From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnaldo Carvalho de Melo Subject: Re: [RFCv4 PATCH 1/2] net: Introduce recvmmsg socket syscall Date: Fri, 9 Oct 2009 16:35:20 -0300 Message-ID: <20091009193520.GD12982@ghostprotocols.net> References: <20090916170738.GC7699@ghostprotocols.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Caitlin Bestler , Chris Van Hoof , Clark Williams , Neil Horman , Nir Tzachar , Nivedita Singhvi , Paul Moore , =?iso-8859-1?Q?R=E9mi?= Denis-Courmont , Steven Whitehouse , Linux Networking Development Mailing List To: David Miller Return-path: Received: from mx1.redhat.com ([209.132.183.28]:60477 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934040AbZJITgZ (ORCPT ); Fri, 9 Oct 2009 15:36:25 -0400 Content-Disposition: inline In-Reply-To: <20090916170738.GC7699@ghostprotocols.net> Sender: netdev-owner@vger.kernel.org List-ID: Em Wed, Sep 16, 2009 at 02:07:38PM -0300, Arnaldo Carvalho de Melo escreveu: > Meaning receive multiple messages, reducing the number of syscalls and > net stack entry/exit operations. > > Next patches will introduce mechanisms where protocols that want to > optimize this operation will provide an unlocked_recvmsg operation. Hi Dave, The second patch in this series has issues, I still have to investigate it properly, study removing the skb_queue_head lock like TCP does, but the first patch seems to be OK and already providing good results at least as reported by Nir, if there aren't any other concerns about the API, can we get it into net-next-2.6? Best Regards, - Arnaldo