Subject: Re: NFS root
To: None <hubert.feyrer@informatik.fh-regensburg.de>
From: Marcus Comstedt <marcus@roxen.com>
List: port-dreamcast
Date: 02/26/2001 19:40:16
>>>>> "Hubert" == Hubert Feyrer <hubert.feyrer@informatik.fh-regensburg.de> writes:

  Hubert> On Mon, 26 Feb 2001, Marcus Comstedt wrote:
  >> I think I've said this before, but I'll say it again:  If you have a
  >> 100Mbps network, you need to change the define NFS_RSIZE in
  >> sys/nfs/nfs.h from 32768 to 16384, or NFS will not work.

  Hubert> Will a value of 16k work on 10MBd NFS?

Yes.  Lower is always safer, but may degrade performance.


  Hubert> If so, could it be made the default, and leave 32k as performance tuning
  Hubert> for 100MBd ethernet, to make things work out of the box?

The problem is that the default value is global for all ports and
NIC:s, and it would probably not be appreciated by people who has
hardware that copes with the 32k buffer if we change it globally.  The
right way to solve this is probably to make the default buffer size
into a kernel parameter.


  // Marcus