1 |
On Mi, Jun 29 2005 at 10:13:10 -0400, Jason Cooper wrote: |
2 |
> > Now _that_ is strange: I recently changed that back from 1000 to 100 but |
3 |
> > with no visible (or audible that is) effect whatsoever. |
4 |
> > |
5 |
> |
6 |
> This is correct behavior. The value of HZ defines the minimum |
7 |
> resolution of system calls like usleep() and select(). At 100Hz, the |
8 |
> minimum resolution is ~20ms (= 2 * 1 / f) for usleep and ~10ms (= 1 / f) |
9 |
> for select(). Setting it to 1000Hz adjusts those values down to ~2ms |
10 |
> and ~1ms respectively. I've only found this useful with old/half-assed |
11 |
> code that polls interfaces. They'll use usleep() to throttle the |
12 |
> polling and not max out the CPU. |
13 |
> |
14 |
> But that is background info, the main point is that it has no effect on |
15 |
> power consumption. |
16 |
> |
17 |
> Do your system monitors (gkrellm, top, whatever) show a maxed out cpu? |
18 |
|
19 |
Nope. There's no indication that there is anything going on. The system |
20 |
load is flat at 0.00 and the cpu seems to be idle according to top. |
21 |
Also i tried to stop all daemons (even those i'd rather not turn off |
22 |
under every-day circumstances) so that there is hardly more than init |
23 |
and a few getties running (not even X) and still the fan would go off |
24 |
after a few minutes and never stop 'til i shut down the computer. |
25 |
|
26 |
Alexander |