Bug 436959 - kernel-2.6.24.3-28.fc8 breaks ACPI battery status
Summary: kernel-2.6.24.3-28.fc8 breaks ACPI battery status
Keywords:
Status: CLOSED UPSTREAM
Alias: None
Product: Fedora
Classification: Fedora
Component: kernel
Version: 8
Hardware: All
OS: Linux
low
low
Target Milestone: ---
Assignee: Kernel Maintainer List
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2008-03-11 12:26 UTC by drago01
Modified: 2008-03-21 16:23 UTC (History)
3 users (show)

Fixed In Version: 2.6.24.3-34
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2008-03-21 16:23:25 UTC
Type: ---


Attachments (Terms of Use)
broken files under /proc/acpi (782 bytes, text/plain)
2008-03-11 15:30 UTC, drago01
no flags Details
correct files under /proc/acpi (779 bytes, text/plain)
2008-03-11 15:30 UTC, drago01
no flags Details
acpidump under the working kernel (-33) (124.11 KB, text/plain)
2008-03-12 17:25 UTC, drago01
no flags Details
acpidump under the broken kernel (-28) (124.11 KB, text/plain)
2008-03-12 17:25 UTC, drago01
no flags Details
requested debug output (123.94 KB, text/plain)
2008-03-15 22:40 UTC, drago01
no flags Details
requested dmesg output with cpufreq.debug=7 (28.52 KB, text/plain)
2008-03-20 09:57 UTC, drago01
no flags Details
acpidump output (124.11 KB, text/plain)
2008-03-20 09:58 UTC, drago01
no flags Details

Description drago01 2008-03-11 12:26:41 UTC
Description of problem:
After this changes:
--------
* Mon Mar 10 2008 Chuck Ebbert <cebbert@redhat.com> 2.6.24.3-25
- Use better fix for clearing the direction flag for x86 signal handlers.

* Mon Mar 10 2008 Chuck Ebbert <cebbert@redhat.com> 2.6.24.3-24
- ACPI: really disable stray GPEs (#251744)
- x86: clear direction flag before calling signal handlers
---------
The kernel no longer reports the battery status correctly and does not detect
when I plug in the AC adapter.
powertop also only reports 1.8 W power usage and never changes (same for
gnome-power-manager). 

kernel-2.6.24.3-18.fc8 works fine. 

If more infos or tests are needed feel free to ask.

Comment 1 drago01 2008-03-11 15:30:23 UTC
Created attachment 297627 [details]
broken files under /proc/acpi

Comment 2 drago01 2008-03-11 15:30:53 UTC
Created attachment 297628 [details]
correct files under /proc/acpi

Comment 3 Chuck Ebbert 2008-03-11 21:21:26 UTC
Patch that most probably caused the was removed in 2.6.24.3-32.

Comment 4 Zhang Rui 2008-03-12 03:20:52 UTC
For the unpatched kernel, could you please do the following test?
echo 0x04 > /sys/module/acpi/parameters/debug_layer
echo 0x8800001f > /sys/module/acpi/parameters/debug_level
plug the AC and attach the dmesg output.
And please attach the acpidump as well. ;)

Comment 5 drago01 2008-03-12 17:24:26 UTC
(In reply to comment #4)
> For the unpatched kernel, could you please do the following test?
> echo 0x04 > /sys/module/acpi/parameters/debug_layer
> echo 0x8800001f > /sys/module/acpi/parameters/debug_level
> plug the AC and attach the dmesg output.
> And please attach the acpidump as well. ;)

There is no such file on my here .. missing config option?

Comment 6 drago01 2008-03-12 17:25:01 UTC
Created attachment 297809 [details]
acpidump under the working kernel (-33)

Comment 7 drago01 2008-03-12 17:25:32 UTC
Created attachment 297810 [details]
acpidump under the broken kernel (-28)

Comment 8 Chuck Ebbert 2008-03-12 22:31:39 UTC
(In reply to comment #5)
> (In reply to comment #4)
> > For the unpatched kernel, could you please do the following test?
> > echo 0x04 > /sys/module/acpi/parameters/debug_layer
> > echo 0x8800001f > /sys/module/acpi/parameters/debug_level
> > plug the AC and attach the dmesg output.
> > And please attach the acpidump as well. ;)
> 
> There is no such file on my here .. missing config option?

You need to install and boot the kernel from the kernel-debug package.



Comment 9 Zhang Rui 2008-03-13 07:20:33 UTC
Well, I don't see why the patch breaks the battery on this laptop.
Any update on this? :)


Comment 10 Chuck Ebbert 2008-03-13 19:54:43 UTC
drago01, have you tested a kernel after 2.6.24.3-32?

Comment 11 drago01 2008-03-13 20:30:00 UTC
(In reply to comment #10)
> drago01, have you tested a kernel after 2.6.24.3-32?

yes 2.6.24.3-33.fc8 is working fine.

Comment 12 drago01 2008-03-13 20:30:35 UTC
(In reply to comment #9)
> Well, I don't see why the patch breaks the battery on this laptop.
> Any update on this? :)
> 

will install the debug kernel tomorrow and provide the debug info.

Comment 13 drago01 2008-03-15 22:40:35 UTC
Created attachment 298166 [details]
requested debug output

Sorry for the delay, here is the requested debug output. I first tried with the
debug build of kernel-2.6.24.3-28.fc8 booted with AC plugged in, enabled the
debug output, unplugged it and replugged it but nothing happened (it does print
stuff though when I press the power button).
Then I did the same with the debug build of kernel-2.6.24.3-34.fc8 (the one
with the patch reverted) and attached the output here. I hope this helps.

Comment 14 Zhang Rui 2008-03-17 05:57:21 UTC
I suspect the patch a7f9b1f24974da287771e2d70b30d9ca7bd66684 is not in kernel-
2.6.24.3-XX.fc8.
Chuck, can you check this?

Comment 15 Chuck Ebbert 2008-03-18 19:16:36 UTC
(In reply to comment #14)
> I suspect the patch a7f9b1f24974da287771e2d70b30d9ca7bd66684 is not in kernel-
> 2.6.24.3-XX.fc8.
> Chuck, can you check this?

No it's not. If that is added, then should the new patch work on top of that?

  https://bugzilla.redhat.com/attachment.cgi?id=297134

Comment 16 Chuck Ebbert 2008-03-19 16:44:19 UTC
Closing this - 2.6.24.3-34 is out now.

Comment 17 drago01 2008-03-20 00:01:54 UTC
Battery  reporting seems to work fine with -47 (grabed from
http://koji.fedoraproject.org/koji/taskinfo?taskID=522761), which has this
patches applied.
But it introduced a new problem :(.
CPU frequency scaling no  longer works. The cpu stays at 2.17 Ghz.
Trying to load the acpi-cpufreq module results into:
"FATAL: Error inserting acpi_cpufreq
(/lib/modules/2.6.24.3-47.fc8/kernel/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.ko):
No such device"

CPU info output:
 cat /proc/cpuinfo 
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 15
model name      : Intel(R) Core(TM)2 CPU         T7400  @ 2.16GHz
stepping        : 6
cpu MHz         : 2166.816
cache size      : 4096 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx lm
constant_tsc arch_perfmon pebs bts rep_good pni monitor ds_cpl vmx est tm2 ssse3
cx16 xtpr lahf_lm
bogomips        : 4337.31
clflush size    : 64
cache_alignment : 64
address sizes   : 36 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 15
model name      : Intel(R) Core(TM)2 CPU         T7400  @ 2.16GHz
stepping        : 6
cpu MHz         : 2166.816
cache size      : 4096 KB
physical id     : 0
siblings        : 2
core id         : 1
cpu cores       : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 10
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx lm
constant_tsc arch_perfmon pebs bts rep_good pni monitor ds_cpl vmx est tm2 ssse3
cx16 xtpr lahf_lm
bogomips        : 4333.51
clflush size    : 64
cache_alignment : 64
address sizes   : 36 bits physical, 48 bits virtual
power management:
----
Is this related to this patches or should I open a new bug?

Comment 18 drago01 2008-03-20 02:08:11 UTC
(In reply to comment #17)
> Battery  reporting seems to work fine with -47 (grabed from
> http://koji.fedoraproject.org/koji/taskinfo?taskID=522761), which has this
> patches applied.
> But it introduced a new problem :(.
> CPU frequency scaling no  longer works. The cpu stays at 2.17 Ghz.
> Trying to load the acpi-cpufreq module results into:
> "FATAL: Error inserting acpi_cpufreq
> (/lib/modules/2.6.24.3-47.fc8/kernel/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.ko):
> No such device"
[...]
> Is this related to this patches or should I open a new bug?

OK, its unrelated to this patches.
"linux-2.6-acpi-fix-sizeof.patch" was causing it.

Comment 19 Zhang Rui 2008-03-20 02:12:46 UTC
do you mean "linux-2.6-acpi-fix-sizeof.patch" caused a regression and the 
cpufreq driver work again after reverting it?

Comment 20 yakui.zhao 2008-03-20 02:20:53 UTC
Will you please add the boot option of "cpufreq.debug=7" and attach the output
of dmesg?
Of course the output of acpidump is needed.

Comment 21 drago01 2008-03-20 09:43:02 UTC
(In reply to comment #19)
> do you mean "linux-2.6-acpi-fix-sizeof.patch" caused a regression and the 
> cpufreq driver work again after reverting it?

yes

Comment 22 drago01 2008-03-20 09:57:31 UTC
Created attachment 298676 [details]
requested dmesg output with cpufreq.debug=7

Comment 23 drago01 2008-03-20 09:58:07 UTC
Created attachment 298677 [details]
acpidump output

Comment 24 Zhang Rui 2008-03-21 06:55:09 UTC
this seems to be a kernel bug, please open a bug at 
http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI
and attach the dmesg output and acpidump there. :)
BTW: please get the acpidump with the latest pmtools at
http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI
(no parameter needed)

Chuck, I think this bug can be closed now, :)

Comment 25 drago01 2008-03-21 16:23:25 UTC
(In reply to comment #24)
> this seems to be a kernel bug, please open a bug at 
> http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI
> and attach the dmesg output and acpidump there. :)
> BTW: please get the acpidump with the latest pmtools at
> http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI
> (no parameter needed)

Done: http://bugzilla.kernel.org/show_bug.cgi?id=10299
 
> Chuck, I think this bug can be closed now, :)

ok, closing as UPSTREAM


Note You need to log in before you can comment on or make changes to this bug.