Messages in this thread | | | Date | Tue, 5 Jun 2007 17:14:12 +0200 | From | "Dmitry Adamushko" <> | Subject | Re: Interesting interaction between lguest and CFS |
| |
On 05/06/07, Matt Mackall <mpm@selenic.com> wrote: > > [...] > > Click into the lguest window and trigger the delay. > > I did: > > while true; do sleep 1; cat /proc/sched_debug > sched_debug.txt; done > > and got this, hopefully inside the window: > > Sched Debug Version: v0.02 > now at 257428593818894 nsecs > > cpu: 0 > .nr_running : 3 > .raw_weighted_load : 2063 > .nr_switches : 242830075 > .nr_load_updates : 30172063 > .nr_uninterruptible : 0 > .jiffies : 64282148 > .next_balance : 0 > .curr->pid : 27182 > .clock : 125650217819008823 > .prev_clock_raw : 257428516403535
The delta (clock - prev_clock_raw) looks insane.
The current time (which doesn't depend on rq_clock() --> sched_clock() is " now at 257428593818894 nsecs " (right at the beginning of the output).
'prev_clock_raw' is updated any time rq_clock() is called - basically upon any scheduling operation (e.g. enqueue/dequeue)
now - prev_clock_raw == 257428593818894 - 257428516403535 == ~ 80 ms.
while 'clock' reports something crazy.. that would explain why there wes a huge "block_max" reported earlier.. I guess, sched_clock() is tsc-based in your case?
Any way to get it switched to jiffies-based one and repeat the test?
-- Best regards, Dmitry Adamushko - 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/
| |