count_history_pages does only call page_cache_prev_hole in rcu_lock
context using address_space mapping. There's no need to have file_ra_state
here.
Cc: Andrew Morton <[email protected]>
Cc: Fengguang Wu <[email protected]>
Signed-off-by: Fabian Frederick <[email protected]>
---
mm/readahead.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/mm/readahead.c b/mm/readahead.c
index 0ca36a7..17b9172 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -326,7 +326,6 @@ static unsigned long get_next_ra_size(struct file_ra_state *ra,
* - thrashing threshold in memory tight systems
*/
static pgoff_t count_history_pages(struct address_space *mapping,
- struct file_ra_state *ra,
pgoff_t offset, unsigned long max)
{
pgoff_t head;
@@ -349,7 +348,7 @@ static int try_context_readahead(struct address_space *mapping,
{
pgoff_t size;
- size = count_history_pages(mapping, ra, offset, max);
+ size = count_history_pages(mapping, offset, max);
/*
* not enough history pages:
--
1.8.4.5
On Tue, 10 Jun 2014, Fabian Frederick wrote:
> count_history_pages does only call page_cache_prev_hole in rcu_lock
> context using address_space mapping. There's no need to have file_ra_state
> here.
>
> Cc: Andrew Morton <[email protected]>
> Cc: Fengguang Wu <[email protected]>
> Signed-off-by: Fabian Frederick <[email protected]>
Acked-by: David Rientjes <[email protected]>
Acked-by: Fengguang Wu <[email protected]>
Thank you!