Alberto Gonzalez wrote: > On Saturday 03 January 2009 09:10:53 Bryan Quigley wrote: > >> I am not an expert (and this might be wrong), but I'm pretty sure mouse >> movements cause hardware interrupts which the CPU and therefore the kernel >> have no choice but to wake up for. >> > > I'm sure that there is no choice but to wake up, otherwise the mouse pointer > wouldn't move. The question is if it needs to wake up as much as 450 times > per > second. I'm no expert either, but I don't think it's a hardware problem. I > guess that the software could use a lower resolution and just wake up 20-40 > times per second which could be enough for most tasks... but what do I know, > that was my original question anyway. > > >> I think hardware would have to be changed for you to really get any better >> results with the mouse movements. Perhaps some sort of mouse-GPU tie-in >> and hardware accelerated mouse movements that don't require the CPU. >> >> I also don't think it should be affecting battery life very much because >> usually when you move the mouse, you will click on something or move over >> some object and it will need to wake up the CPU anyway. >> > > Even with that huge amount of wakeups, moving the mouse doesn't cause any > significant CPU load (much less I/O), so probably it doesn't waste too much > power. But still, I wonder if all the efforts made in lots of programs out > there to minimize those wakeups aren't being spoiled by this mouse thing. > It's not really relevant how much CPU usage there is when moving the mouse, but the fact that the CPU is raised out of a sleep state for no useful work - I'm sure the mouse sensitivity could be turned down for little or no noticeable effect for normal desktop or laptop usage.
As raised before, the question is whether you can modify the hardware to stop generating interrupts as often. I believe old serial mice simply sent data at a constant rate, and an interrupt was generated by the serial controller when a byte was received. I'm unsure how USB mice work. It looks like PS/2 mice had to set the sampling rate when the driver was initialized, so it bodes well that this is software configurable for newer mice as well. Some more research will tell... > >> Of course I might be wrong.... >> -Bryan >> > > Thanks, > Alberto. > > _______________________________________________ > Discuss mailing list > [email protected] > http://mail.lesswatts.org/mailman/listinfo/discuss > _______________________________________________ Discuss mailing list [email protected] http://mail.lesswatts.org/mailman/listinfo/discuss
