Bug#471141: acpi-support: vm and wireless settings to save power while running on battery
Package: acpi-support Version: 0.103-5 Severity: wishlist Tags: patch Here are two scripts from https://wiki.ubuntu.com/ReducedPowerUsage which I have added to /etc/acpi/ac.d and /etc/acpi/battery.d to try to reduce my power usage while on battery. They have to do with tweaking the Linux virtual memory manager and setting the wireless adapter (Intel ipw-3945 in my case) power-saving mode. Francois -- System Information: Debian Release: lenny/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable') Architecture: amd64 (x86_64) Shell: /bin/sh linked to /bin/dash Versions of packages acpi-support depends on: ii acpi-support-base 0.103-5scripts for handling base ACPI eve ii acpid 1.0.6-5.1 Utilities for using ACPI power man ii dmidecode 2.9-1 Dump Desktop Management Interface ii finger0.17-11user information lookup program ii hdparm8.3-1 tune hard disk parameters for high ii laptop-detect 0.13.5 attempt to detect a laptop ii libc6 2.7-9 GNU C Library: Shared libraries ii lsb-base 3.2-4 Linux Standard Base 3.2 init scrip ii nvclock 0.8b3-1Allows you to overclock your nVidi ii powermgmt-base1.30 Common utils and configs for power ii radeontool1.5-5 utility to control ATI Radeon back ii vbetool 1.0-1.1run real-mode video BIOS code to a ii x11-xserver-utils 7.3+2 X server utilities acpi-support recommends no packages. -- no debconf information ac.d__10-vm_setting.sh Description: application/shellscript ac.d__20_wireless_power.sh Description: application/shellscript battery.d__10-vm_setting.sh Description: application/shellscript battery.d__20_wireless_power.sh Description: application/shellscript
Bug#471141: acpi-support: vm and wireless settings to save power while running on battery
Francois Marier wrote: Package: acpi-support Version: 0.103-5 Severity: wishlist Tags: patch Here are two scripts from https://wiki.ubuntu.com/ReducedPowerUsage which I have added to /etc/acpi/ac.d and /etc/acpi/battery.d to try to reduce my power usage while on battery. They have to do with tweaking the Linux virtual memory manager and setting the wireless adapter (Intel ipw-3945 in my case) power-saving mode. Hi Francois, Thank you for contributing. At this point I think I will not include these additions, for several reasons: * They are not clearly intended to make things work (which is what the package is for, basically). * Making things work better (which is what these patches intend to do) usually involves some trade-offs, and those may require conscious decisions by the user, or at least configuration settings which you can turn off. So it's a bit more work than simply including these scripts. * Regarding trade-offs: the scripts may break things. For instance, if the power saving mode of iwl3945 wouldn't have drawbacks, it would probably have been on by default in the hardware. It wouldn't even have been a choice. :-) Probably, the power saving mode for iwl3945 will reduce the effective range of wireless, which is not acceptable if you use your laptop on battery in your garden. People will have a very hard time tracing such a loss of connectivity back to acpi-support if they didn't consciously choose to turn this power saving feature on. * I must say I don't really agree with the VM tweaks. For instance, if this means what I think it does: echo 0 /proc/sys/vm/dirty_expire_centisecs echo 0 /proc/sys/vm/dirty_writeback_centisecs then it will either turn off writeback, or it will make the laptop write everything back to the hard drive immediately. The former situation is very dangerous (power loss = lose everything!), while the latter option is asking for both performance problems and you can forget about ALPM power savings (see http://www.lesswatts.org/tips/disks.php) and spinning down the disk when it is idle. * The VM settings are also tweaked by laptop-mode-tools, which many laptop users install as well. I make an explicit point of not interfering with what laptop-mode-tools handles. And arguably, the combinations of settings applied used by laptop-mode-tools save more power, because they actually make disk I/O more chunky, allowing the hardware to go into power saving mode in between batches. * Laptop mode tools also allows you to enable wireless power saving, and does it better: it supports this for several other types of wireless interfaces as well. So if you want it, it's already there, no programming required. Cheers, Bart -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#471141: acpi-support: vm and wireless settings to save power while running on battery
Hi Francois, On Sun, March 16, 2008 12:04, Francois Marier wrote: Hi Bart, Thanks for the prompt response and taking the time to provide all of these great pointers. Based on your clarifications of what acpi-support does, I'm gonna look into laptop-mode-tools and send patches there if it's missing anything. Those will also end up with me, so I'll consider myself warned. :-) BTW, laptop-mode-tools has a module mechanism, so you can extend it however you like. And I keep a much more open policy for laptop-mode-tools, so any useful power saving stuff is much more likely to be accepted -- and also to be pushed upstream (which is also me :-) ). You were also right about some of the potentially dangerous VM settings. I have tweaked them on my laptop now. Anyways, since acpi-support already recommends laptop-mode-tools, I see no reason for leaving this bug open. Thanks, that's one less bug to keep track of in this package that seems to attract bug reports like shit attracts flies. :-) Cheers, Bart
