linux-kernel - Re: [PATCH] regulator: axp20x: Get rid of AXP20X_xxx

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: <a724fa00-14a2-865b-ed8d-60fa6bda88ba@schinagl.nl>
Date:   Wed, 27 Feb 2019 21:26:32 +0100
From:   Olliver Schinagl <oliver@...inagl.nl>
To:     Mark Brown <broonie@...nel.org>
Cc:     Axel Lin <axel.lin@...ics.com>, Chen-Yu Tsai <wens@...e.org>,
        Priit Laes <plaes@...es.org>,
        Liam Girdwood <lgirdwood@...il.com>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] regulator: axp20x: Get rid of AXP20X_xxx_START/END/STEPS
 defines

On 27-02-2019 21:05, Mark Brown wrote:
> On Wed, Feb 27, 2019 at 08:41:46PM +0100, Olliver Schinagl wrote:
>> On 25-02-2019 18:25, Mark Brown wrote:
>>> If you find you need to describe what the fields are it would be much
>>> more constructive to add a comment at the top of the table saying what
>>> they are.  As things are this isn't helping anyone - as a big pile of
>>> defines it's hard to read the values without context for how they're
>>> used and if you're looking at the table you can't tell what the
>>> regulator actually supports without going and decoding the defines.
>> Then the name of the define should be more constructive, which imo they
>> are reasonably? But as everything with programming, naming things is the
>> he hardest part, right?
> I really don't think that's it - I think that sometimes a data table is
> just a data table.  There are some coding styles that work to avoid
> having raw numbers anywhere in code outside of defines at all costs but
> I do think that goes too far in cases like this where the name of the
> define is at some level just going to summarize what should go in a
> given slot in a table which adds little.

I'm not sure if we're still talking about the same thing or same table;
In any case, this is something up to personal taste in the end, and I am
in the camp that favors (readable, which of course could always be
improved) defines over magic values; especially when it comes to these
bit-selectors. And even if a little less here, to keep things consistent
with all defines is why at least I prefer the one approach. You guys
prefer the raw values. Two flavors of two opinions I suppose :)