Bug 2310

Summary:	No battery informations on Toshiba laptop Tecra S1
Product:	ACPI	Reporter:	Christian Mauduit (ufoot)
Component:	Power-Battery	Assignee:	Shaohua (shaohua.li)
Status:	REJECTED DUPLICATE
Severity:	normal
Priority:	P2
Hardware:	i386
OS:	Linux
Kernel Version:	2.6.3	Subsystem:
Regression:	---	Bisected commit-id:
Attachments:	output of dmesg with a standard 2.6.3 kernel output of dmesg with the "patched" 2.6.3 kernel output of dmidecode output of acpidmp output of lspci -vv content of /proc/interrupts .config of the "standard" 2.6.3 kernel .config of the patched 2.6.3 kernel the patch applied to make the battery APCI module work the patch applied to make the battery APCI module work

Description Christian Mauduit 2004-03-15 07:56:30 UTC

Distribution: Debian GNU/Linux testing (sarge)

Hardware Environment: Toshiba Tecra S1 (laptop)

Software Environment: Official mainstream 2.6.3 kernel, tainted with fglrx (ATI
proprietary driver). However installing or removing the ATI driver does not seem
to affect ACPI problems at all, in any way.

Problem Description: No informations in /proc/battery and also no informations
from acpi programs concerning the battery. Note that the battery module seems to
be loaded OK, however it does not seem to return any valuable information, if
any information at all. /proc/acpi/battery is an empty directory. It does not
work either with a 2.4.25 kernel. However I downloaded a patch on
http://www.a-za-z0-9.de/linux/s1acpifix.html and now I get proper info from my
battery ACPI module and programs. Point is it generates loads of warnings and
possibly errors when loading, and wrecked some other ACPI related modules. I
don't care much for the only information I really need is how much battery I
have left, but I presume if the patch or a derivative of it is to be included in
the stable branch, it has to work correctly all the way. FYI when applying the
patch on 2.6.3 there were 3 rejected lines, but it's rather trivial to apply
them "manually". Of course I'm willing to provide informations to help debugging
this, I have the laptop, basic C programming skills, and can perform more tests
if needed.

Steps to reproduce: Have a Tecra notebook, install ACPI module and/or tools,
then type "/proc/acpi/battery" and "acpi -V"

Comment 1 Christian Mauduit 2004-03-15 07:58:14 UTC

Created attachment 2333 [details]
output of dmesg with a standard 2.6.3 kernel

Comment 2 Christian Mauduit 2004-03-15 08:07:22 UTC

Created attachment 2334 [details]
output of dmesg with the "patched" 2.6.3 kernel

There are a bunch of warnings, more than with a standard kernel, but however in
this case the battery module works.

Comment 3 Christian Mauduit 2004-03-15 08:08:06 UTC

Created attachment 2335 [details]
output of dmidecode

Comment 4 Christian Mauduit 2004-03-15 08:08:52 UTC

Created attachment 2336 [details]
output of acpidmp

Comment 5 Christian Mauduit 2004-03-15 08:09:46 UTC

Created attachment 2337 [details]
output of lspci -vv

Comment 6 Christian Mauduit 2004-03-15 08:10:15 UTC

Created attachment 2338 [details]
content of /proc/interrupts

Comment 7 Christian Mauduit 2004-03-15 08:11:20 UTC

Created attachment 2339 [details]
.config of the "standard" 2.6.3 kernel

Comment 8 Christian Mauduit 2004-03-15 08:11:52 UTC

Created attachment 2340 [details]
.config of the patched 2.6.3 kernel

Comment 9 Christian Mauduit 2004-03-15 08:14:19 UTC

Created attachment 2341 [details]
the patch applied to make the battery APCI module work

The patch is not appliable by a simple patch -p0 on kernel 2.6.3, there are 3
lines rejected but I was able to apply them manually.

Comment 10 Christian Mauduit 2004-03-15 08:17:21 UTC

Created attachment 2342 [details]
the patch applied to make the battery APCI module work

The patch is not appliable by a simple patch -p0 on kernel 2.6.3, there are 3
lines rejected but I was able to apply them manually.

Comment 11 Shaohua 2004-03-21 18:25:53 UTC

ECDT is required on your system. please try the patch in Bug 1690

Comment 12 Christian Mauduit 2004-03-31 05:13:00 UTC

I tried the patch in Bug 1690 and it did solve my problem. Now battery is
recognized, I don't get weird errors at boot time (dmesg looks fined) and other
ACPI modules work as well (thermal for instance, which used to be broken with
the other patch).

BTW, I applied the patch proposed for Bug 1690 on a 2.6.4 source tree, and a few
lines were rejected, but it's rather trivial to apply the rejected changes
manually. If you wish I can post the new .diff appliable on 2.6.4. Maybe I
should post it on the Bug 1690 page?

Thanks for your help!

Comment 13 Shaohua 2004-04-01 01:12:11 UTC

Since the patch in 1690 works for you, assign it as duplicate with 1690

*** This bug has been marked as a duplicate of 1690 ***