On Fri, 4 Jul 1997, Jason Thorpe wrote:

> On Fri, 4 Jul 1997 09:57:18 -0400 (EDT) 
>  "Steven N. Hirsch" <shirsch@ibm.net> wrote:
> 
>  > 1) Why is disk I/O so dog slow?  Iozone shows that the best my Seagate
>  > ST1401N can do is ~350 bytes/sec RW!  This drive can do 1200 bytes/sec. on
>  > X86 linux, and that is on a 386/40 with a slow controller.
> 
> Part of this, undoubtedly, is the SCSI controller hardware in the Macintosh.
> It lacks a DMA engine.  Your Linux 386 system probably has a controller
> that can do DMA.

So I've learned!  Urk.  I had it in my head that Apple were smarter than
that by the time the SE/30 was designed.  Presumably, later vintage M68K
Macs had more sophisticated hardware?

>  > 2) I'm able to mount exported NFS directories from a Linux box by hand,
>  > but amd fails with a bitter complaint from the server about "..insecure
>  > port 1037 - seek mental counseling.."  (I'm serious).  Any insights?
> 
> Err... you're a little ambiguous here... the Linux server prints that
> on the console?  Try telling amd to use a priviledged port when it
> performs the mount.

Oh, yes, I'm not making that up <g>.  I'm not sure why the BSD version of
amd is using an oddball port for the NFS mount request - I _thought_ the
Linux amd was the same code (it's never caused a problem).  

Here's the Linux signon:

Copyright (c) 1990 Jan-Simon Pendry
Copyright (c) 1990 Imperial College of Science, Technology & Medicine
Copyright (c) 1990 The Regents of the University of California.
Unofficial patch level 102.
amd 5.2.2.2 of 1992/05/31 16:53:21 bsd44-beta #0: Sat Feb 22 17:56:24 EST 1997
Built by hirsch@air for an i386 running linux version 2.0 (little-endian).
Map support for: root, passwd, union, nis, ndbm, file, error.
FS: ufs, nfs, nfsx, host, link, linkx, program, union, auto, direct, 
    toplvl, error.

Nothing relative to default port, nor does the man page even indicate an
option to change it..  I'll read the sources later on and delve into this
deeper.

Steve