On Tue, Feb 05, 2013 at 06:26:51PM -0800, Michel Lespinasse wrote:
> Just nitpicking, but:
>
> On Tue, Feb 5, 2013 at 3:57 AM, Mel Gorman <mgorman@xxxxxxx> wrote:
> > +static inline bool zone_is_idx(struct zone *zone, enum zone_type idx)
> > +{
> > + /* This mess avoids a potentially expensive pointer subtraction. */
> > + int zone_off = (char *)zone - (char *)zone->zone_pgdat->node_zones;
> > + return zone_off == idx * sizeof(*zone);
> > +}
>
> Maybe:
> return zone == zone->zone_pgdat->node_zones + idx;
> ?
>
Not a nit at all. Yours is more readable but it generates more code. A
single line function that uses the helper generates 0x3f bytes of code
(mostly function entry/exit) with your version and 0x39 bytes with mine. The
difference in efficiency is marginal as your version uses lea to multiply
by a constant but it's still slightly heavier.
The old code is fractionally better, your version is more readable so it's
up to Andrew really. Right now I think he's gone with his own version with
zone_idx() in the name of readability whatever sparse has to say about
the matter.
--
Mel Gorman
SUSE Labs
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>