[PATCH] swap: scan_swap_map latency breaks

The get_swap_page/scan_swap_map latency can be so bad that even those without
preemption configured deserve relief: periodically cond_resched.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

diff --git a/mm/swapfile.c b/mm/swapfile.c
index fdee145..e675ae5 100644 (file)
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -56,8 +56,6 @@ static DECLARE_MUTEX(swapon_sem);
  */
 static DECLARE_RWSEM(swap_unplug_sem);
 
-#define SWAPFILE_CLUSTER 256
-
 void swap_unplug_io_fn(struct backing_dev_info *unused_bdi, struct page *page)
 {
        swp_entry_t entry;
@@ -84,9 +82,13 @@ void swap_unplug_io_fn(struct backing_dev_info *unused_bdi, struct page *page)
        up_read(&swap_unplug_sem);
 }
 
+#define SWAPFILE_CLUSTER       256
+#define LATENCY_LIMIT          256
+
 static inline unsigned long scan_swap_map(struct swap_info_struct *si)
 {
        unsigned long offset, last_in_cluster;
+       int latency_ration = LATENCY_LIMIT;
 
        /* 
         * We try to cluster swap pages by allocating them sequentially
@@ -117,6 +119,10 @@ static inline unsigned long scan_swap_map(struct swap_info_struct *si)
                                si->cluster_next = offset-SWAPFILE_CLUSTER-1;
                                goto cluster;
                        }
+                       if (unlikely(--latency_ration < 0)) {
+                               cond_resched();
+                               latency_ration = LATENCY_LIMIT;
+                       }
                }
                swap_device_lock(si);
                goto lowest;
@@ -153,6 +159,10 @@ checks:    if (!(si->flags & SWP_WRITEOK))
                        swap_device_lock(si);
                        goto checks;
                }
+               if (unlikely(--latency_ration < 0)) {
+                       cond_resched();
+                       latency_ration = LATENCY_LIMIT;
+               }
        }
        swap_device_lock(si);
        goto lowest;