From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lyude Paul Subject: Re: [PATCH] HID: rmi: Make sure the HID device is opened on resume Date: Mon, 24 Jul 2017 15:46:23 -0400 Message-ID: <1500925583.6243.4.camel@redhat.com> References: <20170723011509.23651-1-lyude@redhat.com> <1500918313.6243.1.camel@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: stable-owner@vger.kernel.org To: Jiri Kosina Cc: Andy Shevchenko , linux-input , Andrew Duggan , stable@vger.kernel.org, Benjamin Tissoires , "linux-kernel@vger.kernel.org" List-Id: linux-input@vger.kernel.org Yeah I noticed that, sorry if my response wasn't very clear! Should probably wait to have my morning coffee before responding to these messages :P On Mon, 2017-07-24 at 21:28 +0200, Jiri Kosina wrote: > On Mon, 24 Jul 2017, Lyude Paul wrote: > > > > > So, call hid_hw_open() in rmi_post_resume() so we make sure > > > > that > > > > the > > > > device is alive before we try talking to it. > > > > > > > > This fixes RMI device suspend/resume over HID. > > > > - int ret; > > > > + int ret = 0; > > > > > > What's the point? > > > > So that we can use the same out: label at the end of the function > > that > > calls hid_hw_close() to return success. This being said though I > > just > > realized that setting ret will initialize it to 0 anyway, so I > > guess > > this can be dropped > > Andy's point was that hid_hw_open() is obviously re-initializing the > ret > before its first use as a return value, so there is no need to > initialize > it at a declaration time. >