Messages in this thread | | | From | Peter Chubb <> | Date | Fri, 12 Sep 2003 11:48:32 +1000 | Subject | Re: [PATCH] schedstat-2.6.0-test5-A1 measuring process scheduling latency |
| |
>>>>> "Rick" == Rick Lindsley <ricklind@us.ibm.com> writes:
PeterC> An alternative is my microstate accounting patch, which lets PeterC> you do the same thing (with rather less intrusiveness) but PeterC> per-thread.
Rick> [ ... snip ... ]
PeterC> My own observations tend me to the idea that a process PeterC> waiting for disk I/O isn't awoken fast enough, at least on my PeterC> laptop.
Rick> I'm not sure that it's any less or more intrusive, but it's at Rick> least another way of doing the same thing. So since you've Rick> taken some measurements, what's the length of time you find your Rick> process waits to hit the processor after getting the I/O it Rick> needs? What's the time it seems to wait when it skips (what's Rick> the cutoff at which you hear a skip versus don't hear one?)
Graph attached, times in seconds per second. (BTW, the patch I sent was bad, because I diffed against the wrong tree. There's a new patch coming soon)
[image] | |