-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Mattia Dongili wrote: > On Sun, Mar 11, 2007 at 10:50:35AM +0100, Thorsten Schmidt wrote: >> Hello, >> >> Am Freitag, 9. März 2007 23:12 schrieb Mattia Dongili: >>> On Fri, Mar 09, 2007 at 09:52:20PM +0100, Thorsten Schmidt wrote: >>>> Am Freitag, 9. März 2007 21:41 schrieb Mike Hommey: >>>>> On Fri, Mar 09, 2007 at 09:25:46PM +0100, Thorsten Schmidt >>>> <[EMAIL PROTECTED]> wrote: >>>>>> Am Freitag, 9. März 2007 20:42 schrieb Mike Hommey: >>>>>>> On Fri, Mar 09, 2007 at 08:32:21PM +0100, Thorsten Schmidt >>>>>> <[EMAIL PROTECTED]> wrote: >>>>>>>> for some time I've been trying to reduce my laptop's power >>>>>>>> consumption (while running on battery power) on my >>>>>>>> Fujitsu-Siemens Lifebook C1320 with Debian etch. >>>>>>>> - I echo powersave to >>>>>>>> /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor - I >>>>>>>> installed laptop-mode in order to shut down my harddrive if it is >>>>>>>> unused. - I don't use swap, /tmp/ is a ramdisk-mount and >>>>>>>> /var/tmp links to /tmp, (I've 1 GB RAM) >>>>>>>> Thus my hd is barely used and my CPU reaches C3. >>>>>>>> >>>>>>>> But for some reason, Windows is more efficient in power saving. >>>>>>>> (3,5h vs. 4,5h battery life while surfing wireless). >>>>>>>> Is there anything else I can do to save battery power? >>>>>>> Basically, no, except waiting for Linux ACPI support to get better. >>>> Is this about to happen soon? >>>> Would switching from vanilla (2.6.20.1) to -mm be an option? >>> probably using CONFIG_NO_HZ=y (yes, available in -mm), but you'll hardly >>> get one more hour. >>> Give it a shot. >> Thanks, I'll do so. - the kernel is compiling now. >> Btw. back in good old 2.4 times, I recently patched ACPI to make it useable. >> Is patching ACPI still a good idea? > > no. Moreover if you switch to -mm you'll get the acpi test branch which > is ahead of plain 2.6 kernels anyway. > > To save some more power in linux you may also want to try unloading > usb modules. If your processor supports it, you could try undervolting, I got about 15 more minutes out of it, but the main gain, is that temperature stays remains a lot lower, and so the fan also works less, which saves noise, and some energy too. Here is where I looked at: http://www.thinkwiki.org/wiki/Pentium_M_undervolting_and_underclocking . The stress testing part is very important ihmo, and I have to tell you that i am not 100% sure it is absolutely safe for the system, although i didn't have any problems.
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFF9HLpmtnb6d8/tK8RAqQtAJoClOWsW1hX1vS7CthrwnsNrCioqgCeNhpd SaxmXzWxHUD3735DZgWBfgA= =PIqx -----END PGP SIGNATURE-----
