Top wakeups causers

[Jim Gettys]

XRes (the X Resource Extension) could use some TLC, if someone is
looking for a nice little project.

The other major X to kernel information interface that would be good to
have is to allow X to charge back pixmap memory usage to the process
that requested the usage.  Right now, unless a user is intelligent
enough to run xrestop or equivalent, they have no insight into their
actual memory usage.
                                    - Jim


On Mon, 2007-05-14 at 11:52 -0400, Adam Jackson wrote:
> On Sun, 2007-05-13 at 17:30 +0300, Aaron Baff wrote:
> 
> > And below all of those is Xorg, which seems to be split between
> > do_setitimer (it_real_fn) and nv_start_rc_timer (nv_kern_rc_timer)
> > functions.
> 
> This is going to come up a lot, so, let's get this one out of the way.
> 
> X uses setitimer() when it's actively servicing requests from clients.
> It schedules an alarm for ~20ms in the future so that it can know when a
> significant amount of time has elapsed, and therefore schedule other
> clients if they're ready.  So at worst, it'd wake up 50 times a second,
> but it'll only do so when it's already busy doing work.
> 
> Getting better accounting information out of X is definitely a goal, but
> right now no one's working on it.  Until then you have to work backwards
> from the other wakeup events to figure out what's waking up your X
> server.
> 
> - ajax
> 
> _______________________________________________
> Power mailing list
> Power at bughost.org
> http://www.bughost.org/mailman/listinfo/power
-- 
Jim Gettys
One Laptop Per Child