From owner-freebsd-hackers  Wed May  1 11:27:43 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id LAA06566
          for hackers-outgoing; Wed, 1 May 1996 11:27:43 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id LAA06560
          Wed, 1 May 1996 11:27:37 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id EAA05707; Thu, 2 May 1996 04:24:59 +1000
Date: Thu, 2 May 1996 04:24:59 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199605011824.EAA05707@godzilla.zeta.org.au>
To: luigi@labinfo.iet.unipi.it, terry@lambert.org
Subject: Re: lmbench IDE anomaly
Cc: current@freebsd.org, hackers@freebsd.org, koshy@india.hp.com,
        msmith@atrad.adelaide.edu.au
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>> The disksort stuff is pessimal.  Contact mday@elbereth.org for details.

>The original poster was doing a couple of lmdd from /dev/rwd0a, the raw
>device. Does disksort get in the way in this case ?

Of course not.

>BTW, running a number of dd on /dev/wd0a works *much* better, you get
>almost n times the bandwidth at least with up to 5 instances, what I
>tried.

/dev/wd0a is a completely different device from /dev/rwd0a.  dd on it
tests the buffer cache and the too-small block size of BLKDEV_IOSIZE
= 2048.

Bruce