[Intel-gfx] [PATCH v2 4/8] proc: Optimise smaps for shmem entries

Matthew Wilcox (Oracle) willy at infradead.org
Thu Sep 10 18:33:14 UTC 2020


Avoid bumping the refcount on pages when we're only interested in the
swap entries.

Signed-off-by: Matthew Wilcox (Oracle) <willy at infradead.org>
Acked-by: Johannes Weiner <hannes at cmpxchg.org>
---
 fs/proc/task_mmu.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index 5066b0251ed8..e42d9e5e9a3c 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -520,16 +520,10 @@ static void smaps_pte_entry(pte_t *pte, unsigned long addr,
 			page = device_private_entry_to_page(swpent);
 	} else if (unlikely(IS_ENABLED(CONFIG_SHMEM) && mss->check_shmem_swap
 							&& pte_none(*pte))) {
-		page = find_get_entry(vma->vm_file->f_mapping,
+		page = xa_load(&vma->vm_file->f_mapping->i_pages,
 						linear_page_index(vma, addr));
-		if (!page)
-			return;
-
 		if (xa_is_value(page))
 			mss->swap += PAGE_SIZE;
-		else
-			put_page(page);
-
 		return;
 	}
 
-- 
2.28.0



More information about the Intel-gfx mailing list