linux-kernel - Re: [PATCH v2 5/8] rtc: isl12022: implement RTC_VL

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <bd17c1a4-6102-3d93-5494-89fdb0a346e5@rasmusvillemoes.dk>
Date:   Thu, 15 Jun 2023 12:53:24 +0200
From:   Rasmus Villemoes <linux@...musvillemoes.dk>
To:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Alexandre Belloni <alexandre.belloni@...tlin.com>
Cc:     Alessandro Zummo <a.zummo@...ertech.it>,
        devicetree@...r.kernel.org, Rob Herring <robh+dt@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Conor Dooley <conor+dt@...nel.org>, linux-rtc@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 5/8] rtc: isl12022: implement RTC_VL_READ ioctl

On 14/06/2023 17.13, Andy Shevchenko wrote:

> When reading such a code the following questions are arisen:
> 1) Can the positive return value be the case?
> 2) If so, what is the meaning of a such?
> 3) Why do we not care about it?
> 
> All this can simply gone if we use
> 
> 	ret = foo(...);
> 	if (ret)
> 		return ret;
> 
> As it's clear that whatever is non-zero we accept as something to be promoted
> to the upper layer. I hope this explains my position.

But we're in a context (in this case an ->ioctl method) where _our_
caller expects 0/-ESOMETHING, so returning something positive would be a
bug - i.e., either way of spelling that if(), the reader must know that
foo() also has those 0/-ESOMETHING semantics.

I honestly didn't think much about it, but looking at the existing code
and the stuff I add, all other places actually do 'if (ret)', so I've
updated this site for consistency.

Rasmus