[PATCH] blk: fix tag shrinking (revive real_max_size)

My patch in commit fa72b903f75e4f0f0b2c2feed093005167da4023 incorrectly
removed blk_queue_tag->real_max_depth.

The original resize implementation was incorrect in the following
points.

 * actual allocation size of tag_index was shorter than real_max_size,
   but assumed to be of the same size, possibly causing memory access
   beyond the allocated area.
 * bits in tag_map between max_deptn and real_max_depth were
   initialized to 1's, making the tags permanently reserved.

In an attempt to fix above two bugs, I had removed allocation optimization
in init_tag_map and real_max_size.  Tag map/index were allocated and freed
immediately during resize.

Unfortunately, I wasn't considering that tag map/index can be resized
dynamically with tags beyond new_depth active.  This led to accessing
freed area after shrinking tags and led to the following bug reporting
thread on linux-scsi.

   http://marc.theaimsgroup.com/?l=linux-scsi&m=112319898111885&w=2

To fix the problem, I've revived real_max_depth without allocation
optimization in init_tag_map, and Andrew Vasquez confirmed that the
problem was fixed.  As Jens is not going to be available for a week, he
asked me to make sure that this patch reaches you.

   http://marc.theaimsgroup.com/?l=linux-scsi&m=112325778530886&w=2

Also, a comment was added to make sure that real_max_size is needed for
dynamic shrinking.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

drivers/block/ll_rw_blk.c

@@ -719,7 +719,7 @@ struct request *blk_queue_find_tag(request_queue_t *q, int tag)
 {
 	struct blk_queue_tag *bqt = q->queue_tags;
 
-	if (unlikely(bqt == NULL || tag >= bqt->max_depth))
+	if (unlikely(bqt == NULL || tag >= bqt->real_max_depth))
 		return NULL;
 
 	return bqt->tag_index[tag];
@@ -798,6 +798,7 @@ init_tag_map(request_queue_t *q, struct blk_queue_tag *tags, int depth)
 
 	memset(tag_index, 0, depth * sizeof(struct request *));
 	memset(tag_map, 0, nr_ulongs * sizeof(unsigned long));
+	tags->real_max_depth = depth;
 	tags->max_depth = depth;
 	tags->tag_index = tag_index;
 	tags->tag_map = tag_map;
@@ -871,12 +872,23 @@ int blk_queue_resize_tags(request_queue_t *q, int new_depth)
 	if (!bqt)
 		return -ENXIO;
 
+	/*
+	 * if we already have large enough real_max_depth.  just
+	 * adjust max_depth.  *NOTE* as requests with tag value
+	 * between new_depth and real_max_depth can be in-flight, tag
+	 * map can not be shrunk blindly here.
+	 */
+	if (new_depth <= bqt->real_max_depth) {
+		bqt->max_depth = new_depth;
+		return 0;
+	}
+
 	/*
 	 * save the old state info, so we can copy it back
 	 */
 	tag_index = bqt->tag_index;
 	tag_map = bqt->tag_map;
-	max_depth = bqt->max_depth;
+	max_depth = bqt->real_max_depth;
 
 	if (init_tag_map(q, bqt, new_depth))
 		return -ENOMEM;
@@ -913,7 +925,7 @@ void blk_queue_end_tag(request_queue_t *q, struct request *rq)
 
 	BUG_ON(tag == -1);
 
-	if (unlikely(tag >= bqt->max_depth))
+	if (unlikely(tag >= bqt->real_max_depth))
 		/*
 		 * This can happen after tag depth has been reduced.
 		 * FIXME: how about a warning or info message here?

include/linux/blkdev.h

@@ -301,6 +301,7 @@ struct blk_queue_tag {
 	struct list_head busy_list;	/* fifo list of busy tags */
 	int busy;			/* current depth */
 	int max_depth;			/* what we will send to device */
+	int real_max_depth;		/* what the array can hold */
 	atomic_t refcnt;		/* map can be shared */
 };