On Tue, 20 May 2014, Mel Gorman wrote:
> Dan Carpenter reported the following bug
>
> The patch a486e00b8283: "mm: page_alloc: calculate classzone_idx
> once from the zonelist ref" from May 17, 2014, leads to the
> following static checker warning:
>
> mm/page_alloc.c:2543 __alloc_pages_slowpath()
> warn: we tested 'nodemask' before and it was 'false'
>
> mm/page_alloc.c
> 2537 * Find the true preferred zone if the allocation is unconstrained by
> 2538 * cpusets.
> 2539 */
> 2540 if (!(alloc_flags & ALLOC_CPUSET) && !nodemask) {
> ^^^^^^^^^
> Patch introduces this test.
>
> 2541 struct zoneref *preferred_zoneref;
> 2542 preferred_zoneref = first_zones_zonelist(zonelist, high_zoneidx,
> 2543 nodemask ? : &cpuset_current_mems_allowed,
> ^^^^^^^^
> Patch introduces this test as well.
>
> 2544 &preferred_zone);
> 2545 classzone_idx = zonelist_zone_idx(preferred_zoneref);
> 2546 }
>
> This patch should resolve it and is a fix to the mmotm patch
> mm-page_alloc-calculate-classzone_idx-once-from-the-zonelist-ref
>
> Signed-off-by: Mel Gorman <mgorman@xxxxxxx>
> ---
> mm/page_alloc.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 0959b09..ebb947d 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -2590,8 +2590,7 @@ restart:
> if (!(alloc_flags & ALLOC_CPUSET) && !nodemask) {
> struct zoneref *preferred_zoneref;
> preferred_zoneref = first_zones_zonelist(zonelist, high_zoneidx,
> - nodemask ? : &cpuset_current_mems_allowed,
> - &preferred_zone);
> + NULL, &preferred_zone);
> classzone_idx = zonelist_zone_idx(preferred_zoneref);
> }
>
The switch from "nodemask : &cpuset_current_mems_allowed" to NULL isn't
described in the changelog but makes sense as well because of
!(alloc_flags & ALLOC_CPUSET).
Acked-by: David Rientjes <rientjes@xxxxxxxxxx>
--
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>