Re: Fw: Re: 2.6.15-rc5-mm1

[Thomas Renninger <trenn-l3A5Bk7waGM@public.gmane.org>]

Voluspa wrote:
> On Sat, 10 Dec 2005 13:20:15 +0300 Roman I Khimov wrote:
>> | Voluspa:
>>> Nope, neither "nocst" nor "acpi=nocst" gave me back the C1
>>> functionality...
>> Yep, same here, just tried.
> 
> Sorry about the red herring. I've patched 2.6.14 with the two acpi
> suggestions for this stable series that Greg KH posted (both single and
> combined):
> 
> [patch 10/26] ACPI: Prefer _CST over FADT for C-state capabilities
> http://marc.theaimsgroup.com/?l=linux-kernel&m=113446277408937&w=2
> 
> [patch 12/26] ACPI: Add support for FADT P_LVL2_UP flag
> http://marc.theaimsgroup.com/?l=linux-kernel&m=113446277427267&w=2
> 
> I thought nr 10 was the cause for the C-state regression, but neither
> effected my machine. C1 was used as it should. Now, I don't have the
> stomach for an un-cheerleaded "git bisect" of approximately four hours
> to find the real culprit. So unless someone waves the pompoms I'll slip
> quietly into the shadows, hoping the issue will disappear by itself
> before 2.6.15 is finalized.
> 
Maybe it is this little patch that also has been applied in rc5?
http://bugzilla.kernel.org/show_bug.cgi?id=5452

If you have ACPI_DEBUG=y compiled in you should have seen:
"BIOS reporting wrong ACPI idfor the processor", then it might be this one.

Good luck,

   Thomas


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click