From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757720Ab0ELUwc (ORCPT ); Wed, 12 May 2010 16:52:32 -0400 Received: from mail.openrapids.net ([64.15.138.104]:40367 "EHLO blackscsi.openrapids.net" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1757632Ab0ELUw1 (ORCPT ); Wed, 12 May 2010 16:52:27 -0400 Date: Wed, 12 May 2010 16:52:25 -0400 From: Mathieu Desnoyers To: Peter Zijlstra Cc: Frederic Weisbecker , Steven Rostedt , Pierre Tardy , Ingo Molnar , Arnaldo Carvalho de Melo , Tom Zanussi , Paul Mackerras , linux-kernel@vger.kernel.org, arjan@infradead.org, ziga.mahkovec@gmail.com, davem Subject: Re: Perf and ftrace [was Re: PyTimechart] Message-ID: <20100512205225.GA31848@Krystal> References: <20100512170734.GA15953@Krystal> <1273686425.1626.142.camel@laptop> <20100512175305.GB32496@Krystal> <1273687212.1626.147.camel@laptop> <20100512180438.GE15953@Krystal> <1273687712.1626.151.camel@laptop> <20100512183704.GD21432@Krystal> <1273690142.1626.158.camel@laptop> <20100512185111.GH21432@Krystal> <1273690878.1626.165.camel@laptop> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1273690878.1626.165.camel@laptop> X-Editor: vi X-Info: http://www.efficios.com X-Operating-System: Linux/2.6.26-2-686 (i686) X-Uptime: 16:49:37 up 109 days, 23:26, 4 users, load average: 0.24, 0.24, 0.19 User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org * Peter Zijlstra (peterz@infradead.org) wrote: > On Wed, 2010-05-12 at 14:51 -0400, Mathieu Desnoyers wrote: > > > > It makes it hard to use splice() or read() if you don't specify the buffer size > > at creation time. That alone seems like a pretty good argument for fixing the > > size before the mmap() call. > > Our read() doesn't actually provide the same data splice()/mmap() would. > But yeah, hence my proposal to create a separate syscall that creates > buffer objects. > Well, it's a bit strange to have read() vs splice() outputting different data, isn't it ? I would think it's rather more natural to have read() and splice() implementations outputting the exact same data, and mmap() just mmapping the ring buffer in userspace. If you really want to export that information in text format, then a different file descriptor could be used for that, with read() and splice() support. Thanks, Mathieu -- Mathieu Desnoyers Operating System Efficiency R&D Consultant EfficiOS Inc. http://www.efficios.com