From mboxrd@z Thu Jan 1 00:00:00 1970 From: SF Markus Elfring Date: Fri, 22 Sep 2017 20:44:23 +0000 Subject: Re: [media] spca500: Use common error handling code in spca500_synch310() Message-Id: List-Id: References: <4921ea61-49cd-4071-e636-c199daddec8e@users.sourceforge.net> <0baa322a-6019-70dc-0245-caae824ccb49@users.sourceforge.net> <25a82438-8760-66fd-23d3-24569078f906@grinta.net> In-Reply-To: <25a82438-8760-66fd-23d3-24569078f906@grinta.net> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Daniele Nicolodi , linux-media@vger.kernel.org Cc: Julia Lawall , Hans Verkuil , Mauro Carvalho Chehab , Sakari Ailus , LKML , kernel-janitors@vger.kernel.org > No one needs to argue about keeping it the way it is. I got an other impression in this case after a bit of information was presented which seems to be contradictory. > I don't see any improvement brought by the proposed change, Do you care if the source code for an error message is present only once in this function? > other than making the code harder to read. I suggest to reconsider this concern. > I find goto statements hard to read, because they inherently make some > information non local. They are justified in error path handling, > if the error path only unwinds what the function did early on. > That's not the case here. I am looking also for change possibilities without such a restriction. > The same applies to dozens of patches you proposed recently. I proposed some software updates to reduce a bit of code duplication. Do you find any corresponding approaches useful? Regards, Markus