checksumming benchmark

[Ted Matsumura]

Folks,
  Our webmaster is out for a day or two more, but please continue to help me
refine the Linux Resources Web page draft which is temporarily residing at:

www.tedm.com/linux.htm <http://www.tedm.com/linux.htm> 

Also - curious if anyone has done the sw correction for tcp checksum offload
as suggested below?

Ted

*	Ted Matsumura, Product Manager tedm@packetengines.com
<mailto:tedm@packetengines.com> 
*	Packet Engines, Inc., Box 14497, Spokane, WA  99214
*	Phone: 509-777-7327, FAX: 509-777-7006
*	N 47'39.435 W 117'14.882


		-----Original Message-----
		From:	pmonta@halibut.imedia.com
[mailto:pmonta@halibut.imedia.com]
		Sent:	Friday, February 19, 1999 7:03 PM
		To:	yellowfin@cesdis1.gsfc.nasa.gov
		Subject:	checksumming benchmark

		I was curious about how much hardware checksumming would
help
		(but too lazy to do the actual work).  So this trivial
patch,
		as suggested by Donald's comment,

					/* Note: checksum -> skb->ip_summed
= CHECKSUM_UNNECESSARY; */
		+			skb->ip_summed =
CHECKSUM_UNNECESSARY;

		increases TCP performance from 378 Mbit/sec to 528 Mbit/sec
(PII/350,
		2.2.2pre).  I suppose the boost would decrease fractionally
when the
		driver actually does the right thing.

		Cheers,
		Peter Monta   pmonta@imedia.com
		Imedia Corp.
		
 | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the
 |  body of the mail, include only the text:
 |   unsubscribe this-list-name youraddress@wherever.org
 | You will be unsubscribed as speedily as possible.