[PATCH] mm: avoid false-positive PageTail() during meminit

[PATCH] mm: avoid false-positive PageTail() during meminit

[Kirill A. Shutemov]

Since compound_head() rework we encode PageTail() into bit 0 of
page->lru.next (aka page->compound_head). We need to make sure that
page->lru is initialized before first use of compound_head() or
PageTail().

My page-flags patchset makes sure that we don't use PG_reserved on
compound pages. That means we have PageTail() check as eary as in
SetPageReserved() in reserve_bootmem_region()

Let's initialize page->lru before that to avoid false positive from
PageTail().

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
---

Andrew, this can be folded into "mm: make compound_head() robust"

---
 mm/page_alloc.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 778eb3a0f103..95fbc43a93dd 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -926,8 +926,6 @@ static void init_reserved_page(unsigned long pfn)
 #else
 static inline void init_reserved_page(unsigned long pfn)
 {
-	/* Avoid false-positive PageTail() */
-	INIT_LIST_HEAD(&pfn_to_page(pfn)->lru);
 }
 #endif /* CONFIG_DEFERRED_STRUCT_PAGE_INIT */
 
@@ -947,6 +945,10 @@ void __meminit reserve_bootmem_region(unsigned long start, unsigned long end)
 			struct page *page = pfn_to_page(start_pfn);
 
 			init_reserved_page(start_pfn);
+
+			/* Avoid false-positive PageTail() */
+			INIT_LIST_HEAD(&page->lru);
+
 			SetPageReserved(page);
 		}
 	}
-- 
2.5.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>