MIME-Version: 1.0
In-Reply-To: <1298669593.3356.2.camel@odin>
References: <1298669593.3356.2.camel@odin>
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 25 Feb 2011 14:10:49 -0800
Message-ID: <AANLkTim_bbv=WUd4DcB6fVmBy+-Yz1GNbDGga3rUeLNt@mail.gmail.com>
Subject: Re: [GIT PULL] Regulator fixes for 2.6.38
To: Liam Girdwood <lrg@slimlogic.co.uk>
Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        Jesper Juhl <jj@chaosbits.net>
Content-Type: text/plain; charset=ISO-8859-1
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1290
Lines: 34

On Fri, Feb 25, 2011 at 1:33 PM, Liam Girdwood <lrg@slimlogic.co.uk> wrote:
>
> Jesper Juhl (1):
>      regulator, mc13xxx: Remove pointless test for unsigned less than zero

So I absolutely detest patches like this.

It's not AT ALL pointless to write

   BUG_ON(x < 0 || x > XYZ);

because tests like that make it obvious that there are no sign issues,
without anybody ever having to worry about the sign. So it makes the
code more readable, and the intention more obvious. There is no
downside.

If it's an unsigned variable, the compiler can optimize away the test
against zero. In fact, it can do so even if it's signed (assuming XYZ
is a positive compile-time constant integer) and turn it into an
unsigned comparison.

So I did the pull, but I want to raise my objection to patches like
this. They are doubly wrong during the late -rc stage, but they are
wrong even otherwise.

If you have a compiler that warns about the test, you have a compiler
that is pure and utter shit. It's that simple.

                                    Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/