Bug 15879 - no dsdt entry for the A/C adapter causing no entry in /proc/acpi/* and /sys/class/power_supply
Summary: no dsdt entry for the A/C adapter causing no entry in /proc/acpi/* and /sys/c...
Status: REJECTED INVALID
Alias: None
Product: ACPI
Classification: Unclassified
Component: Power-Battery (show other bugs)
Hardware: All Linux
: P1 high
Assignee: Zhang Rui
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-04-29 16:26 UTC by Justin P. Mattock
Modified: 2010-05-10 08:43 UTC (History)
4 users (show)

See Also:
Kernel Version: current
Subsystem:
Regression: No
Bisected commit-id:


Attachments
dsdt of the iMac9,1 (207.58 KB, text/x-dsl)
2010-04-29 16:37 UTC, Justin P. Mattock
Details
Xorg.0.log (14.48 KB, text/x-log)
2010-04-29 16:41 UTC, Justin P. Mattock
Details
dmesg iMac9,1(open suse 11.2) (62.34 KB, text/plain)
2010-04-29 16:48 UTC, Justin P. Mattock
Details

Description Justin P. Mattock 2010-04-29 16:26:45 UTC
on my iMac9,1 there is no entry in the dsdt for the A/C adapter causing the kernel to not know what todo i.g. no entry in /proc/acpi/* and /sys/class/power_supply/* stating the current status. online/offline

on one of my machines there is an error in /var/log/Xorg.0.log:

(II) Apr 29 00:01:04 NVIDIA: Using 768.00 MB of virtual memory for
indirect memory access.
(II) Apr 29 00:01:04 NVIDIA(0): Initialized GPU GART.
(WW) Apr 29 00:01:04 NVIDIA(0): ACPI: AC power state information is
not available under
(WW) Apr 29 00:01:04 NVIDIA(0):     /sys/class/power_supply/ , nor under
(WW) Apr 29 00:01:04 NVIDIA(0):     /proc/acpi/ac_adapter/
(WW) Apr 29 00:01:04 NVIDIA(0): ACPI: failed to determine the system's
current power source
(II) Apr 29 00:01:04 NVIDIA(0): Setting mode "nvidia-auto-select"



resulting in nvidia defaulting the system being on the battery(even though iMac's don't have batteries).

regardless of nvidia's status though the entries in /proc/acpi/* and /sys/class/power_supply remain empty.

attached(when I locat it), is the iMac's dsdt dmesg and Xorg.0.log
Comment 1 Justin P. Mattock 2010-04-29 16:37:28 UTC
Created attachment 26184 [details]
dsdt of the iMac9,1

here is the dsdt.dsl of the iMac9,1 for the record
I have two machines that are iMac9,1's but with two different gpu's

A= iMac9,1 with the nvidia gt130 card
that defaults to AC status

B= iMac9,1 with the GeForce 9400 that defaults
to battery status.

both dsdt's look pretty much the same, as well as the dsdt from macosx.
Comment 2 Justin P. Mattock 2010-04-29 16:41:37 UTC
Created attachment 26185 [details]
Xorg.0.log

the GeForce 9400 is the only machine right now the shows this error

(WW) Apr 29 00:01:04 NVIDIA(0): ACPI: AC power state information is not available under
(WW) Apr 29 00:01:04 NVIDIA(0):     /sys/class/power_supply/ , nor under
(WW) Apr 29 00:01:04 NVIDIA(0):     /proc/acpi/ac_adapter/
(WW) Apr 29 00:01:04 NVIDIA(0): ACPI: failed to determine the system's current power source


maybe nvidia defaults this card to "battery", since the majority of gpu's are on laptops(powersaving etc..(but could be wrong)).
Comment 3 Justin P. Mattock 2010-04-29 16:48:09 UTC
Created attachment 26186 [details]
dmesg iMac9,1(open suse 11.2)

and dmesg of the system(a little messy with avc's everywhere).

as it stands I've been looking at this, and am thinking in the area of some dmi blacklist for ac.c and battery that bassically(for machines that are in this situation) tricks the kernel into the status
(given the hardware), either on A/C or battery using proc_mkdir or something.
Comment 4 Justin P. Mattock 2010-05-01 00:35:57 UTC
It seems after receiving a post from lkml, that this is a problem with nvidia.
if everybody agrees then this can be closed.
I've sent in a post to linux-bugs@nvidia.com and linux-nforce-bugs@nvidia.com
so hopefully I can get this resolved.
Comment 5 Zhang Rui 2010-05-10 07:06:01 UTC
please attach the acpidump output of this laptop.
Comment 6 Alexey Starikovskiy 2010-05-10 07:51:19 UTC
Zhang,
This is not notebook, it's a desktop (all-in-one), so it should not have ACPI AC adapter...
Comment 7 Zhang Rui 2010-05-10 07:55:15 UTC
hah, I see. Alexey, thanks for reminding.

close this bug
Comment 8 Justin P. Mattock 2010-05-10 08:43:09 UTC
hi,
received a post from the nvidia people.. so they are aware of this
issue..(no words from them if it might be kernel and such),
so yes at this point please close.. sorry for the newbiness!!

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