The patch titled
Subject: mm-page_alloc-fix-allocation-imbalances-from-speculative-cache-lookup
has been removed from the -mm tree. Its filename was
mm-page_alloc-fix-allocation-imbalances-from-speculative-cache-lookup.patch
This patch was dropped because it is obsolete
------------------------------------------------------
From: Johannes Weiner <hannes@xxxxxxxxxxx>
Subject: mm-page_alloc-fix-allocation-imbalances-from-speculative-cache-lookup
fixes suggested by Hugh
Link: https://lkml.kernel.org/r/YFo7SOni0s0TbXUm@xxxxxxxxxxx
Fixes: e320d3012d25 mm/page_alloc.c: fix freeing non-compound pages
Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Reported-by: Hugh Dickins <hughd@xxxxxxxxxx>
Reported-by: Matthew Wilcox <willy@xxxxxxxxxxxxx>
Acked-by: Hugh Dickins <hughd@xxxxxxxxxx>
Reviewed-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
Acked-by: Hugh Dickins <hughd@xxxxxxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxx>
Cc: Zhou Guanghui <zhouguanghui1@xxxxxxxxxx>
Cc: Zi Yan <ziy@xxxxxxxxxx>
Cc: Shakeel Butt <shakeelb@xxxxxxxxxx>
Cc: Roman Gushchin <guro@xxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/page_alloc.c | 26 +++++++++++++++++---------
1 file changed, 17 insertions(+), 9 deletions(-)
--- a/mm/page_alloc.c~mm-page_alloc-fix-allocation-imbalances-from-speculative-cache-lookup
+++ a/mm/page_alloc.c
@@ -5083,6 +5083,8 @@ static inline void free_the_page(struct
*/
void __free_pages(struct page *page, unsigned int order)
{
+ bool compound = PageHead(page);
+
/*
* Drop the base reference from __alloc_pages and free. In
* case there is an outstanding speculative reference, from
@@ -5094,18 +5096,24 @@ void __free_pages(struct page *page, uns
}
/*
- * The speculative reference will put and free the page.
+ * Ok, the speculative reference will put and free the page.
+ *
+ * - If this was an order-0 page, we're done.
+ *
+ * - If the page was compound, the other side will free the
+ * entire page and we're done here as well. Just note that
+ * freeing clears PG_head, so it can only be read reliably
+ * before the put_page_testzero().
*
- * However, if the speculation was into a higher-order page
- * chunk that isn't marked compound, the other side will know
- * nothing about our buddy pages and only free the order-0
- * page at the start of our chunk! We must split off and free
- * the buddy pages here.
+ * - If the page was of higher order but NOT marked compound,
+ * the other side will know nothing about our buddy pages
+ * and only free the order-0 page at the start of our block.
+ * We must split off and free the buddy pages here.
*
- * The buddy pages aren't individually refcounted, so they
- * can't have any pending speculative references themselves.
+ * The buddy pages aren't individually refcounted, so they
+ * can't have any pending speculative references themselves.
*/
- if (!PageHead(page) && order > 0) {
+ if (order > 0 && !compound) {
split_page_memcg(page, 1 << order);
while (order-- > 0)
free_the_page(page + (1 << order), order);
_
Patches currently in -mm which might be from hannes@xxxxxxxxxxx are
mm-page-writeback-simplify-memcg-handling-in-test_clear_page_writeback.patch
mm-memcontrol-fix-cpuhotplug-statistics-flushing.patch
mm-memcontrol-kill-mem_cgroup_nodeinfo.patch
mm-memcontrol-privatize-memcg_page_state-query-functions.patch
cgroup-rstat-support-cgroup1.patch
cgroup-rstat-punt-root-level-optimization-to-individual-controllers.patch
mm-memcontrol-switch-to-rstat.patch
mm-memcontrol-switch-to-rstat-fix-2.patch
mm-memcontrol-consolidate-lruvec-stat-flushing.patch
kselftests-cgroup-update-kmem-test-for-new-vmstat-implementation.patch
