[08:56] <Administrator> 2.6.17-10.33 kernel panics
[08:57] <Administrator> at least the current one of that version in edgy, is that a known bug?
[09:57] <infinity> Administrator: You might have to be a bit more specific.
[10:01] <Administrator> infinity, ill report more detailed error in a min or 5/10.
[10:21] <Administrator> infinity, kernel panic not syncing: attempted to kill init kernel_thread helper| above the ther error messg |  process swapper and kernel trehad helper i running 2.6.17-10 generic
[10:27] <Administrator> I can debug further, i do not have a serial console or whatsoever..
[10:28] <infinity> If you can take a picture of the panic, perhaps?
[10:28] <infinity> I have to run, but Ill be back in 20/30.
[11:54] <Administrator> infinity, you there?
[11:54] <Administrator> i can send you ...
[11:54] <Administrator> through mail or dcc ...
[11:58] <infinity> adconrad@ubuntu.com
[11:58] <infinity> Though, once we determine that it's a bug, and in what package, I'd prefer if you file a bug and attach the image. :)
[12:05] <Administrator> i am a bit bussy today, i'll mail you the images asap and at the end of the day i'll file the bug if things calm down a bit.
[03:08] <Trae> what package is mkinitrd in?
[03:09] <makx> you update-initramfs from initramfs-tools
[03:09] <infinity> mkinitrd is well and truley obsolete.
[03:10] <makx> yes bug on ftp.debian.org is placed, it will soon be gone
[03:10] <Trae> A friend of mine was trying to help me get a fresh kernel compiled
[03:10] <Trae> he doesn't run Ubuntu though, he runs Gentoo
[03:12] <zul_> different ways of working but we dont use mkinitrd as infinity said
[03:13] <Trae> I'm in /usr/src/linux (it's a symlink to /usr/src/linux-2.6.18.1   I've already done make && make install && make modules_install  and edited grub.  
[03:14] <Trae> I got a kernel panic though when It tried to boot
[03:14] <Trae> Cannot open root device hda1
[03:14] <Trae> it says put in a valid root= thingy
[03:14] <Trae> heh
[03:14] <zul_> Trae: check the wiki there is documentation on how to compile a kernel
[03:14] <makx> why do you want a hand build kernel?
[03:15] <Trae> makx: you haven't seen me whinging in here aboutthe laptop overheat bug? :)
[03:15] <Trae> makx: hehe
[03:16] <Trae> I'm a graphic designer, dealing with this kernel stuff is way over my head sadly.
[03:16] <makx> hehe is not a reason :-P
[03:16] <Trae> :P
[03:16] <Trae> so here is the odd thing
[03:16] <makx> yes so use the provided linux images
[03:16] <Trae> oh let me find the bug
[03:16] <Trae> https://launchpad.net/bugs/22336
[03:17] <Trae> check this out... I was able to compile the kernel for like an hour without the laptop overheating or rebooting
[03:17] <Trae> yet 4 mins of playing a Youtube video, and the thing shuts down on me!?
[03:18] <makx> hmm mjg59 is not around he is the local acpi guru
[03:18] <makx> Trae what release are you using?
[03:19] <Trae> Dapper Drake
[03:19] <makx> did you try the Edgy release candidate - the live cd?
[03:20] <Trae> I tried beta
[03:20] <Trae> hmm
[03:20] <makx> and?
[03:20] <Trae> no haven't tried the RC
[03:20] <Trae> the Beta had the sameproblem.
[03:20] <Trae> I will try the RC though
[03:20] <Trae> and of course use the Final
[03:21] <makx> beurk
[03:21] <makx> hmm if i were you i'll wait for the acpi master to show up
[03:22] <infinity> Trae: Laptop overheating bug?
[03:22] <Trae> infinity: https://launchpad.net/bugs/22336
[03:27] <infinity> Okay, this is sad to admit, but I suffer from this bug, and just "cope".
[03:27] <infinity> I'll get my laptop in front of BenC and mjg59 in Mountain View, and see if the three of us can nail it down.
[03:28] <infinity> It actually seems to be bad hardware design mixed with slow response times on throttling.
[03:28] <thom> mjg59's not gonna be there afaik :(
[03:28] <Trae> like clockwork
[03:29] <infinity> (The bad hardware design in my case being that the video RAM is WAY too close to the CPU, and when I overheat in a split second, the video corrupts and the machine crashes)
[03:29] <Trae> 4mins into the youtube video and kerpow
[03:29] <Trae> heh
[03:29] <infinity> thom: Oh, well, BenC and I can deal then.
[03:29] <Trae> infinity: really you have this too? 
[03:29] <Trae> infinity: you a kernel developer?
[03:29] <Trae> infinity: w00p!
[03:30] <Trae> wish I was going to be in Mtn View too.
[03:30] <Trae> infinity: whoa,that's quite insightful
[03:30] <Trae> re: ram being too close
[03:31] <Trae> infinity: I will volunteer to help you do whatever you need on this laptop.  Even give youguys ssh access if need be
[03:31] <zul_> infinity: ill be there as well
[03:31] <Trae> infinity: traemccombs {at} gmail d o t com
[03:31] <Trae> :)
[03:31] <Trae> I'll try and lurk here in case you might need me.
[03:35] <Trae> Keybuk: morning!
[03:35] <Trae> :)
[03:35] <Keybuk> heyhey
[03:35] <Trae> Keybuk: infinity says he suffers from the same bug!
[03:35] <Keybuk> oh, that's cool
[03:35] <Trae> Keybuk: and is going to work with BenC in Mtn. View
[03:36] <Trae> Keybuk: sucky for him, good for us hopefully.
[03:36] <zul> argh...who the hell puts their log files in html
[03:37] <infinity> MSN messenger?
[03:37] <zul> no exceed on demand
[03:40] <infinity> Trae: I can't say if it's the *same* bug, as the variables (CPU type, drivers involved as a result, etc) are pretty different, but the end effect is certainly the same (rapid rise in temperature, throttling and/or cooling not kicking in fast enough to cope)
[03:41] <infinity> Trae: So, it could be an underlying timing bug that lives a bit higher up the stack than the individal CPU power drivers.
[03:41] <infinity> Trae: Anyhow, I don't have any time (or the knowlege of that part of the codebase) to look at it in the next week, so there's no hope of it making edgy, but maybe we'll work up some patches for you to test a few weeks after release.
[04:01] <Trae> infinity: no worries, as always, triple my money back if I'm not 100% satisfied :)
[04:02] <tuxmaniac> Hi All.
[04:03] <tuxmaniac> My Card reader on my Compaq laptop not working since I updated to the 2.6.17 series
[04:03] <tuxmaniac> It worked in the Dapper kernel versions
[05:11] <Keybuk> BenC: around?
[06:20] <BenC> Keybuk: yeah
[06:26] <Keybuk> BenC: have been trying to trace the reboot() syscall to find out whether the kernel actually calls sync() before powering down or not
[06:26] <Keybuk> but not having much luck
[06:27] <BenC> Keybuk: I don't know for sure, but I do know that I see scsi disk sync messages on reboot on power down
[06:28] <BenC> filesystems should sync on umount (or remount ro)
[06:28] <mjg59> infinity: On your Thinkpad?
[06:28] <BenC> and I think the block layer handles hw level syncs for any devices with dirty cache
[06:28] <infinity> mjg59: Yeah.
[06:28] <infinity> mjg59: I can blow it up pretty spectacularly.
[06:29] <infinity> mjg59: It *seems* to be heat/throttle related, but given that the same sort of things that generate heat also do other evil things to one's system, I can't be completely sure.
[06:31] <mjg59> infinity: Entirely not Linux's problem
[06:31] <infinity> Well, in the words of the bug submitter, "I can't reproduce this in Windows".
[06:31] <mjg59> As far as I can tell, at least. Is it really getting to more than 91 degrees?
[06:32] <mjg59> When I say "not Linux's problem", I mean "The hardware may be making assumptions that are not guaranteed"
[06:32] <mjg59> Thinkpads don't have any OS-level fan control
[06:32] <mjg59> And they're not supposed to throttle until 90 degrees CPU temp or so
[06:32] <infinity> Yeah, I've been curious about that.
[06:32] <infinity> Fan usage patterns seem drastically different in Win32 and Linux, but I'd swear the kernel does no fan control.
[06:33] <infinity> So I have no idea WTF is up with that.
[06:33] <mjg59> If you have no /proc/acpi/fan/*, the kernel does no fan control
[06:33] <infinity> I do not. :)
[06:33] <mjg59> For whatever reason, Windows probably drives the hardware in such a way that it generates less heat
[06:34] <infinity> The only time the fan on this EVER spins into vacuum mode is on POST.  Which is worrying.
[06:34] <infinity> (In Windows, it'll vacuum every once in a while on its own)
[06:34] <mjg59> The processor is probably never getting hot enough to trigger it
[06:49] <Keybuk> BenC: hmm, I'm trying to determine whether the "sync(); sleep(2)" is necessary in /sbin/reboot
[06:49] <Keybuk> we already know that ifdown'ing the interfaces and putting the drives to standby is unnecessary
[07:49] <BenC> Keybuk: Probably not necessary, but "better safe than sorry"