Aubrey Li wrote:
> 2009/1/16 Mark Haywood <Mark.Haywood at sun.com>:
>   
>> And this was the beginning of the problem that we are seeing with Bug 6113.
>> Notice that though in these tables the first 9 P-states are identical, the
>> _PPC is zero. Gotta love it.
>>
>> Hi,
>>
>> An OpenSolaris user reported a problem with SpeedStep support on his laptop.
>> His problem was that OpenSolaris was not finding the supported P-states.
>> After a few email exchanges, we found that his _PSS table had a strangely
>> defined set of P-states (see the attached stbl.dsl). There are 10 P-states
>> returned by the _PSS, but the first 9 are duplicates. So, really there are
>> only 2 uniquely defined P-states. The current P-state parsing code in
>> Solaris doesn't allow for duplicates in the middle of the table (it does
>> handle them at the end of the table since we've seen that case before).
>> Though I consider this to be a questionable _PSS defintion, I think we can
>> support it easy enough by ignoring consecutive duplicates altogether.
>>
>> I placed a webrev of the fix at:
>>
>> http://cr.opensolaris.org/~mhaywood/6716347/
>>
>> And  I welcome any comments.
>>
>> Thanks!
>> Mark
>>
>>     
> It looks like 6716347 doesn't exist under your folder.
>   

Sorry. I wasn't asking for a code review. I was forwarding on a 
discussion from the middle of last year to point out that we'd run into 
the issue of non-unique P-states (with ASUS) and that we'd modified the 
ACPI _PSS parsing code to ignore the duplicates. At that time, they were 
returning a _PPC value of zero even though the first 9 P-states were 
duplicates. Now they are returning a _PPC value that points to the last 
duplicate value. And apparently in their very latest BIOS they got 
things right and stopped supplying the duplicate entries.

> -Aubrey
> _______________________________________________
> tesla-dev mailing list
> tesla-dev at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/tesla-dev
>   


Reply via email to