Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964962Ab2EQMLw (ORCPT ); Thu, 17 May 2012 08:11:52 -0400 Received: from hrndva-omtalb.mail.rr.com ([71.74.56.122]:18251 "EHLO hrndva-omtalb.mail.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754098Ab2EQMLt (ORCPT ); Thu, 17 May 2012 08:11:49 -0400 X-Authority-Analysis: v=2.0 cv=cssZYiEi c=1 sm=0 a=ZycB6UtQUfgMyuk2+PxD7w==:17 a=XQbtiDEiEegA:10 a=XlIr_nCh6zIA:10 a=5SG0PmZfjMsA:10 a=Q9fys5e9bTEA:10 a=n7lzCUN1pRk5MbKrU8gA:9 a=mKWEfrWPLEBYV7PoiZQA:7 a=PUjeQqilurYA:10 a=ZycB6UtQUfgMyuk2+PxD7w==:117 X-Cloudmark-Score: 0 X-Originating-IP: 74.67.80.29 Message-ID: <1337256707.6724.101.camel@gandalf.stny.rr.com> Subject: Re: [RFC][PATCH] tracing: Remove useless 4 bytes of padding from every event From: Steven Rostedt To: Ingo Molnar Cc: Arjan van de Ven , Dave Jones , Linus Torvalds , LKML , Ingo Molnar , Frederic Weisbecker , David Sharp , Vaibhav Nagarnaik , Peter Zijlstra , Andrew Morton Date: Thu, 17 May 2012 08:11:47 -0400 In-Reply-To: <20120517080651.GC17419@gmail.com> References: <1337175871.6724.46.camel@gandalf.stny.rr.com> <1337198404.6724.81.camel@gandalf.stny.rr.com> <4FB408F2.8060002@linux.intel.com> <20120516201431.GA29661@redhat.com> <4FB40B2B.703@linux.intel.com> <20120517080651.GC17419@gmail.com> Content-Type: text/plain; charset="ISO-8859-15" X-Mailer: Evolution 3.2.2-1 Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1409 Lines: 33 On Thu, 2012-05-17 at 10:06 +0200, Ingo Molnar wrote: > Nor should we waste too much time over these 4 bytes really. Is > the kernel really in such a good shape that we must spend our > time trying to break working apps, over a mostly cosmetic detail > in an ABI which will soon be messed up with our next set up > mistakes anyway? ;-) > 4 bytes is not cosmetic for a 32 byte event. That's 1/8th overhead. If we could get rid of 4 bytes from struct page, would we do that? It's only just 4 bytes for ever 4096 bytes. Just a 1/1024 overhead. Of course perf events are much bigger than 32 bytes. It's one of the biggest complaints I hear about perf, the size of its events. We should be trying hard to fix that. And we are not spending time trying to break tools. The tools were already broken. The main motivator for getting parse-events into powertop was that the older version (due to this binary interface) could not be used on a system running a 32bit userspace on a 64bit kernel. With the proper parsing, it not only was able to run on a 32/64 environment, it can also parse the events like perf does and not need direct padding of space. -- Steve -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/