From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932457AbcHSLxb (ORCPT ); Fri, 19 Aug 2016 07:53:31 -0400 Received: from mx1.redhat.com ([209.132.183.28]:48730 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932418AbcHSLx3 (ORCPT ); Fri, 19 Aug 2016 07:53:29 -0400 From: Josh Poimboeuf To: Thomas Gleixner , Ingo Molnar , "H . Peter Anvin" Cc: x86@kernel.org, linux-kernel@vger.kernel.org, Andy Lutomirski , Linus Torvalds , Steven Rostedt , Brian Gerst , Kees Cook , Peter Zijlstra , Frederic Weisbecker , Byungchul Park , Nilay Vaish Subject: [PATCH 6/8] ftrace/x86: implement HAVE_FUNCTION_GRAPH_RET_ADDR_PTR Date: Fri, 19 Aug 2016 06:53:00 -0500 Message-Id: In-Reply-To: References: X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.31]); Fri, 19 Aug 2016 11:53:28 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Use the more reliable version of ftrace_graph_ret_addr() so we no longer have to worry about the unwinder getting out of sync with the function graph ret_stack index, which can happen if the unwinder skips any frames before calling ftrace_graph_ret_addr(). This fixes this issue (and several others like it): $ cat /proc/self/stack [] save_stack_trace_tsk+0x22/0x40 [] proc_pid_stack+0xb9/0x110 [] proc_single_show+0x54/0x80 [] seq_read+0x108/0x3e0 [] __vfs_read+0x37/0x140 [] vfs_read+0x99/0x140 [] SyS_read+0x58/0xc0 [] entry_SYSCALL_64_fastpath+0x1f/0xbd [] 0xffffffffffffffff $ echo function_graph > /sys/kernel/debug/tracing/current_tracer $ cat /proc/self/stack [] return_to_handler+0x0/0x27 [] print_context_stack+0xfc/0x100 [] return_to_handler+0x0/0x27 [] dump_trace+0x12b/0x350 [] return_to_handler+0x0/0x27 [] save_stack_trace_tsk+0x22/0x40 [] return_to_handler+0x0/0x27 [] proc_pid_stack+0xb9/0x110 [] return_to_handler+0x0/0x27 [] proc_single_show+0x54/0x80 [] return_to_handler+0x0/0x27 [] seq_read+0x108/0x3e0 [] return_to_handler+0x0/0x27 [] __vfs_read+0x37/0x140 [] return_to_handler+0x0/0x27 [] vfs_read+0x99/0x140 [] 0xffffffffffffffff Enabling function graph tracing causes the stack trace to change in two ways: First, the real call addresses are confusingly interspersed with 'return_to_handler' addresses. This issue will be fixed by the next patch. Second, the stack trace is offset by two frames, because the unwinder skipped the first two frames and got out of sync with the ret_stack index. This patch fixes this issue. Signed-off-by: Josh Poimboeuf --- arch/x86/include/asm/ftrace.h | 2 ++ arch/x86/kernel/ftrace.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h index 37f67cb..eccd0ac 100644 --- a/arch/x86/include/asm/ftrace.h +++ b/arch/x86/include/asm/ftrace.h @@ -14,6 +14,8 @@ #define ARCH_SUPPORTS_FTRACE_OPS 1 #endif +#define HAVE_FUNCTION_GRAPH_RET_ADDR_PTR + #ifndef __ASSEMBLY__ extern void mcount(void); extern atomic_t modifying_ftrace_code; diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index ae3b1fb..8639bb2 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -1029,7 +1029,7 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent, } if (ftrace_push_return_trace(old, self_addr, &trace.depth, - frame_pointer, NULL) == -EBUSY) { + frame_pointer, parent) == -EBUSY) { *parent = old; return; } -- 2.7.4