>-----Original Message----- >From: [EMAIL PROTECTED] >[mailto:[EMAIL PROTECTED] On Behalf Of Török Edvin >Sent: Thursday, June 21, 2007 5:38 AM >To: Florian Reitmeir >Cc: [email protected] >Subject: Re: CPU C4 vs. C3 states, BIOS support and 1.5W power savings > >On 6/21/07, Florian Reitmeir <[EMAIL PROTECTED]> wrote: >> >> it shows me max C3 in 2.6.21 , and 2.6.20 show me max. C4. > >2.6.20 is not tickless IIRC, so C4 won't have a big impact. > >If you had C4 in 2.6.20, but not in 2.6.21 I suggest to open a >bugreport on bugzilla.kernel.org. > >Perhaps use acpidump to dump the _CST tables. It is a bit complicated >how to do that, you need to dump the SSDT, disassemble with iasl, >locate the CST physical address, and length, dump again from that >address, and decompile again. > >Are you using same BIOS with 2.6.21 vs. 2.6.20? >
Florian, Yes. If you see this regression between 2.6.20 and 2.6.21, please open a bugzilla at bugzilla.kernel.org and assign it to me.. I don't know of any changes in C-state detection code in that time. Also, can you double check you checked both kernels with same BIOS and same setting (both on A/C power or both on battery). Typically BIOS make additional C-states visible on battery. Thanks, Venki _______________________________________________ Power mailing list [email protected] http://www.bughost.org/mailman/listinfo/power
