Juliusz, On Wed, Dec 21, 2011 at 05:37:39PM +0100, Juliusz Chroboczek wrote: > This issue appears to be worked around by > linux-image-3.1.0-1-amd64 3.1.5-1 > Note that it is *not* fixed by > linux-image-3.2.0-rc4-amd64 3.2~rc4-1~experimental.1
I agree that kernel 3.1.5 (and 3.1.6) does not exhibit the "method parse/execution failed" ACPI error, whereas 3.2~rc4 does. > In case you're interested in what's going on under the hood -- the > kernel is now able to put the processor in one of the low-power > "package" states. Previously, the cores were able to switch to > low-power states, but the package remained in PC0. > > cr CPU %c0 GHz TSC %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6 %pc7 > 1.03 0.80 2.49 5.21 0.01 0.00 93.74 1.22 0.00 2.07 84.59 > 0 0 2.32 0.80 2.49 3.02 0.03 0.00 94.63 1.22 0.00 2.07 84.59 > 0 2 1.05 0.80 2.49 4.28 0.03 0.00 94.64 1.22 0.00 2.07 84.59 > 1 1 0.60 0.80 2.49 6.55 0.00 0.00 92.85 1.22 0.00 2.07 84.59 > 1 3 0.14 0.80 2.49 7.01 0.00 0.00 92.85 1.22 0.00 2.07 84.59 But what seems to really make a difference is adding i915.i915_enable_rc6=1 to GRUB_CMDLINE_LINUX in /etc/default/grub (and run update-grub), not using a different kernel. Despite the ACPI error, using this option with kernel 3.2 boosts battery life and turbostat shows the package reaching PC7 more than 90% of the time. In other words, I am not convinced this ACPI issue is related with the battery life (but I must admit I do not understand a lot about power-saving stuff). Best regards, -- Gabriel -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111228144359.gc22...@kerneis.info
