Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 17 Jul 2007 15:17:39 +0200
From:      Gary Jennejohn <gary.jennejohn@freenet.de>
To:        Jeff Roberson <jroberson@chesapeake.net>
Cc:        current@freebsd.org
Subject:   Re: ULE/SCHED_SMP diff for 7.0
Message-ID:  <20070717151739.c292d3fe.gary.jennejohn@freenet.de>
In-Reply-To: <20070716233030.D92541@10.0.0.1>
References:  <20070716233030.D92541@10.0.0.1>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 16 Jul 2007 23:35:38 -0700 (PDT)
Jeff Roberson <jroberson@chesapeake.net> wrote:

> http://people.freebsd.org/~jeff/ule.diff
> 
> This patch is scheduled for inclusion in 7.0.  I would like anyone
> who cares to run it to validate that it does not create any stability
> or performance regression over the existing ULE.  This patch replaces
> ULE with SCHED_SMP, which will now no longer exist as a seperate fork
> of ULE.
> 
> Briefly, this is still a very suitable scheduler for uniprocessor
> machines while providing stronger affinity and other performance
> improvements for multiprocessor machines.
> 
> Even "works for me!" type responses are welcome so I know roughly how
> many people have tested before I commit this close to release.
> 

This is definitely an improvement over the old ULE.

With this: CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ and
64-bit FreeBSD-current I did ``make -j4 buildworld'' and saw that CPU0
isn't idle nearly as much as it was with the previous version of ULE.
CPU1 was hardly ever idle, but CPU0 was ususally at 5-10%, whereas
before it was more like 20%.

BTW /usr/obj is on a ZFS share. Can't say whether that has any
relevance.

--- 
Gary Jennejohn



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070717151739.c292d3fe.gary.jennejohn>