1 files changed, 39 insertions, 49 deletions

diff --git a/block/blk-tag.c b/block/blk-tag.c
index 32667beb03e..a185b86741e 100644
--- a/block/blk-tag.c
+++ b/block/blk-tag.c
@@ -5,6 +5,7 @@
 #include <linux/module.h>
 #include <linux/bio.h>
 #include <linux/blkdev.h>
+#include <linux/slab.h>
 
 #include "blk.h"
 
@@ -26,19 +27,17 @@ struct request *blk_queue_find_tag(struct request_queue *q, int tag)
 EXPORT_SYMBOL(blk_queue_find_tag);
 
 /**
- * __blk_free_tags - release a given set of tag maintenance info
+ * blk_free_tags - release a given set of tag maintenance info
  * @bqt:	the tag map to free
  *
- * Tries to free the specified @bqt@.  Returns true if it was
- * actually freed and false if there are still references using it
+ * Drop the reference count on @bqt and frees it when the last reference
+ * is dropped.
  */
-static int __blk_free_tags(struct blk_queue_tag *bqt)
+void blk_free_tags(struct blk_queue_tag *bqt)
 {
-	int retval;
-
-	retval = atomic_dec_and_test(&bqt->refcnt);
-	if (retval) {
-		BUG_ON(bqt->busy);
+	if (atomic_dec_and_test(&bqt->refcnt)) {
+		BUG_ON(find_first_bit(bqt->tag_map, bqt->max_depth) <
+							bqt->max_depth);
 
 		kfree(bqt->tag_index);
 		bqt->tag_index = NULL;
@@ -48,9 +47,8 @@ static int __blk_free_tags(struct blk_queue_tag *bqt)
 
 		kfree(bqt);
 	}
-
-	return retval;
 }
+EXPORT_SYMBOL(blk_free_tags);
 
 /**
  * __blk_queue_free_tags - release tag maintenance info
@@ -67,33 +65,18 @@ void __blk_queue_free_tags(struct request_queue *q)
 	if (!bqt)
 		return;
 
-	__blk_free_tags(bqt);
+	blk_free_tags(bqt);
 
 	q->queue_tags = NULL;
 	queue_flag_clear_unlocked(QUEUE_FLAG_QUEUED, q);
 }
 
 /**
- * blk_free_tags - release a given set of tag maintenance info
- * @bqt:	the tag map to free
- *
- * For externally managed @bqt@ frees the map.  Callers of this
- * function must guarantee to have released all the queues that
- * might have been using this tag map.
- */
-void blk_free_tags(struct blk_queue_tag *bqt)
-{
-	if (unlikely(!__blk_free_tags(bqt)))
-		BUG();
-}
-EXPORT_SYMBOL(blk_free_tags);
-
-/**
  * blk_queue_free_tags - release tag maintenance info
  * @q:  the request queue for the device
  *
  *  Notes:
- *	This is used to disabled tagged queuing to a device, yet leave
+ *	This is used to disable tagged queuing to a device, yet leave
  *	queue in function.
  **/
 void blk_queue_free_tags(struct request_queue *q)
@@ -147,7 +130,6 @@ static struct blk_queue_tag *__blk_queue_init_tags(struct request_queue *q,
 	if (init_tag_map(q, tags, depth))
 		goto fail;
 
-	tags->busy = 0;
 	atomic_set(&tags->refcnt, 1);
 	return tags;
 fail:
@@ -158,7 +140,6 @@ fail:
 /**
  * blk_init_tags - initialize the tag info for an external tag map
  * @depth:	the maximum queue depth supported
- * @tags: the tag to use
  **/
 struct blk_queue_tag *blk_init_tags(int depth)
 {
@@ -186,7 +167,8 @@ int blk_queue_init_tags(struct request_queue *q, int depth,
 		tags = __blk_queue_init_tags(q, depth);
 
 		if (!tags)
-			goto fail;
+			return -ENOMEM;
+
 	} else if (q->queue_tags) {
 		rc = blk_queue_resize_tags(q, depth);
 		if (rc)
@@ -203,9 +185,6 @@ int blk_queue_init_tags(struct request_queue *q, int depth,
 	queue_flag_set_unlocked(QUEUE_FLAG_QUEUED, q);
 	INIT_LIST_HEAD(&q->tag_busy_list);
 	return 0;
-fail:
-	kfree(tags);
-	return -ENOMEM;
 }
 EXPORT_SYMBOL(blk_queue_init_tags);
 
@@ -271,7 +250,7 @@ EXPORT_SYMBOL(blk_queue_resize_tags);
  * @rq: the request that has completed
  *
  *  Description:
- *    Typically called when end_that_request_first() returns 0, meaning
+ *    Typically called when end_that_request_first() returns %0, meaning
  *    all transfers have been done for a request. It's important to call
  *    this function before end_that_request_last(), as that will put the
  *    request back on the free list thus corrupting the internal tag list.
@@ -282,16 +261,9 @@ EXPORT_SYMBOL(blk_queue_resize_tags);
 void blk_queue_end_tag(struct request_queue *q, struct request *rq)
 {
 	struct blk_queue_tag *bqt = q->queue_tags;
-	int tag = rq->tag;
-
-	BUG_ON(tag == -1);
+	unsigned tag = rq->tag; /* negative tags invalid */
 
-	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?
-		 */
-		return;
+	BUG_ON(tag >= bqt->real_max_depth);
 
 	list_del_init(&rq->queuelist);
 	rq->cmd_flags &= ~REQ_QUEUED;
@@ -313,7 +285,6 @@ void blk_queue_end_tag(struct request_queue *q, struct request *rq)
 	 * unlock memory barrier semantics.
 	 */
 	clear_bit_unlock(tag, bqt->tag_map);
-	bqt->busy--;
 }
 EXPORT_SYMBOL(blk_queue_end_tag);
 
@@ -338,6 +309,7 @@ EXPORT_SYMBOL(blk_queue_end_tag);
 int blk_queue_start_tag(struct request_queue *q, struct request *rq)
 {
 	struct blk_queue_tag *bqt = q->queue_tags;
+	unsigned max_depth;
 	int tag;
 
 	if (unlikely((rq->cmd_flags & REQ_QUEUED))) {
@@ -351,10 +323,29 @@ int blk_queue_start_tag(struct request_queue *q, struct request *rq)
 	/*
 	 * Protect against shared tag maps, as we may not have exclusive
 	 * access to the tag map.
+	 *
+	 * We reserve a few tags just for sync IO, since we don't want
+	 * to starve sync IO on behalf of flooding async IO.
 	 */
+	max_depth = bqt->max_depth;
+	if (!rq_is_sync(rq) && max_depth > 1) {
+		switch (max_depth) {
+		case 2:
+			max_depth = 1;
+			break;
+		case 3:
+			max_depth = 2;
+			break;
+		default:
+			max_depth -= 2;
+		}
+		if (q->in_flight[BLK_RW_ASYNC] > max_depth)
+			return 1;
+	}
+
 	do {
-		tag = find_first_zero_bit(bqt->tag_map, bqt->max_depth);
-		if (tag >= bqt->max_depth)
+		tag = find_first_zero_bit(bqt->tag_map, max_depth);
+		if (tag >= max_depth)
 			return 1;
 
 	} while (test_and_set_bit_lock(tag, bqt->tag_map));
@@ -366,9 +357,8 @@ int blk_queue_start_tag(struct request_queue *q, struct request *rq)
 	rq->cmd_flags |= REQ_QUEUED;
 	rq->tag = tag;
 	bqt->tag_index[tag] = rq;
-	blkdev_dequeue_request(rq);
+	blk_start_request(rq);
 	list_add(&rq->queuelist, &q->tag_busy_list);
-	bqt->busy++;
 	return 0;
 }
 EXPORT_SYMBOL(blk_queue_start_tag);