Re: [PATCH 3/6] mm: add is_migrate_isolate_page_nolock() for cases where locking is undesirable

From: Joonsoo Kim
Date: Tue Mar 04 2014 - 19:39:19 EST


On Fri, Feb 28, 2014 at 03:15:01PM +0100, Vlastimil Babka wrote:
> This patch complements the addition of get_pageblock_migratetype_nolock() for
> the case where is_migrate_isolate_page() cannot be called with zone->lock held.
> A race with set_pageblock_migratetype() may be detected, in which case a caller
> supplied argument is returned.
>
> Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
> ---
> include/linux/page-isolation.h | 24 ++++++++++++++++++++++++
> mm/hugetlb.c | 2 +-
> 2 files changed, 25 insertions(+), 1 deletion(-)
>
> diff --git a/include/linux/page-isolation.h b/include/linux/page-isolation.h
> index 3fff8e7..f7bd491 100644
> --- a/include/linux/page-isolation.h
> +++ b/include/linux/page-isolation.h
> @@ -2,10 +2,30 @@
> #define __LINUX_PAGEISOLATION_H
>
> #ifdef CONFIG_MEMORY_ISOLATION
> +/*
> + * Should be called only with zone->lock held. In cases where locking overhead
> + * is undesirable, consider the _nolock version.
> + */
> static inline bool is_migrate_isolate_page(struct page *page)
> {
> return get_pageblock_migratetype(page) == MIGRATE_ISOLATE;
> }
> +/*
> + * When called without zone->lock held, a race with set_pageblock_migratetype
> + * may result in bogus values. The race may be detected, in which case the
> + * value of race_fallback argument is returned. For details, see
> + * get_pageblock_migratetype_nolock().
> + */
> +static inline bool is_migrate_isolate_page_nolock(struct page *page,
> + bool race_fallback)
> +{
> + int migratetype = get_pageblock_migratetype_nolock(page, MIGRATE_TYPES);
> +
> + if (unlikely(migratetype == MIGRATE_TYPES))
> + return race_fallback;
> +
> + return migratetype == MIGRATE_ISOLATE;
> +}
> static inline bool is_migrate_isolate(int migratetype)
> {
> return migratetype == MIGRATE_ISOLATE;
> @@ -15,6 +35,10 @@ static inline bool is_migrate_isolate_page(struct page *page)
> {
> return false;
> }
> +static inline bool is_migrate_isolate_page_nolock(struct page *page)
> +{
> + return false;
> +}
> static inline bool is_migrate_isolate(int migratetype)
> {
> return false;

Nitpick.
You need race_fallback parameter for is_migrate_isolate_page_nolock().

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