Subject: Re: packet capturing
To: Jonathan Stone <jonathan@dsg.stanford.edu>
From: Jason Thorpe <thorpej@wasabisystems.com>
List: tech-kern
Date: 01/21/2004 14:15:15
--Apple-Mail-4--220649061
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=US-ASCII; format=flowed


On Jan 21, 2004, at 1:02 PM, Jonathan Stone wrote:

> Ask Jason.  I know FreeBSD's em driver does; I have packet traces of
> the Intel-supplied FreeBSD em driver in FreeBSD 4.x slurping in a
> 64k-byte SMB packet train with one (1) interrupt at the end of the 
> train.

NetBSD's "wm" driver also uses interrupt mitigation features of the 
Intel Gig-E chip.  In addition to the normal "interrupt or not" bit in 
the Rx descriptor, there's also an Rx interrupt mitigation timer that 
inserts a delay of the specified time between deciding that an 
interrupt is to be delivered and actually delivering it, thus giving 
more time for additional packets to arrive.

It certainly might be a good exercise to tune the timers, or add other 
mitigation techniques, but some mitigation is currently being done.

         -- Jason R. Thorpe <thorpej@wasabisystems.com>


--Apple-Mail-4--220649061
content-type: application/pgp-signature; x-mac-type=70674453;
	name=PGP.sig
content-description: This is a digitally signed message part
content-disposition: inline; filename=PGP.sig
content-transfer-encoding: 7bit

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (Darwin)

iD8DBQFADvnzOpVKkaBm8XkRAkroAKDVN8tzzqYQd4d+8l8fJOK1gxL52gCdEHvt
F5i7bKYyYNXHp9Rvf5aMARQ=
=lCV9
-----END PGP SIGNATURE-----

--Apple-Mail-4--220649061--