Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree

On Tue, 7 Nov 2006, Siddha, Suresh B wrote:

> tasklet_schedule doesn't schedule if there is already one scheduled.

Correct. The effect of this is to only allow load balancing on one cpu at a 
time. Subsequent attempts to schedule load balancing on other cpus are 
ignored until the cpu that is load balancing has finished. Then the 
others (hopefully) get a turn.

A pretty interesting unintended effect. It certainly solves the concurrent 
load balancing scalability issues and would avoid the need to stagger 
load balancing. Wonder how fair it would be?

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

References:
- Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
  - From: Ingo Molnar <[email protected]>
- Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
  - From: Christoph Lameter <[email protected]>
- Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
  - From: "Siddha, Suresh B" <[email protected]>
- Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
  - From: Christoph Lameter <[email protected]>
- Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
  - From: "Siddha, Suresh B" <[email protected]>

Prev by Date: 2.6.19-rc1: Volanomark slowdown
Next by Date: Re: [PATCH 12/12] cciss: fix for iostat
Previous by thread: Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
Next by thread: Re: + sched-use-tasklet-to-call-balancing.patch added to -mm tree
Index(es):
- Date
- Thread

[Index of Archives] [Kernel Newbies] [Netfilter] [Bugtraq] [Photo] [Stuff] [Gimp] [Yosemite News] [MIPS Linux] [ARM Linux] [Linux Security] [Linux RAID] [Video 4 Linux] [Linux for the blind] [Linux Resources]