enabling it and seeing how it goes for my machine [15:15] i mean disabling apic [15:16] i got strange cpu-level (presumably) hangs that whacked up a fair number of drivers on my system [15:16] tg3, and iwlagn in particular [15:16] then something in acpi started complaining in dmesg [15:16] and psmouse lost synchronization [15:16] hello! is there somewhere around kernel-ppa for karmic? I am getting 404 error with deb http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu karmic main [15:16] deb-src http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu karmic main [15:16] deb http://kernel.ubuntu.com/~kernel-ppa/mainline/ubuntu karmic main [15:16] deb-src http://kernel.ubuntu.com/~kernel-ppa/mainline/ppa/ubuntu karmic main [15:17] apw: http://pastebin.com/f322a1676 [15:18] hyperair, what were the few lines before those [15:18] Michalxo, the mainline series have to be installed directly from kernel.ubuntu.com/~kernel-ppa [15:20] aha, so do I have a "bug" somewhere in my sources lines? [15:20] I will remove "mainline" line [15:20] Michalxo, right, there is no 'mainline' PPA [15:20] but this "deb http://ppa.launchpad.net/kernel-ppa/ppa/ubuntu karmic main" should work, right? [15:21] Michalxo, it ought to, but there are no packages there [15:21] aha, so how/from where can I get "the latest" kernels? [15:21] Michalxo, https://wiki.ubuntu.com/KernelTeam/MainlineBuilds [15:23] oh, so there is only manual isntalling.. no apt-like thing [15:24] Michalxo, that is correct. we did it that way with malice aforethought. [15:26] aha.. [15:27] thank you very much :-) [15:27] enjoy and have a happy "holidays" [15:33] apw: http://pastebin.com/f33d3433a [16:15] RAOF, about? [16:15] hyperair, thats most of it, can you get another 3 lines or so in case it said anything just before [16:16] RAOF, your nouveau-scratchpad branch thingy seems to be down [16:16] hyperair, what is there looks like a bug to me, and not something apic on or off would necessarily fix [16:17] hyperair, get it in a bug too [16:49] apw: http://pastebin.com/f3a0db8f4 this would be everything [16:52] ..weird, why does noapic turn off one of my CPUs?! [16:53] hyperair, I think its because its the APIC subsystem that is responsible for enabling subsequent CPUs [16:53] ah hell. [16:53] is there another way to enable my other CPUs? [16:54] not that i know of [16:54] they would be on if it had detected them [16:54] * hyperair sighs [16:54] this sucks [16:55] that trace shows your intel wireless breaking cause it failed to load the firmware, due to a memory allocation failure [16:55] hmm is that so? [16:55] but it was working fine before that [16:55] at the top there yes [16:55] there was some SW microcode error [16:55] right the card looks to have 'crashed' and the restarted failed as we couldn't get the firmware [16:56] indeed [16:56] in fact, everything was crashing one by one [16:56] including psmouse [16:56] and tg3 [16:56] tg3 has a very interesting side effect [16:56] it will be able to receive, but not send [16:58] don't see anything about the mouse there [17:04] hmm weird [17:12] apw: okay, this time around, with noapic and nolapic, i still get the hangs, but i get these two lines.. [17:13] [ 5245.062585] iwlagn 0000:04:00.0: index 0 already used in uCode key table. [17:13] [ 5257.876381] iwlagn 0000:04:00.0: Hardware error detected. Restarting. [17:13] thats definatly the card being unhappy [17:13] indeed [17:13] though normally they recover and continue [17:13] it did recover and continue [17:13] i'm on a wireless connection now [17:13] that is just luck based on how long after boot [17:14] and i didn't disconnect at all [17:14] luck? [17:14] as the previous one was an allocation order 4 which failed, which is a pretty big allocation [17:14] ah i see [17:14] yeah ... the longer you are up the less likely an order-4 allocation is possible [17:14] i see. [17:14] its not sensible for the driver to rely on it [17:15] then again the microcode should not be going pop ever eit [17:15] either [17:15] heh yeah [17:15] well actually i think it's all related to some other issue.. [17:15] possibly timing or something? [17:15] when playing 3d games i get these weird spurious hangs that get increasingly frequent [17:16] hrm [17:16] when they get severe enough, everything starts going haywire [17:16] mouse, tg3, iwlagn [17:16] those are the most prominent ones [17:16] in the case of tg3, a suspend/resume fixes it [17:16] rmmod and modprobe again doesn't work [17:17] * hyperair brb. need to reclaim my lost CPU core [17:25] I have a few old linux-image and linux-headers packages, and even when I mark them as auto, aptitude won't remove them, because they have "Provides: linux-image/headers". is that how it's supposed to be? [18:09] apw: With what reverted? [18:09] the patch you were having trouble with [18:09] I'm running 2.6.32.2 now, we shall see if it breaks [18:09] i think .2 has that patch yes [18:10] Oh. Well, my trouble starts after a fairly immense merge from drm-intel-next before 2.6.32-rc1 [18:10] Reverting the entirety of i915 back to before that seems to help [18:11] But, that's such a volume of changes, that it doesn't say a huge amount [18:11] I'm getting some flickers with 2.6.32.2 [18:11] Different kind of visual artifacts than I was seeing with 2.6.32.1, but still a problem [18:12] I shall continue to run with it to see if the complete blank screen failure case happens. [18:12] Unfortunately I have no way to reproduce it on demand, I just have to use the computer for a few hours [18:27] [ 129.431995] [drm:radeon_cp_reset] *ERROR* radeon_cp_reset called without lock held, held -2147483648 owner ebddd840 ee217600 [18:27] meh [21:32] apw: Hi. Are you still here? [21:33] apw: It seems I've got dns problems - http://121.44.54.198/kernel-lucid-nouveau.git/ will get you the branch. [21:41] apw: dns fixed; http://cooperteam.net/kernel-lucid-nouveau.git should work now.