From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754519Ab2LCQNy (ORCPT ); Mon, 3 Dec 2012 11:13:54 -0500 Received: from arrakis.dune.hu ([78.24.191.176]:41935 "EHLO eddie.linux-mips.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1751412Ab2LCQNx (ORCPT ); Mon, 3 Dec 2012 11:13:53 -0500 Date: Mon, 3 Dec 2012 17:13:46 +0100 From: Ralf Baechle To: Steven Rostedt Cc: David Daney , Alan Cooper , linux-mips@linux-mips.org, linux-kernel@vger.kernel.org Subject: Re: MIPS Function Tracer question Message-ID: <20121203161346.GA29573@linux-mips.org> References: <50B7E91C.6070403@gmail.com> <1354545648.6276.202.camel@gandalf.local.home> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1354545648.6276.202.camel@gandalf.local.home> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Dec 03, 2012 at 09:40:48AM -0500, Steven Rostedt wrote: > The issue is with x86. Gcc wont compile if you have -pg and > -fomit-frame-pointer on x86. I originally forced function tracing to > select FRAME_POINTER, but because now on x86 with -mfentry, -pg no > longer requires frame pointers being set, I just let -pg complain one > way or the other. I believe that gcc by default will not add frame > pointers. Thus adding function tracing just prevents > -fomit-frame-pointer from being set, and if -pg requires frame pointers > it will automatically enable them otherwise they should not be enabled. On architectures such as MIPS where a frame pointer is not required for debugging -O and higher imply -fomit-frame-pointer. Ralf