Re: [PATCH 1/4] kernel.h: Implement DIV_ROUND_CLOSEST_ULL

From: Javi Merino <javi.merino@arm.com>
To: Jeff Epler <jepler@unpythonic.net>
Cc: "akpm@linux-foundation.org" <akpm@linux-foundation.org>,
	"intel-gfx@lists.freedesktop.org"
	<intel-gfx@lists.freedesktop.org>,
	"dri-devel@lists.freedesktop.org"
	<dri-devel@lists.freedesktop.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	Daniel Vetter <daniel.vetter@intel.com>,
	Jani Nikula <jani.nikula@linux.intel.com>,
	David Airlie <airlied@linux.ie>,
	"Darrick J. Wong" <djwong@us.ibm.com>,
	Guenter Roeck <linux@roeck-us.net>, Alex Elder <elder@ieee.org>
Subject: Re: [PATCH 1/4] kernel.h: Implement DIV_ROUND_CLOSEST_ULL
Date: Tue, 24 Mar 2015 11:49:12 +0000	[thread overview]
Message-ID: <20150324114912.GA3982@e104805> (raw)
In-Reply-To: <20150323170416.GA6161@e104805>

On Mon, Mar 23, 2015 at 05:04:16PM +0000, Javi Merino wrote:
> On Mon, Mar 23, 2015 at 12:34:04PM +0000, Jeff Epler wrote:
> > On Fri, Mar 20, 2015 at 11:14:40AM +0000, Javi Merino wrote:
> > > +/*
> > > + * Same as above but for u64 dividends.  divisor must be a 32-bit
> > > + * number.
> > > + */
> > > +#define DIV_ROUND_CLOSEST_ULL(x, divisor)(		\
> > > +{							\
> > > +	unsigned long long _tmp = (x) + (divisor) / 2;	\
> > > +	do_div(_tmp, divisor);				\
> > > +	_tmp;						\
> > > +}							\
> > > +)
> > 
> > The macro evaluates 'divisor' twice.
> 
> Good catch.  That needs to be fixed.  I could do the typeof trick that
> DIV_ROUND_CLOSEST() does but it's probably better to just create a
> static function as Alex Elder suggests.  I'll send a v2 tomorrow with
> a static function instead.

Nah, thinking about it and seeing the other macros in this file
(esp. DIV_ROUND_UP() and DIV_ROUND_UP_ULL()) I don't think a static
function gives you anything.  I'll just modify the macro to look similar
to DIV_ROUND_CLOSEST()

Cheers,
Javi