Re: [patch] Real-Time Preemption, -RT-2.6.12-rc6-V0.7.48-00

From: Ingo Molnar
Date: Wed Jun 22 2005 - 05:18:02 EST

* Ingo Molnar <mingo@xxxxxxx> wrote:

> * William Weston <weston@xxxxxxxxx> wrote:
> > Attached are two typical traces and the .config from my Xeon/HT box,
> > currently running -50-06 with a normal desktop workload (X, wmaker,
> > ten dockapps, several xterms, and firefox).
> the second trace seems to be a cross-CPU wakeup bug. It's not
> completely clear from the trace what happened - but we measured the
> latency of a task (wmcube-3191), where the wakeup happened on CPU#0
> and wmcube-3191 was queued to CPU#1 which was idle at that time. The
> bug is that it wasnt until timestamp 306us that this actually happened
> - and CPU#1 was just idling around in default_idle() for no good
> reason. CPU#1 should have run wmcube-3191 at around timestamp 13us.

ok, managed to reproduce it on a HT box, and it turned out to be a bug
in the SMT scheduler: the dependent sleeper logic incorrectly delayed
high-prio tasks, causing these latencies. I fixed it in the -50-10
kernel - could you redo your tests with that kernel (or later versions)?

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at