From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Grover Subject: Re: [RFC 1/2] net: Introduce recvmmsg socket syscall Date: Wed, 3 Jun 2009 18:44:22 -0700 Message-ID: References: <20090520230652.GB5956@ghostprotocols.net> <20090522200611.GB967@hmsreliant.think-freely.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Arnaldo Carvalho de Melo , David Miller , netdev@vger.kernel.org, Chris Van Hoof , Clark Williams To: Neil Horman Return-path: Received: from mail-qy0-f189.google.com ([209.85.221.189]:62964 "EHLO mail-qy0-f189.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752354AbZFDBt2 convert rfc822-to-8bit (ORCPT ); Wed, 3 Jun 2009 21:49:28 -0400 Received: by qyk27 with SMTP id 27so723530qyk.33 for ; Wed, 03 Jun 2009 18:49:30 -0700 (PDT) In-Reply-To: <20090522200611.GB967@hmsreliant.think-freely.org> Sender: netdev-owner@vger.kernel.org List-ID: On Fri, May 22, 2009 at 1:06 PM, Neil Horman wr= ote: > Not to throw more questions into the mix again, but didn't Ingo write= a > batching syscall a while back, which let you issue several syscalls i= n one trap > to kernel space? =A0I understand that your approach has some efficien= cy gains over > that, but did that ever get accepted upstream? =A0Is the overlap ther= e sufficient > to make this approach redundant? =A0Or are the gains in performance h= ere > sufficient to warrant this new call? I couldn't find this via Google or any posts to LKML by Ingo on this? I'm very interested in multiple send/recvmsg support. So, it's definitely going to happen? :-) Regards -- Andy