Re: SMP: bind process to cpu

From: Christoph Hellwig (hch@caldera.de)
Date: Sat Feb 17 2001 - 12:47:56 EST

Next message: Douglas Gilbert: "Re: Flushing buffer and page cache"
Previous message: brian@worldcontrol.com: "Re: XOR [ was: Linux stifles innovation... ]"
In reply to: Manfred Spraul: "Re: SMP: bind process to cpu"
Next in thread: Andrew Morton: "Re: SMP: bind process to cpu"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

[Nick, I've added you to the Cc list so you can look at
it for future versions of your patch]

On Sat, Feb 17, 2001 at 03:13:45PM +0100, Manfred Spraul wrote:
> You must also update wake_process_synchroneous(), otherwise you can get
> lost wakeups with pipes.
>
> Something like
>
> > if (!(p->cpus_allowed & (1 << smp_processor_id()))
> > reschedule_idle(p);
>
> must be added after add_to_runqueue().

Ok.

> Ingo Molnar did some testing with tux2, and under high load wakeups were
> lost without such a patch.

(s/tux2/tux/ I suppose)

Yepp - but tux is again not userspace...

Christoph

-- 
Of course it doesn't work. We've performed a software upgrade.
-
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/

Next message: Douglas Gilbert: "Re: Flushing buffer and page cache"
Previous message: brian@worldcontrol.com: "Re: XOR [ was: Linux stifles innovation... ]"
In reply to: Manfred Spraul: "Re: SMP: bind process to cpu"
Next in thread: Andrew Morton: "Re: SMP: bind process to cpu"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

This archive was generated by hypermail 2b29 : Fri Feb 23 2001 - 21:00:15 EST