linux-kernel - Re: [PATCH] checkpatch: make the line length warnings match the coding style document

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 for Android: free password hash cracker in your pocket

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

Message-ID: <983e6452a7f2af14ca7edfa56cd2e2997172a771.camel@perches.com>
Date:   Tue, 22 Dec 2020 08:22:06 -0800
From:   Joe Perches <joe@...ches.com>
To:     Christoph Hellwig <hch@....de>
Cc:     Matthew Wilcox <willy@...radead.org>, apw@...onical.com,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        linux-kernel@...r.kernel.org,
        linux-doc <linux-doc@...r.kernel.org>,
        Miguel Ojeda <miguel.ojeda.sandonis@...il.com>
Subject: Re: [PATCH] checkpatch: make the line length warnings match the
 coding style document

On Tue, 2020-12-22 at 14:12 +0100, Christoph Hellwig wrote:
> On Mon, Dec 21, 2020 at 08:08:20PM -0800, Joe Perches wrote:
> > On Thu, 2020-12-10 at 13:27 -0800, Joe Perches wrote:
> > > On Thu, 2020-12-10 at 20:09 +0000, Matthew Wilcox wrote:
> > > > On Thu, Dec 10, 2020 at 12:05:04PM -0800, Joe Perches wrote:
> > > > > Also, given the ever increasing average identifier length, strict
> > > > > adherence to 80 columns is sometimes just not possible without silly
> > > > > visual gymnastics.  The kernel now has quite a lot of 30+ character
> > > > > length function names, constants, and structs.
> > > > 
> > > > maybe checkpatch should warn for identifiers that are 30+ characters
> > > > long?  address the problem at its source ..
> > > 
> > > Hard to know when to warn as patches could just add uses of already
> > > existing names and emitting warnings for those would just be annoying.
> > > 
> > > Maybe something that tests long identifier additions of
> > > defines/functions/macros/structs but not their uses and maybe only
> > > then in patches and not files.
> > > 
> > > Perhaps:
> > 
> > Anyone care that this should be added or not added to checkpatch?
> 
> It is pretty useless.

Maybe so, if only because I chose a high value for the max id length
to avoid controversy.  I would prefer something like 20.

> What we need is a patch that doesn't make people
> uselessly add overly long lines against the intent of the coding style
> document.  I have submitted a pretty reasonable one, and I'm open to
> alternatives, but we need to to stop people submitting code that does
> not fit the coding style all the time because checkpatch doesn't
> complain.

Having checkpatch complain about > 80 column lines didn't stop
patches before, likely it wouldn't stop patches now.

Emitting yet more messages for trivial lines > 80 columns is also
against the intent of the commit that changed the line length maximum.

commit bdc48fa11e46f867ea4d75fa59ee87a7f48be144
checkpatch/coding-style: deprecate 80-column warning

The effect of your patch might as well revert the checkpatch portion
of that commit.

I think that's not a great idea for the reason in the commit message.