lkml.org 
[lkml]   [2004]   [Sep]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] Yielding processor resources during lock contention

> OK that sounds alight.
>
> If it is a "I'm going to be spinning for ages, so schedule me off *now*"
> then I think it is wrong... but if it just allows you to keep the hypervisor
> preemption turned on, then fine.

The hypervisor can preempt us anywhere but we can give it clues so it
tends to preempt us at better times.

Within a few instructions the hypervisor should be able to answer the
big question: is the cpu Im waiting for currently scheduled on another
cpu. If it is then it might be best for the hypervisor to leave us
alone.

If it isnt then the spin time will most likely be dominated by how
long it takes to get that cpu scheduled again and the hypervisor should
make an effort to schedule it.

Anton
-
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/

\
 
 \ /
  Last update: 2005-03-22 14:06    [W:0.119 / U:0.876 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site