Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uwm.edu!cs.utexas.edu!rice!sun-spots-request
From: jeffz@sun.com (Jeffrey Zank - SBus Software)
Newsgroups: comp.sys.sun
Subject: Re: DMA clarification
Keywords: Hardware
Message-ID: <8291@brazos.Rice.edu>
Date: 29 May 90 21:59:47 GMT
Sender: root@rice.edu
Organization: Sun-Spots
Lines: 19
Approved: Sun-Spots@rice.edu
X-Refs:  Original: v9n177, Replies: v9n180 v9n185
X-Sun-Spots-Digest: Volume 9, Issue 185, message 11

| >You need to be clear here.  If you are using DVMA, 600 KBytes sounds about
| >right for mb_ampalloc() and mb_mapfree() type of usage.  Total space for
| >DVMA is about 1 MByte, and its shared with all other devices that do DVMA.
| >You really cannot "reconfigure" your kernel to add more space, you are
| >limited by other kernel resources that currently are not configurable
| >(PMEGS, and IOPB Memory for example).
| 
| Please post a correction - DVMA is a softwared fiction on the SBus
| machines - you can see all of the kernel context from the SBus. There
| is no 1MB DVMA window in hardware on that machine.

I have been asked to clarify this a bit more, so;

I do not refer to DVMA as being a hardware limit here.  I was attempting
to point out that under 4.1 you can *expect* to map up about 600kbytes
worth of DVMA (using mb_ampalloc()) and that the 1 MB is not configurable
by the user (without hacking up the kernel).  The 4/60 and 4/65 no longer
have  the hardware limits that forced the 1 MB limit.    You should look
at bp_mapin() for larger size mappings.