Messages in this thread | | | Date | Sun, 12 May 2002 18:57:09 -0400 | From | Kurt Wall <> | Subject | Re: Question about cpu time accuracy. |
| |
Scribbling feverishly on May 12, Jennifer Huang managed to emit: > Hi all, > > I have a question about cpu time accuracy. > > I am using kernel 2.4.18. But, when I tried "utime" > and "nanosleep" to get a process suspended, it only > worked in 10ms granularity, and it's no way to sleep > for 1 microsecond.
The standard kernel timer has a resolution of 1/HZ, which is 10ms on an x86. You could try a scheduling policy of SCHED_FIFO or SCHED_RR, but this only gets you 2ms resolution.
> Anyone can help me out of this?
There are patches available for high resolution timers:
http://sourceforge.net/projects/high-res-timers/ http://www.cs.wisc.edu/paradyn/libhrtime/
Kurt -- Anarchy may not be the best form of government, but it's better than no government at all. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |