Subject: Re: vm problems
To: Matthias Pfaller <leo@dachau.marco.de>
From: Sverre Froyen <sverre@sst.nrel.gov>
List: port-pc532
Date: 08/14/1995 08:06:15
Matthias Pfaller writes:
> 
> The ld.so copied with msave did differ from the original ld.so within
> one page. From looking at the damaged page it looks like the page was
> a gcc data page. ...
...
> Now I suspect that the problem is not in the ncr driver but somewhere
> in the pc532's vm code. Note that this is the first time that we are
> running with an interrupt driven disk driver. Any hint where to start
> searching for the problem?

Here is a me too.  Maybe :-)

For a while I have been trying to get a Syquest 270MB drive to work
with my pc532.  I have added code to sd.c to handle removable media
such as floppy drives and Syquest drives and my SCSI floppy works
fine.  The Syquest, on the other hand, does not work fine.  Here are
the symptoms.

Using "current" sources from the beginning of May, the drive would
occasionally produce bad data on read.  Writes appear to be OK.  There
is no indication from the driver that anything went wrong, you have to
inspect the resulting file to see it.  For instance,

	cp <syquest>/bin/csh /tmp/csh
	cmp <syquest>/bin/csh /tmp/csh

would say files differ.

With the present "current" sources, the errors appear MUCH more often.
For files of any length, it is virtually impossible to obtain a good
copy.  For sources from about a week ago, I would also get an occasional
kernel panic, but the panics appear to have gone away in the latest source.

My other hard drives appears to work fine (i have not noticed any problems).

I thought the problem was caused by timing problems in the ncr driver.
Note that I am not doing any memory mapping.


Sverre

-- 
Sverre Froyen (+1-303-384-6612; FAX: +1-303-384-6531)
NREL (National Renewable Energy Laboratory), Golden, CO 80401
email: sverre_froyen@nrel.gov