Re: Re: [Swsusp-devel] swsusp and ac status

["Michael Frank" <mhf-jjFNsPSvq+iXDw4h08c5KA@public.gmane.org>]

On Mon, 28 Jun 2004 00:28:42 +0200, Karol Kozimor <sziwan-DETuoxkZsSqrDJvtcaxF/A@public.gmane.org> wrote:

> Thus wrote Michael Frank:
>> Too bad, this implies that the AC module does not get reinitialized after
>> resume and I suspect that acpi core is broken on 2.6 in this regard.
>
> As far as I know, there is no need to reinitialize the AC module, as it
> does not cache the results -- upon read, it calls the appropriate _PSR
> method which is supposed to return the correct AC adapter state, as simple
> as it gets. This way, any problems would indicate a faulty BIOS.
> Best regards,
>

Sorry, the BIOS does not know about suspend.

Please consider It just booted cold and the resuming kernel
without any ACPI modules won't have touched the AC method.

----> Thus the BIOS is now in a post boot state, like before
	vmlinux starts on boot!

----> The resumed kernel is in the state it was upon suspend.

Therefor IMO cause is the kernel which was resumed and
continues to talk to the BIOS which is in a state quite different
than what it was pefore suspend...

I concur that the BIOS in this case may break, but as it worked
prior to suspend it cannot be blamed for not handling such abuse.

A simple solution is for the resumed kernel to  start talking to
BIOS as it would have on a normal boot, perhaps executing
just a subset of the init code.

If it also would disable all modules prior to suspend, it
would likely fix all problems eliminating the reuqirement
to unload ACPI modules....

Best Regards



-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical experts, no vendor pitches,
unmatched networking opportunities. Visit www.blackhat.com