From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754177AbdKNNsY (ORCPT ); Tue, 14 Nov 2017 08:48:24 -0500 Received: from eddie.linux-mips.org ([148.251.95.138]:43230 "EHLO cvs.linux-mips.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751836AbdKNNsQ (ORCPT ); Tue, 14 Nov 2017 08:48:16 -0500 Date: Tue, 14 Nov 2017 14:35:24 +0100 From: Ralf Baechle To: jiaxun.yang@flygoat.com Cc: linux-mips@linux-mips.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH 3/4] MIPS: Loongson64: Yeeloong add platform driver Yeeloong is a laptop with a MIPS Loongson 2F processor, AMD CS5536 chipset, and KB3310B controller. Message-ID: <20171114133524.GE13046@linux-mips.org> References: <20171112063617.26546-1-jiaxun.yang@flygoat.com> <20171112063617.26546-3-jiaxun.yang@flygoat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20171112063617.26546-3-jiaxun.yang@flygoat.com> User-Agent: Mutt/1.9.1 (2017-09-22) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sun, Nov 12, 2017 at 02:36:16PM +0800, jiaxun.yang@flygoat.com wrote: > + asm(".set noreorder\n"); > + /* input enable */ > + outl(0x00000800, (gpio_base | 0xA0)); > + /* revert the input */ > + outl(0x00000800, (gpio_base | 0xA4)); > + /* event-int enable */ > + outl(0x00000800, (gpio_base | 0xB8)); > + asm(".set reorder\n"); I forgot to comment on this .set noreorder thing. GCC expects gas to be in reorder mode at the end of every bit of inline assembler, so above code is at the mercy of GCC. Not sure what the intent was here, was it to avoid GCC from reordering the outl() calls? These are already marked volatlile and should not be reordered by GCC. .set (no)reorder on the other hand is a directive to GAS and GCC won't be influenced at all by it. Cheers, Ralf