From mboxrd@z Thu Jan 1 00:00:00 1970 From: "David S. Miller" Subject: Re: RFC: NAPI packet weighting patch Date: Thu, 09 Jun 2005 15:22:28 -0700 (PDT) Message-ID: <20050609.152228.112623409.davem@davemloft.net> References: <1118237775.6382.34.camel@localhost.localdomain> <20050609150546.61b0fee7@unknown-215.office.pdx.osdl.net> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: jesse.brandeburg@intel.com, hadi@cyberus.ca, john.ronciak@intel.com, mitch.a.williams@intel.com, mchan@broadcom.com, buytenh@wantstofly.org, jdmason@us.ibm.com, netdev@oss.sgi.com, Robert.Olsson@data.slu.se, ganesh.venkatesan@intel.com Return-path: To: shemminger@osdl.org In-Reply-To: <20050609150546.61b0fee7@unknown-215.office.pdx.osdl.net> Sender: netdev-bounce@oss.sgi.com Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org From: Stephen Hemminger Date: Thu, 9 Jun 2005 15:05:46 -0700 > > I have profile data, here is an example of 5tx/5rx threads, where the > > throughput was 1236Mb/s total, 936tx, 300rx, on 2.6.12-rc5 with old TSO > > (the original problem case) we are at 100% cpu and generating 3289 ints/s, > > with no hardware drops reported prolly due to my replenish patch > > CPU: P4 / Xeon with 2 hyper-threads, speed 2791.36 MHz (estimated) > > Counted GLOBAL_POWER_EVENTS events (time during which processor is not stopped) with a unit mask of 0x01 (mandatory) count 100000 > > samples % image name symbol name > > 533687 8.1472 vmlinux pskb_expand_head > > 428726 6.5449 vmlinux __copy_user_zeroing_intel > > 349934 5.3421 vmlinux _read_lock_irqsave > > We should kill all reader/writer locks in the fastpath. reader locks are > more expensive than spinlocks unless they are going to be held for a fairly > large window. True, but I see no reason why it should have any influence here. Let's not get distracted by this in our analysis of the problem.