*BSD News Article 25334


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!sgiblab!spool.mu.edu!howland.reston.ans.net!usenet.ins.cwru.edu!eff!news.kei.com!news.byu.edu!cwis.isu.edu!u.cc.utah.edu!cs.weber.edu!terry
From: terry@cs.weber.edu (A Wizard of Earth C)
Newsgroups: comp.os.386bsd.development
Subject: Re: [FreeBSD 1.0R] DMA Problems?
Date: 26 Dec 1993 22:10:34 GMT
Organization: Weber State University, Ogden, UT
Lines: 39
Message-ID: <2fl24q$jn2@u.cc.utah.edu>
References: <jmonroyCIB20s.FF8@netcom.com> <2f8kjq$ft2@u.cc.utah.edu> <jmonroyCIHJA2.oy@netcom.com>
NNTP-Posting-Host: cs.weber.edu

In article <jmonroyCIHJA2.oy@netcom.com> jmonroy@netcom.com (Jesus Monroy Jr) writes:
>: No.  No way.  Not a chance.
>:
>	No to what?.... I mentioned three things... no to what?

No, the statement that "the 'RAM refresh' process had priority over the
'FDC transfer'" is incorrect.

Thus the conclusion "the problem stemmed from the possibility that ..."
is also incorrect, as is the statement that begins "If this is true..."
and says "there is no way around this problem" is also false.

>: The main interaction between memory refresh and DMA is that you can
>: kill memory off and parity error yourself to death by holding the bus
>: so long that refreshes get skipped.  Dynamic RAM gets pissy when you
>: skip its refresh.  8-).

>	my information says that you can skip a few "RAM refresh"
>	cycles... what are you saying?

You can.  Tell me: how do you allow "skipping", but deterministically
sufficient skipping for the potential well to discharge to the point
that the memory contents are no longer reliable?  There is no hardware
"skip count"... DRAM refresh is not a realtime process which will
interrupt other things on your motherboard to safeguard your memory
contents.

How many you can skip is dependent on the chip and the bus rate and
the on-time for the refresh... in other words, how deep the potential
well is, how long it is charged on initial write, and how long it is
charged on each refresh.  And there is *no* way to probe this from
software.


					Terry Lambert
					terry@cs.weber.edu
---
Any opinions in this posting are my own and not those of my present
or previous employers.