[PATCH RFC v3 17/19] mm: page_alloc: clear PG_zeroed on buddy merge if not both zero

["Michael S. Tsirkin" <mst@redhat.com>]

When two buddy pages merge in __free_one_page(), preserve
PG_zeroed on the merged page only if both buddies have the
flag set.  Otherwise clear it.

Without this, a zeroed page (freed via free_frozen_pages_zeroed
from balloon deflate) could merge with a non-zero buddy.  The merged
page would inherit PG_zeroed, and a later __GFP_ZERO allocation
would skip zeroing stale data in the non-zero half.

The page reporting path is not affected: it sets PG_zeroed during
allocation (page_del_and_expand), not on free list pages.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Assisted-by: Claude:claude-opus-4-6
Assisted-by: cursor-agent:GPT-5.4-xhigh
---
 mm/page_alloc.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 9311374bbd2d..122b49a6d435 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -991,6 +991,8 @@ static inline void __free_one_page(struct page *page,
 	unsigned long buddy_pfn = 0;
 	unsigned long combined_pfn;
 	struct page *buddy;
+	bool buddy_zeroed;
+	bool page_zeroed;
 	bool to_tail;
 
 	VM_BUG_ON(!zone_is_initialized(zone));
@@ -1029,6 +1031,8 @@ static inline void __free_one_page(struct page *page,
 				goto done_merging;
 		}
 
+		buddy_zeroed = PageZeroed(buddy);
+
 		/*
 		 * Our buddy is free or it is CONFIG_DEBUG_PAGEALLOC guard page,
 		 * merge with it and move up one order.
@@ -1047,10 +1051,17 @@ static inline void __free_one_page(struct page *page,
 			change_pageblock_range(buddy, order, migratetype);
 		}
 
+		page_zeroed = PageZeroed(page);
+		__ClearPageZeroed(page);
+		__ClearPageZeroed(buddy);
+
 		combined_pfn = buddy_pfn & pfn;
 		page = page + (combined_pfn - pfn);
 		pfn = combined_pfn;
 		order++;
+
+		if (page_zeroed && buddy_zeroed)
+			__SetPageZeroed(page);
 	}
 
 done_merging:
-- 
MST