From owner-freebsd-hackers  Fri Sep 12 18:15:02 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id SAA23302
          for hackers-outgoing; Fri, 12 Sep 1997 18:15:02 -0700 (PDT)
Received: from pat.idi.ntnu.no (0@pat.idi.ntnu.no [129.241.103.5])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id SAA23286
          for <hackers@FreeBSD.ORG>; Fri, 12 Sep 1997 18:14:58 -0700 (PDT)
Received: from idt.unit.no (tegge@ikke.idi.ntnu.no [129.241.111.65])
	by pat.idi.ntnu.no (8.8.6/8.8.6) with ESMTP id DAA01842;
	Sat, 13 Sep 1997 03:14:28 +0200 (MET DST)
Message-Id: <199709130114.DAA01842@pat.idi.ntnu.no>
To: dkelly@hiwaay.net
Cc: grog@lemis.com, hackers@FreeBSD.ORG
Subject: Re: Do *you* have problems with floppies? 
In-Reply-To: Your message of "Fri, 12 Sep 1997 05:53:34 -0500"
References: <199709121053.FAA14361@nospam.hiwaay.net>
Mime-Version: 1.0
X-Mailer: Mew version 1.70 on Emacs 19.34.1
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Sat, 13 Sep 1997 03:14:28 +0200
From: Tor Egge <Tor.Egge@idi.ntnu.no>
Sender: owner-freebsd-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

> So this time I kicked off bonnie and got
> 
>  51 pci irq9	(Adaptec 2940)
>  21 fdc0 irq6
>  blks 5245	(sd0, "SEAGATE ST32550N 0021")
> 
> Processing EVEVEVEEVEVVVVEVVVVVEVVVVVVVVVVVVVVVVVVV done.
> 
> The V's were in the "Reading intelligently..." phase.

I have similar experiences when iozone is running in the background.
MB is ASUS P65UP5. AHA2940 and AHA2940UW scsi controllers.


Using fdformat gives the following output:

	Format 1440K floppy `/dev/rfd0.1440'? (y/n): y
	Processing VVVVVEVVVVVVEVVVVVVVEEVVVVVVVVVVVVVVVVVV done.

with errors as:

fd0c: hard error reading fsbn 420 of 414-431 (ST0 44<abnrml,top_head> ST1 4<sec_not_fnd> ST2 0 cyl 11 hd 1 sec 7)

Aborting iozone and running fdformat -v (only verify) reproduces the
same errors, indicating that bad contents was written to some of the
sector headers on the disk during the format operation.

with no scsi bus traffic, I was able to run fdformat 10 times in a row
without encountering any errors. 

Disabling interrupts in fd_cmd and fd_read_status did not help. This
indicates that `busy waiting' will not help.

It looks very much like a DMA conflict or starvation. 

- Tor Egge