From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754986AbcEZSnq (ORCPT ); Thu, 26 May 2016 14:43:46 -0400 Received: from mx1.redhat.com ([209.132.183.28]:33009 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754548AbcEZSnp (ORCPT ); Thu, 26 May 2016 14:43:45 -0400 Date: Thu, 26 May 2016 13:43:43 -0500 From: Josh Poimboeuf To: Linus Torvalds Cc: Ingo Molnar , Linux Kernel Mailing List , DRI Subject: [PATCH] drm/vmwgfx: fix "duplicate frame pointer save" warning Message-ID: <20160526184343.fdtjjjg67smmeekt@treble> References: <20160525175631.44xgcylo24ysftss@treble> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.6.0.1 (2016-04-01) X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.38]); Thu, 26 May 2016 18:43:44 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, May 25, 2016 at 04:51:21PM -0700, Linus Torvalds wrote: > On Wed, May 25, 2016 at 10:56 AM, Josh Poimboeuf wrote: > > > > I used to have a STACKTOOL_IGNORE_INSN macro that would tell the tool to > > skip warnings for specific instructions in inline asm: > > > > Here's an example of it how it was used: > > Ok, looking at that, I'm starting to suspect that it is simpler to > just use STACK_FRAME_NON_STANDARD and mark the two functions that use > this particular inline asm with the odd %rbp problem. > > It's a rather special case, after all. That's fine with me, it is indeed a rare case. We can always add the per-instruction macro later if needed. Here's a patch. --- From: Josh Poimboeuf Subject: [PATCH] drm/vmwgfx: fix "duplicate frame pointer save" warning objtool reports the following warnings: drivers/gpu/drm/vmwgfx/vmwgfx_msg.o: warning: objtool: vmw_send_msg()+0x107: duplicate frame pointer save drivers/gpu/drm/vmwgfx/vmwgfx_msg.o: warning: objtool: vmw_host_get_guestinfo()+0x252: duplicate frame pointer save To quote Linus: "The reason is that VMW_PORT_HB_OUT() uses a magic instruction sequence (a "rep outsb") to communicate with the hypervisor (it's a virtual GPU driver for vmware), and %rbp is part of the communication. So the inline asm does a save-and-restore of the frame pointer around the instruction sequence. I actually find the objtool warning to be quite reasonable, so it's not exactly a false positive, since in this case it actually does point out that the frame pointer won't be reliable over that instruction sequence. But in this particular case it just ends up being the wrong thing - the code is what it is, and %rbp just can't have the frame information due to annoying magic calling conventions." Silence the warnings by telling objtool to ignore the two functions which use the VMW_PORT_HB_{IN,OUT} macros. Reported-by: Linus Torvalds Signed-off-by: Josh Poimboeuf --- drivers/gpu/drm/vmwgfx/vmwgfx_msg.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c index 6de283c..f0374f9 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c @@ -28,6 +28,7 @@ #include #include #include +#include #include #include "drmP.h" #include "vmwgfx_msg.h" @@ -194,7 +195,7 @@ static int vmw_send_msg(struct rpc_channel *channel, const char *msg) return -EINVAL; } - +STACK_FRAME_NON_STANDARD(vmw_send_msg); /** @@ -304,6 +305,7 @@ static int vmw_recv_msg(struct rpc_channel *channel, void **msg, return 0; } +STACK_FRAME_NON_STANDARD(vmw_recv_msg); /** -- 2.4.11