From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752822Ab1ARRBT (ORCPT ); Tue, 18 Jan 2011 12:01:19 -0500 Received: from hrndva-omtalb.mail.rr.com ([71.74.56.122]:37060 "EHLO hrndva-omtalb.mail.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752530Ab1ARRBS (ORCPT ); Tue, 18 Jan 2011 12:01:18 -0500 X-Authority-Analysis: v=1.1 cv=3uSaImBeuprzHBlOOPjkqgu+7PcxSRW0m2Aphm9Zmck= c=1 sm=0 a=I98iFNk1EVwA:10 a=Q9fys5e9bTEA:10 a=OPBmh+XkhLl+Enan7BmTLg==:17 a=IU6EI_ANgppSYHbGRbkA:9 a=JluYTsPGgyFOaazmsDh-zoYhleIA:4 a=PUjeQqilurYA:10 a=OPBmh+XkhLl+Enan7BmTLg==:117 X-Cloudmark-Score: 0 X-Originating-IP: 67.242.120.143 Subject: Re: [PATCH] tracing: remove superfluous sub instructions From: Steven Rostedt To: Jiri Olsa Cc: Frederic Weisbecker , mingo@elte.hu, linux-kernel@vger.kernel.org In-Reply-To: <20110118164148.GD1941@jolsa.brq.redhat.com> References: <1295364498-9288-1-git-send-email-jolsa@redhat.com> <20110118155159.GA1791@nowhere> <1295367762.12215.2.camel@gandalf.stny.rr.com> <20110118164148.GD1941@jolsa.brq.redhat.com> Content-Type: text/plain; charset="ISO-8859-15" Date: Tue, 18 Jan 2011 12:01:17 -0500 Message-ID: <1295370077.12215.13.camel@gandalf.stny.rr.com> Mime-Version: 1.0 X-Mailer: Evolution 2.30.3 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 2011-01-18 at 17:41 +0100, Jiri Olsa wrote: > ops, missed this one.. > > would it make sense to update the IP inside the function_trace_probe_call > function to save one instruction in the entry code used by all? or it's > not worth it.. That would require function_trace_probe_call() to handle arch dependencies. I would like to keep arch dependent code in the arch subsystem when possible. This change only saves a subtraction to a register. It only gets called if we are tracing functions. This modification is far into the noise that is caused by the tracer. I don't think it is worth it as it will cause head aches with knowing how to handle the ip from the users of the function tracer. I do plan on making the function tracer a bit more generic for common users. I would like to keep the ip at the call site of mcount. -- Steve