ACPI Power Switching Fails Toshiba L30

[Danny Wood]

Binary package hint: linux-image

First problem is that the laptop will neither startup nor shutdown without being plugged into AC power.

Second issue is that when the AC is removed whilst in Ubuntu the kernel does an Oops (attatched).

The ACPI worked fine in Gutsy, so far I have tried compiling a vanilla 2.6.24 which gave the same results, then tried the latest 2.6.26-rc1, this was marginally better but still gave an oops. (2.6.26-rc1 would start but not shutdown correctly)

Current Kernel Version:
Linux danny-laptop 2.6.24-17-generic #1 SMP Thu May 1 14:31:33 UTC 2008 i686 GNU/Linux

This was an issue with the default hardy kernel, 2.6.24-17, Vanilla 2.6.24, Vanilla 2.6.26-rc1.
Im going to recompile the 2.6.23 kernel to see if the ACPI works again on that version.

Please inform me if anymore information is required.

[Danny Wood (danwood76)]

Ok so I recompiled 2.6.23-17 of the linux kernel and now my ACPI works perfectly.
Im assuming there was some big ACPI change between 2.6.23 and 2.6.24 which has caused my problems.

BTW I am also running the latest BIOS (no thanks to Toshiba for their windows only BIOS update)

[Danny Wood (danwood76)]

I have tried booting this laptop from a fedora beta live CD (with 2.6.24 kernel) and it has the same issue.
So I believe this is actually a kernel bug.

I am trying to debug this myself but am not really sure where to start, ive tried forcing it to use the correct DSDT but this changed nothing.

[Danny Wood (danwood76)]

After further kernel compilation and testing I have managed to compile the latest ubuntu kernel and got it to work.

It turns out the config option CPU_IDLE (CPU idle PM Support) was causing my issues.
Enabling this in any version kernel later than 2.6.24 causes the kernel oops and system instability in my laptop.

System SPecs:

CPU: Intel Pentium Dual Core T2080 @ 1.73 GHz
Chipset: ATI SB400
RAM: 1GB
GPU: ATI Xpress 200

[zOu (kervoaz-gmail)]

Hi,

I confirm the problem with Toshiba L30-105. work fine only on AC power.
Problem before and after BIOS update.
Problem before and after last ubuntu update.

BR

[Danny Wood (danwood76)]

Disabling the kernel option I mentioned earlier fixes the issue completely, though it does require recompiling the kernel.

[Merayo (merayo)]

Hi from France

I also confirm the problem with Toshiba L30,but E11 (after 105 model), This laptop work fine but only on AC power connector.
The Bios update didn't change anything, like the updates.

Does anyone can tell me how to recompil the kernel ?

Thanks

[Danny Wood (danwood76)]

Heres a good guide/info:
https://help.ubuntu.com/community/Kernel/Compile#head-8587f42b684c3694ba72bdcc89e461ec289cbf02

At the config stage I would use the xconfig option and use the search function for CPU_IDLE, switch it off, save the config and continue with the build.
Once you have installed reboot and make sure you load the correct kernel from the grub menu.

Good luck.

[Merayo (merayo)]

There is news today

after the boot without ac power, the splash stopped and gived to me this information :

[ 36.209012] usplash[1223] segfault at b783e421 eip b7f7f13e esp bfcaa380 error 6

and nothing, i must reboot the computer

To be continued...

[Danny Wood (danwood76)]

Is that after recompiling the kernel?

[Merayo (merayo)]

No before recompiling the kernel
I didn't make it yet

To be continued...

[James Tarin (spam-meltwaterventures)]

I think this may be the relevant kernel bug:

http://bugzilla.kernel.org/show_bug.cgi?id=10394

[Danny Wood (danwood76)]

Interesting.
I applied the patch in that bug to the latest ubuntu source (-19), I had to modify it a little to get it to work on this version of the source, and now my laptop works when running on battery with a ubuntu kernel.
I think this patch needs to be applied to the ubuntu source in the future update.

Any idea if any devs actually read these bugs and want to add it to the source patches?

[James Tarin (spam-meltwaterventures)]

Hi Danwood,

can you post your modified patch here so that I can look at it / use it too?

many thanks

james

[Danny Wood (danwood76)]

Hi, here it is.
When I say I modded the patch I actually added the bits that didnt work manually.

I have remade the patch from the original using diff and just tested it on a fresh ubuntu source and it patched fine with the correct output.
So it should work fine.

[Danny Wood (danwood76)]

BTW to patch this you just stick it in the source dir and run patch like so:
patch -i cpuidle_init_issues_mod.patch -p1

[James Tarin (spam-meltwaterventures)]

Forgot to say that I also confirm this bug on the Toshiba Satellite L30. The fact that this kernel patch appears to fix it indicates that it most likely is related to http://bugzilla.kernel.org/show_bug.cgi?id=10394.

Danwood, thanks for supplying this patch. I will try it at the weekend, when I get a moment.

James

[James Tarin (spam-meltwaterventures)]

Patch supplied by danwood76 worked fine on my 2.6.24-18 kernel source. Fixed the AC power problem. Suspend and hibernate still totally broken though. Probablyl will try TuxOnIce and see if that fixes it.

[Danny Wood (danwood76)]

With regards to the suspend issue, this bug has also been squashed:
https://bugs.launchpad.net/ubuntu/hardy/+source/linux/+bug/226279

If you get the -19 kernel source the suspend issue is fixed, this patch works fine with it aswell.
Check the other bug, -19 should be in the latest set of updates I think.

[James Tarin (spam-meltwaterventures)]

Yep, suspend, hibernate and resume working fine with -19 on the L30. However, the above patch does not play 100% nicely with suspend and resume (if the AC power is taken out whilst hibernated, a kernel trace is encountered upon resume).

[Danny Wood (danwood76)]

I just tried what you said.
BTW my kernel is a fresh -19 source (from repos) with my patch added.

I booted up with power in and did a hibernate, then when the machine was completely off I removed the power and booted back up.
I have no kernel errors or oops or anything dis ordinary in dmesg.

Are you sure you're booting the patched kernel?

[James Tarin (spam-meltwaterventures)]

I suppose it is possible I made an error. I'm not going to try again in a hurry as i've recompiled without CPU_IDLE and all is fine. Don't have time to recompile again in very near future as am moving house and wife having a baby. Thanks for your help.

[taras (val-taras)]

Im new to linux and Ubuntu.
I downloaded patch. My system already upgraded to -19 kernel.
But I have no idea how to apply patch.
When I try to do "patch -i cpuidle_init_issues_mod.patch -p1" I have to type the name of the file to be patched.
I have no idea which file/files I have to patch.
Do I have to place patch file to /usr/src/linux-headers-2.6.24-19-generic and then "cd" to this directory?
Will really appreciate step by step instruction.

[travauxpublics (travauxpublics)]

Thanks for information
Sorry, but, what does it mean exactly ?
I understand that we can hope an update at next days, is it really like that
?

Good bye

2008/7/4 danwood76 <email address hidden>:

> ** Changed in: linux (Ubuntu)
> Status: New => Fix Released
>
> --
> ACPI Power Switching Fails Toshiba L30
> https://bugs.launchpad.net/bugs/228375
> You received this bug notification because you are a direct subscriber
> of the bug.
>

[Danny Wood (danwood76)]

The patch I released earlier in the thread (post 16) based on the patch from the main Linux kernel thread fixes the issue, this can be applied to the ubuntu kernel source and recompiled.
Weather or not it gets included in the Ubuntu kernel is up to the kernel dev team, although I am unsure as to weather they have even viewed this bug yet.

[Danny Wood (danwood76)]

This is fixed in the, now in alpha, intrepid ibex based on the 2.6.26 kernel.

So if you cant recompile the kernel it might be worth running intrepid although there are a few bugs floating around at this stage. (as you would expect)

[Launchpad Janitor (janitor)]

Per a decision made by the Ubuntu Kernel Team, bugs will longer be assigned to the ubuntu-kernel-team in Launchpad as part of the bug triage process. The ubuntu-kernel-team is being unassigned from this bug report. Refer to https://wiki.ubuntu.com/KernelTeamBugPolicies for more information. Thanks.