[v11,14/25] mm: Document why we don't set PageReadahead

Commit Message

Matthew Wilcox April 14, 2020, 3:02 p.m. UTC

From: "Matthew Wilcox (Oracle)" <[e-mail]>

If the page is already in cache, we don't set PageReadahead on it.

Signed-off-by: Matthew Wilcox (Oracle) <[e-mail]>
Reviewed-by: Christoph Hellwig <[e-mail]>
Reviewed-by: William Kucharski <[e-mail]>
---
 mm/readahead.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

Patch

diff --git a/mm/readahead.c b/mm/readahead.c
index ae231a5312cb..73cb59ed5cff 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -195,9 +195,12 @@  void page_cache_readahead_unbounded(struct address_space *mapping,
 
 		if (page && !xa_is_value(page)) {
 			/*
-			 * Page already present?  Kick off the current batch of
-			 * contiguous pages before continuing with the next
-			 * batch.
+			 * Page already present?  Kick off the current batch
+			 * of contiguous pages before continuing with the
+			 * next batch.  This page may be the one we would
+			 * have intended to mark as Readahead, but we don't
+			 * have a stable reference to this page, and it's
+			 * not worth getting one just for that.
 			 */
 			read_pages(&rac, &page_pool, true);
 			continue;