Subject: Re: New hash algorithm - FNV - for use in the kernel
To: David Laight <David.Laight@btinternet.com>
From: Bill Studenmund <wrstuden@netbsd.org>
List: tech-perform
Date: 11/27/2001 16:48:27
On Wed, 28 Nov 2001, David Laight wrote:

> >  > But there is no reason why it shouldn't optimise for what it expects to
> >  > see most often....
> >
> > And what would that be?  Solaris?  NetBSD?  Tru64?
>
> This is NFS, all the implementations are clones of each other,
> I'd guess that there is little difference between the implementations.

At least the way the BSD NFS implementation works, the file handle is
supplied by the underlying file system. I'd expect other systems are the
same. The problem is that you (well such an optimization) assume(s) that
the file handles are all the same. While NFS may be similar, think about
file systems out there with 64-bit inode numbers. Opaque file handles deal
wiht that, but if we don't include all the file handle, we can be ignoring
parts of the handle which will vary significantly.

Take care,

Bill