From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754693Ab0CMGuc (ORCPT ); Sat, 13 Mar 2010 01:50:32 -0500 Received: from mail-gy0-f174.google.com ([209.85.160.174]:49229 "EHLO mail-gy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754367Ab0CMGuI (ORCPT ); Sat, 13 Mar 2010 01:50:08 -0500 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=sender:from:to:cc:subject:date:message-id:x-mailer:in-reply-to :references; b=PDzgsFjWg2wk+jnpouW5H8ETeC2WNG6n3qSc99UkNMXkYZ43p2P1vaaOKaNPpYkITN suTZih5QWD3KQ70Hvltm6/8/0l+sMBNCSMoDWqYDWisGarmPxnjN7ynZ96ZPVF0LGSGK kWWFzk7VFheVtDE1UXWqxWpvlCUKagr+owKJs= From: Rabin Vincent To: linux-kernel@vger.kernel.org Cc: linux-arm-kernel@lists.infradead.org, Rabin Vincent , Steven Rostedt , Frederic Weisbecker , Ingo Molnar , Abhishek Sagar , =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Subject: [PATCH 02/10] ARM: ftrace: document mcount formats Date: Sat, 13 Mar 2010 12:19:39 +0530 Message-Id: <1268462987-8271-3-git-send-email-rabin@rab.in> X-Mailer: git-send-email 1.7.0 In-Reply-To: <1268462987-8271-1-git-send-email-rabin@rab.in> References: <1268462987-8271-1-git-send-email-rabin@rab.in> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Add a comment describing the mcount variants and how the callsites look like. Acked-by: Frederic Weisbecker Signed-off-by: Rabin Vincent --- v2: Reword/expand comment. arch/arm/kernel/entry-common.S | 36 ++++++++++++++++++++++++++++++++++++ 1 files changed, 36 insertions(+), 0 deletions(-) diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 0b042bd..f05a35a 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -92,6 +92,42 @@ ENDPROC(ret_from_fork) #define CALL(x) .long x #ifdef CONFIG_FUNCTION_TRACER +/* + * When compiling with -pg, gcc inserts a call to the mcount routine at the + * start of every function. In mcount, apart from the function's address (in + * lr), we need to get hold of the function's caller's address. + * + * Older GCCs (pre-4.4) inserted a call to a routine called mcount like this: + * + * bl mcount + * + * These versions have the limitation that in order for the mcount routine to + * be able to determine the function's caller's address, an APCS-style frame + * pointer (which is set up with something like the code below) is required. + * + * mov ip, sp + * push {fp, ip, lr, pc} + * sub fp, ip, #4 + * + * With EABI, these frame pointers are not available unless -mapcs-frame is + * specified, and if building as Thumb-2, not even then. + * + * Newer GCCs (4.4+) solve this problem by introducing a new version of mcount, + * with call sites like: + * + * push {lr} + * bl __gnu_mcount_nc + * + * With these compilers, frame pointers are not necessary. + * + * mcount can be thought of as a function called in the middle of a subroutine + * call. As such, it needs to be transparent for both the caller and the + * callee: the original lr needs to be restored when leaving mcount, and no + * registers should be clobbered. (In the __gnu_mcount_nc implementation, we + * clobber the ip register. This is OK because the ARM calling convention + * allows it to be clobbered in subroutines and doesn't use it to hold + * parameters.) + */ #ifdef CONFIG_DYNAMIC_FTRACE ENTRY(mcount) stmdb sp!, {r0-r3, lr} -- 1.7.0