Re: [PATCH] add scheduler domains for ia64

From: Jesse Barnes <jbarnes_at_engr.sgi.com>
Date: 2004-08-20 12:22:35
On Thursday, August 19, 2004 10:11 pm, Nick Piggin wrote:
> Sorry I haven't replied earlier.  I think this looks good, provided
> it does the right thing for you (I can't test it myself). Send it to
> Andrew to get merged if you'd like.

Yep, it's been working ok so far.  There's still more we can do, but this is a 
good start I think.  Andrew, this version applies on top of 2.6.8.1-mm2 but 
overwrites most of the earlier node-span patch by moving bits from arch/ia64 
to kernel/sched.c, so let me know if you want the patch in a different 
format.

This patch adds some more NUMA specific logic to the creation of scheduler 
domains.  Domains spanning all CPUs in a large system are too large to 
schedule across efficiently, leading to livelocks and inordinate amounts of 
time being spent in scheduler routines.  With this patch applied, the node 
scheduling domains for NUMA platforms will only contain a specified number of 
nearby CPUs, based on the value of SD_NODES_PER_DOMAIN.  It also allows 
arches to override SD_NODE_INIT, which sets the domain scheduling parameters 
for each node's domain.  This is necessary especially for large systems.

Possible future directions:
  o multilevel node hierarchy (e.g. node domains could contain 4 nodes worth 
of CPUs, supernode domains could contain 32 nodes worth, etc. each with their 
own SD_NODE_INIT values)
  o more tweaking of SD_NODE_INIT values for good load balancing vs. overhead 
tradeoffs

Signed-off-by: Jesse Barnes <jbarnes@sgi.com>

Thanks,
Jesse

-
To unsubscribe from this list: send the line "unsubscribe linux-ia64" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Received on Thu Aug 19 22:24:22 2004

This archive was generated by hypermail 2.1.8 : 2005-08-02 09:20:29 EST