From mboxrd@z Thu Jan 1 00:00:00 1970 From: SF Markus Elfring Date: Tue, 22 Aug 2017 14:36:35 +0000 Subject: Re: [alsa-devel] ALSA: pcsp: Use common error handling code in snd_card_pcsp_probe() Message-Id: <139363f6-e059-defb-357e-f18645ba9768@users.sourceforge.net> List-Id: References: <08ee0d6b-788b-2845-6964-e1e55c2d2292@users.sourceforge.net> <20170822121625.syvr64kwyh5xjexg@mwanda> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit To: Takashi Iwai , alsa-devel@alsa-project.org Cc: kernel-janitors@vger.kernel.org, LKML , Dan Carpenter >> I got the impression that the functions which are called at the updated places >> by the function “snd_card_pcsp_probe” indicate a successful execution >> only by zero so far. > > You have the impression, great. This aspect is also a general programming interface issue for some functions. > And what's the reason to drop the negative check? * I find it a bit safer when the error predicate is “return value != 0”. * It is also a small source code reduction. > It's not clearer, not better readable. It seems that we have got different development opinions this time. > And, the worst part is that you've done it silently even without > mentioning in the change log at all. That's really bad. > Just don't do it. I found it not relevant enough for the commit message. > For example, the control API functions may return the positive number > when the value got changed, 0 for else, and a negative number for the > error. The functions returning some numbers may return positive > numbers, of course. Did I touch any specific function calls which belong to this programming interface category? Regards, Markus