From: Bo Liu <[email protected]>
In try_to_unuse(), swcount is a local copy of *swap_map, including the
SWAP_HAS_CACHE bit; but a wrong comparison against swap_count(*swap_map),
which masks off the SWAP_HAS_CACHE bit, succeeded where it should fail.
That had the effect of resetting the mm from which to start searching
for the next swap page, to an irrelevant mm instead of to an mm in which
this swap page had been found: which may increase search time by ~20%.
But we're used to swapoff being slow, so never noticed the slowdown.
Remove that one spurious use of swap_count(): Bo Liu thought it merely
redundant, Hugh rewrote the description since it was measurably wrong.
Signed-off-by: Bo Liu <[email protected]>
Signed-off-by: Hugh Dickins <[email protected]>
Reviewed-by: KAMEZAWA Hiroyuki <[email protected]>
Cc: [email protected]
---
mm/swapfile.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
--- 2.6.32-rc5/mm/swapfile.c 2009-10-05 04:20:31.000000000 +0100
+++ linux/mm/swapfile.c 2009-10-28 19:31:43.000000000 +0000
@@ -1151,8 +1151,7 @@ static int try_to_unuse(unsigned int typ
} else
retval = unuse_mm(mm, entry, page);
- if (set_start_mm &&
- swap_count(*swap_map) < swcount) {
+ if (set_start_mm && *swap_map < swcount) {
mmput(new_start_mm);
atomic_inc(&mm->mm_users);
new_start_mm = mm;