From owner-freebsd-hackers  Thu May 29 10:34:12 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id KAA27423
          for hackers-outgoing; Thu, 29 May 1997 10:34:12 -0700 (PDT)
Received: from inetfw.sonycsl.co.jp (inetfw.sonycsl.co.jp [203.137.129.4])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA27416
          for <freebsd-hackers@FreeBSD.ORG>; Thu, 29 May 1997 10:34:10 -0700 (PDT)
Received: from hotaka.csl.sony.co.jp (hotaka.csl.sony.co.jp [43.27.98.57]) by inetfw.sonycsl.co.jp (8.8.3/3.5W) with ESMTP id RAA22107; Thu, 29 May 1997 17:33:46 GMT
Received: from localhost (localhost [127.0.0.1]) by hotaka.csl.sony.co.jp (8.8.4/3.3W3) with ESMTP id CAA20576; Fri, 30 May 1997 02:33:30 +0900 (JST)
Message-Id: <199705291733.CAA20576@hotaka.csl.sony.co.jp>
To: freebsd-hackers@FreeBSD.ORG
cc: Chuck Cranor <chuck@maria.wustl.edu>
Subject: misaligned access to EDO RAM
Date: Fri, 30 May 1997 02:33:30 +0900
From: Kenjiro Cho <kjc@csl.sony.co.jp>
Sender: owner-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk


Does anyone know if it is OK to perform misaligned-burst-access to EDO 
RAM?

The DMA unit of an ENI ATM PCI card seems to lock up when it performs
misaligned-burst-access crossing 1KB boundary. 
(e.g. 64-byte-burst from 0xf02663c4 but no problem with 512B boundary
e.g. from 0xf02661c4)

It happens with a machine with ECC EDO DIMM (dell optiplex GXpro).
The lockup doesn't occur with other machines with fastpage DRAM
(all P6-200s with Natoma).

I suspect the 1K boundary comes from the EDO internal structure, but
I'm not sure if the specs (EDO, Natoma, PCI) allow to do that.

Any idea?  Am I missing something?

--kj