[04:24] <mxpxpod> BenC: ping
[10:52] <marcreichelt> hi there
[10:57] <marcreichelt> my ubuntu kernel seems to have problems with acpi - nearly every time when my laptop is under heavy use (e.g. reading big files from my external usb disk or burning a cd) it just hangs up with a kernel panic
[10:59] <marcreichelt> a shot of the screen can be found here: http://www.marcreichelt.de/misc/kernelpanic.jpg
[10:59] <marcreichelt> I tried to deactivate all modules containing the word "acpi", but this didn't fix the problem
[11:00] <marcreichelt> I also tried to deactivate acpi on boot, with "acpi=off" as kernel parameter
[11:01] <marcreichelt> the result was that the system frozed after approx. 30 seconds after boot time
[11:01] <marcreichelt> do you have any ideas? is this a bug in the ubuntu kernel? what can I do?
[11:40] <mjg59> marc_learning: "acpi_processor_idle" just means that the system wasn't doing anything at the time
[11:40] <mjg59> But you can check by unloading the "processor" module
[01:29] <marcreichelt> back again
[01:34] <marcreichelt> hmm
[01:35] <marcreichelt> processor depends on speedstep_centrino, and I can't unload that one, because it is in use
[01:41] <marcreichelt> but it doesn't say by which module it is used
[01:42] <Mithrandir> it's probably used by powernowd or something similar.
[01:43] <infinity> No, you can't unload centrino.  mjg59 and I ran into that a week or two ago.  It's thpethul.
[01:44] <infinity> The only way I could "unload" it was to not load it in the first place.
[01:45] <marcreichelt> k
[01:45] <marcreichelt> how can I deactivate it on boot?
[01:45] <marcreichelt> so that "processor" and speedstep_centrino aren't started at boot time?
[01:45] <infinity> Blacklist them in modprobe.
[01:45] <marcreichelt> Mithrandir, I killed powernowd
[01:46] <infinity> Alternately, move the files somewhere else so you're sure nothing can load them. :)
[01:48] <marcreichelt> moment...
[01:48] <marcreichelt> brb
[01:57] <marcreichelt> ok, I created the file /etc/hotplug/blacklist.d/asuslaptop-problems and wrote the line "processor" into it
[01:57] <marcreichelt> is this correct?
[01:59] <marcreichelt> I think it should be
[01:59] <marcreichelt> well, I'm going to reboot now
[02:00] <marcreichelt> see you in a few minutes
[02:12] <marcreichelt> hello again
[02:12] <marcreichelt> my module "processor" is still running
[02:12] <marcreichelt> even a line in /etc/hotplug/blacklist did not work
[02:13] <marcreichelt> you still mean that it would be good to rename the module?
[02:14] <marcreichelt> k, I renamed the processor module
[02:14] <marcreichelt> brb
[02:19] <marcreichelt> hi again
[02:20] <marcreichelt> he still loads the module "processor" - but I renamed processor.ko to processor.ko.old... :-/
[02:20] <marcreichelt> maybe I have to create a new initrd?
[02:21] <marcreichelt> or kernel?
[02:22] <marcreichelt> or is there another way to fix my problem?
[02:23] <marcreichelt> I would hack myself into the kernel to solve this problem if I would have the time to
[02:24] <marcreichelt> the problem seems to be in processor.c
[02:36] <marcreichelt> :-(
[02:47] <marcreichelt> brr
[02:48] <marcreichelt> again a kernel panic
[02:49] <makx> what ubuntu version are you using
[02:49] <makx> blacklists should be under /etc/modprobe.d
[02:49] <marcreichelt> Breezy
[02:50] <marcreichelt> Kernel: 2.6.12-10-386
[02:50] <makx> yes put it there
[02:50] <marcreichelt> moment
[02:51] <marcreichelt> makx, how should the blacklist look like?
[02:51] <marcreichelt> just a file containing the line "processor"?
[02:55] <marcreichelt> ?
[02:56] <makx> blacklist evbug
[02:56] <makx> for example
[02:56] <marcreichelt> k
[02:56] <marcreichelt> for me it is blacklist processor, then
[02:57] <marcreichelt> ok, brb
[03:02] <marcreichelt> no, it didn't work
[03:02] <marcreichelt> he is still loading the module "processor"
[03:03] <marcreichelt> is there a kernel version which does not contain this bug?
[03:32] <zul> heylo
[03:50] <BenC> marcreichelt: dpkg-reconfigure linux-image-`uname -r`
[03:50] <BenC> that will generate a new initramfs
[03:52] <mxpxpod> BenC: have you had any problems lately with your ae?
[03:52] <BenC> none
[03:52] <BenC> other than the weird eth0_clashed thing, but that's not particular to ae
[03:53] <mxpxpod> BenC: check this out... someone in my apartment installed a wireless router with authentication on the same channel as mine and my ae gets a ton of DEAUTH messages from it
[03:53] <mxpxpod> this causes klogd to go crazy and take up like 10% cpu
[03:53] <BenC> hmm, sounds like softmac actually
[03:53] <mxpxpod> if I switch to another channel, I don't get it
[03:54] <mxpxpod> oh, really?
[03:54] <BenC> maybe I need to turn down the verbosity
[03:54] <BenC> the messages I think come from softmac
[03:54] <mxpxpod> yeah, dmesg is full of DEAUTH's
[03:54] <mxpxpod> it's quite frustrating... so I switched channels
[03:55] <BenC> actually, it's right from ieee80211.ko
[06:09] <lamont> BenC: can you easily tell me when the last update of the hppa patch in 2.6.15 happened?  (was -13 built with the hppa patchset, or without, to be specific...)
[06:13] <BenC> I've never applied an hppa patchset per se
[06:13] <BenC> I merely got what was there to work
[06:14] <BenC> lamont: so it's 99% stock 2.6.15 code for hppa, with just one backport for a compat syscall I can't recall off hand
[06:19] <lamont> BenC: please consider merging the following patches (minimum) from ftp://ftp.parisc-linux.org/patches
[06:19] <lamont> diff-2.6.12-rc3-tulip_21142_phyinit, diff-2.6.3-tulip, diff-2.6.6-tulip_stop_rxtx, diff-tulip_phy_reset-03
[06:19] <lamont> all 4 are in akpm's tree
[06:22] <lamont> those patches make tulip work on modern hardware...  bug exists on hppa as well as some i386 platforms
[07:16] <marcreichelt> BenC: just leave the processor blacklisted and generate a new initramfs?
[07:21] <fs> what's the current status of the megaraid newgen/legacy patch? 
[07:33] <fs> never mind, found it: it's merged upstream.
[07:35] <lamont> BenC: I'll start working through the hppa patchset to see what makes sense for us to include
[07:36] <lamont> as in, what's low-risk and/or hppa specific
[08:44] <marcreichelt> ok, I give up
[08:44] <marcreichelt> the module 'processor' is still running
[08:45] <marcreichelt> maybe I'm just too stupid for this :-/
[09:03] <BenC> marcreichelt: rm -f the module
[09:07] <marcreichelt> hmm
[09:07] <marcreichelt> or better mv module module.bak?
[09:09] <marcreichelt> ok
[09:09] <marcreichelt> root@tux:/lib/modules/2.6.12-10-386/kernel/drivers/acpi # mv processor.ko processor.ko.bak
[09:09] <marcreichelt> now I do dpkg-reconfigure linux-image-`uname -r`
[09:10] <marcreichelt> hmm
[09:10] <marcreichelt> root@tux:/lib/modules/2.6.12-10-386/kernel/drivers/acpi # dpkg-reconfigure linux-image-`uname -r`
[09:10] <marcreichelt> Not touching initrd symlinks since we are being reinstalled (2.6.12-10.26)
[09:10] <marcreichelt> Not updating image symbolic links since we are being updated (2.6.12-10.26)
[09:10] <marcreichelt> it seems that he didn't noticed anything
[10:23] <marcreichelt> I have to go
[10:23] <marcreichelt> see you