The patch titled
Subject: mm: don't use compound_head() in virt_to_head_page()
has been removed from the -mm tree. Its filename was
mm-dont-use-compound_head-in-virt_to_head_page-v3.patch
This patch was dropped because it was folded into mm-dont-use-compound_head-in-virt_to_head_page.patch
------------------------------------------------------
From: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Subject: mm: don't use compound_head() in virt_to_head_page()
Change from v2: Add some code comments
Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Acked-by: Christoph Lameter <cl@xxxxxxxxx>
Cc: Pekka Enberg <penberg@xxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Cc: Jesper Dangaard Brouer <brouer@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/mm.h | 17 +++++++++++++++++
1 file changed, 17 insertions(+)
diff -puN include/linux/mm.h~mm-dont-use-compound_head-in-virt_to_head_page-v3 include/linux/mm.h
--- a/include/linux/mm.h~mm-dont-use-compound_head-in-virt_to_head_page-v3
+++ a/include/linux/mm.h
@@ -446,6 +446,12 @@ static inline struct page *compound_head
return tail;
}
+/*
+ * Since either compound page could be dismantled asynchronously in THP
+ * or we access asynchronously arbitrary positioned struct page, there
+ * would be tail flag race. To handle this race, we should call
+ * smp_rmb() before checking tail flag. compound_head_by_tail() did it.
+ */
static inline struct page *compound_head(struct page *page)
{
if (unlikely(PageTail(page)))
@@ -453,6 +459,11 @@ static inline struct page *compound_head
return page;
}
+/*
+ * If we access compound page synchronously such as access to
+ * allocated page, there is no need to handle tail flag race, so we can
+ * check tail flag directly without any synchronization primitive.
+ */
static inline struct page *compound_head_fast(struct page *page)
{
if (unlikely(PageTail(page)))
@@ -539,6 +550,12 @@ static inline struct page *virt_to_head_
{
struct page *page = virt_to_page(x);
+ /*
+ * We don't need to worry about synchronization of tail flag
+ * when we call virt_to_head_page() since it is only called for
+ * already allocated page and this page won't be freed until
+ * this virt_to_head_page() is finished. So use _fast variant.
+ */
return compound_head_fast(page);
}
_
Patches currently in -mm which might be from iamjoonsoo.kim@xxxxxxx are
mm-slub-optimize-alloc-free-fastpath-by-removing-preemption-on-off.patch
mm-dont-use-compound_head-in-virt_to_head_page.patch
mm-slab_common-use-kmem_cache_free.patch
mm-vmstatc-fix-cleanup-ifdefs.patch
mm-vmstatc-fix-cleanup-ifdefs-fix-2.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
