On Tue, Mar 05, 2019 at 10:14:29PM -0500, Qian Cai wrote:
> I don't understand this part.
>
> @@ -2279,14 +2286,24 @@ static enum compact_result compact_zone_order(struct
> zone *zone, int order,
> .ignore_skip_hint = (prio == MIN_COMPACT_PRIORITY),
> .ignore_block_suitable = (prio == MIN_COMPACT_PRIORITY)
> };
> + struct capture_control capc = {
> + .cc = &cc,
> + .page = NULL,
> + };
> +
> + if (capture)
> + current->capture_control = &capc;
>
>
> That check will always be true as it is,
>
It's a defensive check allowing for the possibility that
try_to_compact_pages() is passed NULL. Originally the structure was
different but I preserved the NULL check to avoid potential surprises.
It could be changed but I don't think it'll help. I aim to setup a machine
with your config today, try again to reproduce the problem and look at
the patch again to see can I spot how it could corrupt anything.
--
Mel Gorman
SUSE Labs
