This bug is a regression. Things worked well in 2.6.25 and 2.6.24 kernels. Earliest failing kernel version: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1b7fc5aae8867046f8d3d45808309d5b7f2e036a Distribution: Debian Lenny using vanilla 2.6.26.3 kernel (also affects Debian 2.6.26 kernels) Hardware Environment: Asus Eee PC model 701 (reported on model 900 too) Software Environment: Debian Lenny Problem Description: During boot process the system reports a bogus temperature of 144C and shuts down, making it impossible to boot from battery. ACPI: Critical trip point Critical temperature reached (144 C), shutting down. See also http://bugs.debian.org/494984 Steps to reproduce: The bug shows most often (but not always) when booting from a battery which is slightly de-charged. Most of the time it is *not* possible to boot under these conditions, but sometimes everything works well (even a few times in a row, which makes it hard to track down). When on ac-adapter the bug is rarely seen. Using git bisect I have found commit 1b7fc5aae8867046f8d3d45808309d5b7f2e036a to introduce the problem The problem gets fixed with the following change: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=9d699ed92a459cb408e2577e8bbeabc8ec3989e1 I have tried this two line change with kernel 2.6.26.3, where it works well. Please apply this patch to the stable kernel tree. Robert Epprecht
Hi, Robert Now the following commit is already shipped in the 2.6.27-rc4 kernel. >commit 9d699ed92a459cb408e2577e8bbeabc8ec3989e1 >Author: Zhao Yakui <yakui.zhao@intel.com> > Date: Mon Aug 11 10:33:31 2008 +0800 >ACPI: Avoid bogus EC timeout when EC is in Polling mode
Hi, Zhao Yakui. Yes I know. Works well for me (best together with patched ec.c version from Alexey) But for distributions shipping with 2.6.26 kernels it is essential, that it get's applied to stable now. These kernels will still be around for quite some time. An exemple is our DebianEeePC project, based on Lenny. Robert Epprecht
I already sent the patch to the stable@kernel.org. Maybe it will take some time to merge it to the 2.6.26.x stable kernel. So the bug will be marked as resovled.
Hi, Robert Will you please attach the output of dmesg on the 2.6.26.3 kernel?(Of course the patch in comment #1 should be appled). Thanks.
Created attachment 18055 [details] dmesg 2.6.26.3 with patched ec.c on Asus Eee PC 701 The attachment is the dmesg output after booting 2.6.26.3 (with ec.c including the two line change mentioned in comment #1) on an Asus Eee PC 701 4G surf. Robert Epprecht
I have the same problem but if I put 2GB RAM to my eeepc 701 this problem disappear. With the original 512MB RAM I have the problem. I don't understand it, but I could have solved it.
shipped in 2.6.26.6 closed. Author: Zhao Yakui <yakui.zhao@intel.com> 2008-09-23 01:38:13 Committer: Greg Kroah-Hartman <gregkh@suse.de> 2008-10-08 23:23:01 Parent: 1b7fc5aae8867046f8d3d45808309d5b7f2e036a (ACPI: EC: Use msleep instead of udelay while waiting for event.) Branches: master, remotes/origin/master Follows: v2.6.26.5 Precedes: v2.6.26.6 ACPI: Avoid bogus EC timeout when EC is in Polling mode commit 9d699ed92a459cb408e2577e8bbeabc8ec3989e1 upstream