linux-kernel - Re: [PATCH] checkpatch: warn if changed lines exceeds a maximum size

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] [day] [month] [year] [list]

Message-ID: <20180201090348.61c9be3f@lwn.net>
Date:   Thu, 1 Feb 2018 09:03:48 -0700
From:   Jonathan Corbet <corbet@....net>
To:     "Brown, Nicholas" <nb930b@...l.att.com>
Cc:     "joe@...ches.com" <joe@...ches.com>,
        "apw@...onical.com" <apw@...onical.com>,
        "me@...in.cc" <me@...in.cc>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] checkpatch: warn if changed lines exceeds a maximum
 size

On Thu, 1 Feb 2018 15:42:35 +0000
"Brown, Nicholas" <nb930b@...l.att.com> wrote:

> > I think the metric is too simplistic and
> > not particularly useful.  
> 
> I'm not sure it's any more simplistic than than the character line
> length limit, which is there to prompt thought on code nesting levels,
> etc. And as it has to be explicitly configured it allows developers the
> discretion to determine a code change size that meaningful in a given
> situation.

The line-length limit relates directly to code readability and
non-infringement of a developer's sacred right to work unimpeded on an
80x24 xterm (the last vt100 died, unfortunately).

A line-count limit lacks even that justification.  The rule on splitting
patches is entirely about logical changes that can be reviewed
independently.  Some of those changes involve a lot of lines, others do
not.  If the correct splits do not come to you when you're writing the
changelogs for your patches (or before), a tool nagging about line counts
really isn't going to help.  I would expect it to miss most patches
actually in need of splitting while complaining about many patches that
are just fine.

Thanks for working to improve the tools - they certainly can afford a lot
of improvement!  But my own suggestion would be to look a bit further for
improvements that will be truly helpful to the development community.

Thanks,

jon