| Laptop mode |
| =========== |
| |
| This small doc describes the 2.4 laptop mode patch. |
| |
| Last updated 2003-05-25, Jens Axboe <axboe@suse.de> |
| |
| Introduction |
| ------------ |
| |
| A few properties of the Linux vm makes it virtually impossible to attempt |
| to spin down the hard drive in a laptop for a longer period of time (more |
| than a handful of seconds). This means you are lucky if you can even reach |
| the break even point with regards to power consumption, let alone expect any |
| decrease. |
| |
| One problem is the age time of dirty buffers. Linux uses 30 seconds per |
| default, so if you dirty any data then flusing of that data will commence |
| at most 30 seconds from then. Another is the journal commit interval of |
| journalled file systems such as ext3, which is 5 seconds on a stock kernel. |
| Both of these are tweakable either from proc/sysctl or as mount options |
| though, and thus partly solvable from user space. |
| |
| The kernel update daemon (kupdated) also runs at specific intervals, flushing |
| old dirty data out. Default is every 5 seconds, this too can be tweaked |
| from sysctl. |
| |
| So what does the laptop mode patch do? It attempts to fully utilize the |
| hard drive once it has been spun up, flushing the old dirty data out to |
| disk. Instead of flushing just the expired data, it will clean everything. |
| When a read causes the disk to spin up, we kick off this flushing after |
| a few seconds. This means that once the disk spins down again, everything |
| is up to date. That allows longer dirty data and journal expire times. |
| |
| It follows that you have to set long expire times to get long spin downs. |
| This means you could potentially loose 10 minutes worth of data, if you |
| set a 10 minute expire count instead of just 30 seconds worth. The biggest |
| risk here is undoubtedly running out of battery. |
| |
| Settings |
| -------- |
| |
| The main knob is /proc/sys/vm/laptop_mode. Setting that to 1 switches the |
| vm (and block layer) to laptop mode. Leaving it to 0 makes the kernel work |
| like before. When in laptop mode, you also want to extend the intervals |
| desribed above. See the laptop-mode.sh script for how to do that. |
| |
| It can happen that the disk still keeps spinning up and you don't quite |
| know why or what causes it. The laptop mode patch has a little helper for |
| that as well, /proc/sys/vm/block_dump. When set to 1, it will dump info to |
| the kernel message buffer about what process caused the io. Be very careful |
| when playing with this setting, it is advisable to shut down syslog first! |
| |
| Result |
| ------ |
| |
| Using the laptop-mode.sh script with its default settings, I get the full |
| 10 minutes worth of drive spin down. Provided your work load is cached, |
| the disk will only spin up every 10 minutes (well actually, 9 minutes and 55 |
| seconds due to the 5 second delay in flushing dirty data after the last read |
| completes). I can't tell you exactly how much extra battery life you will |
| gain in laptop mode, it will vary greatly on the laptop and workload in |
| question. The only way to know for sure is to try it out. Getting 10% extra |
| battery life is not unrealistic. |
| |
| Notes |
| ----- |
| |
| Patch only changes journal expire time for ext3. reiserfs uses a hardwire |
| value, should be trivial to adapt though (basically just make it call |
| get_buffer_flushtime() and uses that). I have not looked at other |
| journalling file systems, I'll happily accept patches to rectify that! |