linux-kernel - Re: [PATCH v3 1/2] checkpatch: add Kconfig 'default n' test

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]

Date:	Wed, 08 Jun 2016 21:42:40 +0200
From:	Paul Bolle <pebolle@...cali.nl>
To:	Yingjoe Chen <yingjoe.chen@...iatek.com>,
	Andy Whitcroft <apw@...onical.com>
Cc:	Joe Perches <joe@...ches.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	linux-kernel@...r.kernel.org, srv_heupstream@...iatek.com,
	Andi Kleen <ak@...ux.intel.com>
Subject: Re: [PATCH v3 1/2] checkpatch: add Kconfig 'default n' test

On di, 2016-06-07 at 21:16 +0800, Yingjoe Chen wrote:
> On Mon, 2016-06-06 at 20:10 +0100, Andy Whitcroft wrote:
> > > Is it obvious that a Kconfig has "default n" ?
> > > This seems to work, but is this useful?
>
> While sending patch for upstream, I saw maintainers request it to be
> removed. So I think it might worth adding check to it.
> Some examples from google:
> 
> http://lists.infradead.org/pipermail/linux-arm-kernel/2012-September/1
> 20733.html
> https://lkml.org/lkml/2013/3/16/153
> https://lkml.org/lkml/2016/5/23/657

There's one rather subtle case where setting "default n" is, sort of,
useful. See lkml.kernel.org/r/<178407860.0zoJnDfCo1@...ticalops> . 

(I seem to remember disagreeing here. Ie, in my view setting defaults
for a specific Kconfig symbol at two different places is confusing at
best. People probably weren't convinced by my objections. I also
remember diving into this by looking at the various places where a
Kconfig symbol was being set twice. I must have ended that endeavor when
it became clear to me I was't making any progress.)

Even though there's  a corner case where "default n" is useful, it could
still be worth to add a checkpatch check that warns about it. But I
can't say I feel strongly about this either way.

Paul Bolle