Bug 843

Summary:	Getting very high CPU load on Laptop with ACPI
Product:	ACPI	Reporter:	Hanno Boeck (hanno)
Component:	Power-Battery	Assignee:	Shaohua (shaohua.li)
Status:	REJECTED DUPLICATE
Severity:	normal	CC:	acpi-bugzilla
Priority:	P2
Hardware:	i386
OS:	Linux
Kernel Version:	2.5.72 (and others with new acpi code)	Subsystem:
Regression:	---	Bisected commit-id:
Attachments:	gkrellm screenshot showing cpu-usage. dmesg-output of kernel 2.4 dmesg-output of kernel 2.5 DSDT lspci-output original DSDT for ASUS S1300N bios version 0205 dmesg from 2.4.23-pre4 /proc/interrupts

Description Hanno Boeck 2003-06-22 13:55:21 UTC

I own a fujitsu-siemens lifebook C6155.
With several linux-kernels containing the new acpi (2.5.x, 2.4.21+patch,
2.4.21-ac1), I get very high CPU-load every few seconds. It's even impossible to
listen mp3s.
top shows me that it's the process events on 2.5-kernels and keventd on 2.4-kernels.
This does not happen with 2.4.21-vanilla (without acpi-patch, but with acpi
enabled and working).
So the bug seems to be in the new acpi-system.
Attached is dmesg-output (with 2.4 and 2.5), the dsdt, lspci-output and a
screenshot of gkrellm showing the cpu-usage (on a system running nothing that
consumes cpu-power).

Comment 1 Hanno Boeck 2003-06-22 13:55:58 UTC

Created attachment 449 [details]
gkrellm screenshot showing cpu-usage.

Comment 2 Hanno Boeck 2003-06-22 13:56:29 UTC

Created attachment 450 [details]
dmesg-output of kernel 2.4

Comment 3 Hanno Boeck 2003-06-22 13:56:58 UTC

Created attachment 451 [details]
dmesg-output of kernel 2.5

Comment 4 Hanno Boeck 2003-06-22 13:57:29 UTC

Created attachment 452 [details]
DSDT

Comment 5 Hanno Boeck 2003-06-22 13:57:51 UTC

Created attachment 453 [details]
lspci-output

Comment 6 sebastian henschel 2003-07-16 03:46:57 UTC

this seems to be true for all ASUS centrino notebooks, too.

there are reports about it for ASUS M2400N and M3xxxN in the list archives.
i am currently working on a S1300N which shows the same behaviour.
also unloading the thermal modules generates an oops and crashes the keyboard. i
can init 0 via the powerbutton, though.

i will attach the asl for the DSDT.

Comment 7 sebastian henschel 2003-07-16 03:49:34 UTC

Created attachment 542 [details]
original DSDT for ASUS S1300N bios version 0205

sorry, cannot update to 0206 due to lack of windows. the dos-tool for flashing
does not seem to work. :(

Comment 8 Shaohua 2003-09-09 19:50:04 UTC

can you try recent kernel? if still not work, please attach '/proc/interrupts' 
and dmesg. Thanks a lot.

Comment 9 Hanno Boeck 2003-09-13 09:17:19 UTC

Just tried again with kernel 2.4.23-pre4. Still the same problem.
With ACPI Debug enabled, I get this debug-string on every high cpu load:
[ACPI Debug] String: B1ST

On /proc/acpi, I get
battery BAT1 00000080 00000001

I'll attach new dmesg and /proc/interrupts

Comment 10 Hanno Boeck 2003-09-13 09:20:40 UTC

Created attachment 876 [details]
dmesg from 2.4.23-pre4

Comment 11 Hanno Boeck 2003-09-13 09:21:15 UTC

Created attachment 877 [details]
/proc/interrupts

Comment 12 Shaohua 2003-09-14 21:55:43 UTC

Dmesg shows that Battery keeps notifying status change. can you 
attach /proc/acpi/battery/BAT1/state & /proc/acpi/battery/BAT1/info? did you 
recharging the battery? Thanks a lot.

Comment 13 Luming Yu 2003-09-17 01:01:43 UTC

Would you please have patch at bug 1186 a try? thanks a lot.

Comment 14 Shaohua 2003-09-17 19:40:15 UTC

I find a piece of insteresting code in the dsdt:
Method (_Q09, 0, NotSerialized)
{
    Decrement (SMCN)
    If (LNot (SMCN))
    {
        Store (0x05, SMCN)
    }

    If (LNot (LLess (SMCN, 0x05)))
    {
        SMSL ()
    }

    BPOL ()
}
method SMSL() will print '[ACPI Debug] String: B1ST' as seen in dmesg. SMSL() 
executed so frequently that it reduced system overhead. In the method '_Q09', 
only when 'SMCN' equals to 0x05, SMSL will execute. that is 'SMSL' execute one 
time when '_Q09' execute five time. This seems to show that somebody knows 
that 'SMSL' will execute frequently, so he use this method to reduce the 
frequence of executing 'SMSL'. But his plan failed. I guess changing 0x05 
in '_Q09' to 0x20 or other number greater than 0x05 will partially resolve 
your problem.

Comment 15 Shaohua 2003-11-04 23:07:46 UTC

please try the patch in Bug 1433

Comment 16 Shaohua 2004-05-12 22:17:38 UTC

Could you please verify if you have new BIOS. We find similar issue which is 
caused by BIOS. Thanks.

Comment 17 Frederic Detienne 2004-08-29 14:13:43 UTC

Hi,

I have an IBM T-22 laptop and I see the problem too. The issue is triggered by
plugging/unplugging the power cord (going AC<->battery).

I currently run 2.6.8.1; what additional information can I provide ? I do not
see the issue mentioned in #1433.

thanks,

    Frederic Detienne

Comment 18 Shaohua 2004-11-03 23:47:39 UTC


*** This bug has been marked as a duplicate of 1691 ***