> Then I made a second test with running a Whetstone benchmark in background.
> The problem is less frequent here (maybe 2 or 3 flists out of 10), but it
> can still be reproduced! So I hope it has nothing to do with DMA or a disk
I was doing some tests running 'openssl speed' and would get a few
> For comparison: I was running on an A3000 with CSPPC 68060/50, 128MB, using
> a disk attached to cbiiisc(4).
I'm currently running an A4000 with CyberStorm III.
I also did some tests doing a du -s and dd on the IDE drive. That also
causes the same types of errors. The IDE driver doesn't use DMA, so it
makes it less likely to be DMA problems (although my source tree was still
on the SCSI drive).
I've also seen the same thing on a 4.0 system, and I even had a 3.0
system with a 3.1 kernel - which does the same thing.
I was thinking this morning that the list creation process is probably
using pipes; after find a pipe(2) problem on vax a while back, that could
potentially be a source of the problem. Also, to totally eliminate the
possiblity of DMA and/or SCSI driver problems, I'll try putting enough of
the tree on the IDE drive and run entirely off the IDE drive.
Michael L. Hitch mhitch@... Computer Consultant
Information Technology Center
Montana State University Bozeman, MT USA