[PATCH] swap: scan_swap_map drop swap_device_lock

get_swap_page has often shown up on latency traces, doing lengthy scans while
holding two spinlocks.  swap_list_lock is already dropped, now scan_swap_map
drop swap_device_lock before scanning the swap_map.

While scanning for an empty cluster, don't worry that racing tasks may
allocate what was free and free what was allocated; but when allocating an
entry, check it's still free after retaking the lock.  Avoid dropping the lock
in the expected common path.  No barriers beyond the locks, just let the
cookie crumble; highest_bit limit is volatile, but benign.

Guard against swapoff: must check SWP_WRITEOK before allocating, must raise
SWP_SCANNING reference count while in scan_swap_map, swapoff wait for that to
fall - just use schedule_timeout, we don't want to burden scan_swap_map
itself, and it's very unlikely that anyone can really still be in
scan_swap_map once swapoff gets this far.

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/include/linux/swap.h b/include/linux/swap.h
index 93f0eca..db3b5de 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -107,6 +107,8 @@
 	SWP_USED	= (1 << 0),	/* is slot in swap_info[] used? */
 	SWP_WRITEOK	= (1 << 1),	/* ok to write to this swap?	*/
 	SWP_ACTIVE	= (SWP_USED | SWP_WRITEOK),
+					/* add others here before... */
+	SWP_SCANNING	= (1 << 8),	/* refcount in scan_swap_map */
 };
 
 #define SWAP_CLUSTER_MAX 32
diff --git a/mm/swapfile.c b/mm/swapfile.c
index c70248a..fdee145 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -98,10 +98,12 @@
 	 * But we do now try to find an empty cluster.  -Andrea
 	 */
 
+	si->flags += SWP_SCANNING;
 	if (unlikely(!si->cluster_nr)) {
 		si->cluster_nr = SWAPFILE_CLUSTER - 1;
 		if (si->pages - si->inuse_pages < SWAPFILE_CLUSTER)
 			goto lowest;
+		swap_device_unlock(si);
 
 		offset = si->lowest_bit;
 		last_in_cluster = offset + SWAPFILE_CLUSTER - 1;
@@ -111,10 +113,12 @@
 			if (si->swap_map[offset])
 				last_in_cluster = offset + SWAPFILE_CLUSTER;
 			else if (offset == last_in_cluster) {
+				swap_device_lock(si);
 				si->cluster_next = offset-SWAPFILE_CLUSTER-1;
 				goto cluster;
 			}
 		}
+		swap_device_lock(si);
 		goto lowest;
 	}
 
@@ -123,10 +127,12 @@
 	offset = si->cluster_next;
 	if (offset > si->highest_bit)
 lowest:		offset = si->lowest_bit;
+checks:	if (!(si->flags & SWP_WRITEOK))
+		goto no_page;
 	if (!si->highest_bit)
 		goto no_page;
 	if (!si->swap_map[offset]) {
-got_page:	if (offset == si->lowest_bit)
+		if (offset == si->lowest_bit)
 			si->lowest_bit++;
 		if (offset == si->highest_bit)
 			si->highest_bit--;
@@ -137,16 +143,22 @@
 		}
 		si->swap_map[offset] = 1;
 		si->cluster_next = offset + 1;
+		si->flags -= SWP_SCANNING;
 		return offset;
 	}
 
+	swap_device_unlock(si);
 	while (++offset <= si->highest_bit) {
-		if (!si->swap_map[offset])
-			goto got_page;
+		if (!si->swap_map[offset]) {
+			swap_device_lock(si);
+			goto checks;
+		}
 	}
+	swap_device_lock(si);
 	goto lowest;
 
 no_page:
+	si->flags -= SWP_SCANNING;
 	return 0;
 }
 
@@ -1111,10 +1123,6 @@
 	err = try_to_unuse(type);
 	current->flags &= ~PF_SWAPOFF;
 
-	/* wait for any unplug function to finish */
-	down_write(&swap_unplug_sem);
-	up_write(&swap_unplug_sem);
-
 	if (err) {
 		/* re-insert swap space back into swap_list */
 		swap_list_lock();
@@ -1128,10 +1136,28 @@
 			swap_info[prev].next = p - swap_info;
 		nr_swap_pages += p->pages;
 		total_swap_pages += p->pages;
+		swap_device_lock(p);
 		p->flags |= SWP_WRITEOK;
+		swap_device_unlock(p);
 		swap_list_unlock();
 		goto out_dput;
 	}
+
+	/* wait for any unplug function to finish */
+	down_write(&swap_unplug_sem);
+	up_write(&swap_unplug_sem);
+
+	/* wait for anyone still in scan_swap_map */
+	swap_device_lock(p);
+	p->highest_bit = 0;		/* cuts scans short */
+	while (p->flags >= SWP_SCANNING) {
+		swap_device_unlock(p);
+		set_current_state(TASK_UNINTERRUPTIBLE);
+		schedule_timeout(1);
+		swap_device_lock(p);
+	}
+	swap_device_unlock(p);
+
 	destroy_swap_extents(p);
 	down(&swapon_sem);
 	swap_list_lock();
@@ -1431,6 +1457,8 @@
 		}
 
 		p->lowest_bit  = 1;
+		p->cluster_next = 1;
+
 		/*
 		 * Find out how many pages are allowed for a single swap
 		 * device. There are two limiting factors: 1) the number of