On Mon, Mar 22, 1999 at 05:07:00PM -0800, Jason Thorpe wrote:
> On Mon, 22 Mar 1999 12:48:21 -0800 
>  Jason Thorpe <thorpej@nas.nasa.gov> wrote:
> 
>  >  > That is, what units is this in?  I'm guessing 2048 (cluster size)
>  >  > bytes.
>  > 
>  > No, "clusters" in this sense is the historic "page cluster", i.e.
>  > VM page size (as opposed to hardware page size :-)  In the NetBSD
>  > kernel, CLBYTES == PAGE_SIZE :-)
>  > 
>  > Confusing enough yet? :-)
>  > 
>  > So, on i386, 4k.  On Alpha, 8k.  On SPARC, 8k.  Etc.
> 
> BTW, after reexamination of the code, NMBCLUSTERS does in fact mean
> "number of mbuf clusters", not "number of VM pages in map".  I have
> clarified this in options(4), and fixed the ports which defined it
> incorrectly (the PowerPC ports).

Side note: the UVM thesis says that if it is upped to the VM page size,
network drivers could be changed to benefit from page loadout etc.

But I guess THIS is nothing which will be done in time for 1.4.

	-is