Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759282AbcCVK5A (ORCPT ); Tue, 22 Mar 2016 06:57:00 -0400 Received: from mail-oi0-f53.google.com ([209.85.218.53]:36448 "EHLO mail-oi0-f53.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759230AbcCVK4v (ORCPT ); Tue, 22 Mar 2016 06:56:51 -0400 MIME-Version: 1.0 In-Reply-To: <56E789C1.2070907@electromag.com.au> References: <1457968758-22670-1-git-send-email-geert+renesas@glider.be> <56E789C1.2070907@electromag.com.au> Date: Tue, 22 Mar 2016 11:56:50 +0100 Message-ID: Subject: Re: [PATCH] gpio: 74x164: Implement gpiochip.set_multiple() From: Linus Walleij To: Phil Reid , Axel Lin Cc: Geert Uytterhoeven , Alexandre Courbot , Gabor Juhos , Miguel Gaio , "linux-gpio@vger.kernel.org" , "linux-kernel@vger.kernel.org" Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 729 Lines: 21 On Tue, Mar 15, 2016 at 5:04 AM, Phil Reid wrote: > pca953x does not update the cached data unless the write operation > was successful. Which I folowed with the implement of set_multiple. > However a number of other drivers update regardless. > eg chip->buffer is updated even if the write_config fails. > > What is the preferred approach? Obviously the other drivers are buggy and need to be fixed. It can't be too many since the number of drivers with failable write operations aren's so many. (And I guess they seldom fail so it's not a big real world problem.) But it should be fixed. Add Axel Lin to the thread, he's awesome at finding semantic bugs. Patches accepted :) Yours, Linus Walleij