block: change the request allocation/congestion logic to be sync/async based

This makes sure that we never wait on async IO for sync requests, instead
of doing the split on writes vs reads.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

2 files changed, 41 insertions, 23 deletions

diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h
index bee52abb8a4d..0ec2c594868e 100644
--- a/include/linux/backing-dev.h
+++ b/include/linux/backing-dev.h
@@ -24,8 +24,8 @@ struct dentry;
  */
 enum bdi_state {
 	BDI_pdflush,		/* A pdflush thread is working this device */
-	BDI_write_congested,	/* The write queue is getting full */
-	BDI_read_congested,	/* The read queue is getting full */
+	BDI_async_congested,	/* The async (write) queue is getting full */
+	BDI_sync_congested,	/* The sync queue is getting full */
 	BDI_unused,		/* Available bits start here */
 };
 
@@ -215,18 +215,18 @@ static inline int bdi_congested(struct backing_dev_info *bdi, int bdi_bits)
 
 static inline int bdi_read_congested(struct backing_dev_info *bdi)
 {
-	return bdi_congested(bdi, 1 << BDI_read_congested);
+	return bdi_congested(bdi, 1 << BDI_sync_congested);
 }
 
 static inline int bdi_write_congested(struct backing_dev_info *bdi)
 {
-	return bdi_congested(bdi, 1 << BDI_write_congested);
+	return bdi_congested(bdi, 1 << BDI_async_congested);
 }
 
 static inline int bdi_rw_congested(struct backing_dev_info *bdi)
 {
-	return bdi_congested(bdi, (1 << BDI_read_congested)|
-				  (1 << BDI_write_congested));
+	return bdi_congested(bdi, (1 << BDI_sync_congested) |
+				  (1 << BDI_async_congested));
 }
 
 void clear_bdi_congested(struct backing_dev_info *bdi, int rw);
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 465d6babc847..67dae3bd881c 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -38,6 +38,10 @@ struct request;
 typedef void (rq_end_io_fn)(struct request *, int);
 
 struct request_list {
+	/*
+	 * count[], starved[], and wait[] are indexed by
+	 * BLK_RW_SYNC/BLK_RW_ASYNC
+	 */
 	int count[2];
 	int starved[2];
 	int elvpriv;
@@ -66,6 +70,11 @@ enum rq_cmd_type_bits {
 	REQ_TYPE_ATA_PC,
 };
 
+enum {
+	BLK_RW_ASYNC	= 0,
+	BLK_RW_SYNC	= 1,
+};
+
 /*
  * For request of type REQ_TYPE_LINUX_BLOCK, rq->cmd[0] is the opcode being
  * sent down (similar to how REQ_TYPE_BLOCK_PC means that ->cmd[] holds a
@@ -103,7 +112,7 @@ enum rq_flag_bits {
 	__REQ_QUIET,		/* don't worry about errors */
 	__REQ_PREEMPT,		/* set for "ide_preempt" requests */
 	__REQ_ORDERED_COLOR,	/* is before or after barrier */
-	__REQ_RW_SYNC,		/* request is sync (O_DIRECT) */
+	__REQ_RW_SYNC,		/* request is sync (sync write or read) */
 	__REQ_ALLOCED,		/* request came from our alloc pool */
 	__REQ_RW_META,		/* metadata io request */
 	__REQ_COPY_USER,	/* contains copies of user pages */
@@ -438,8 +447,8 @@ struct request_queue
 #define QUEUE_FLAG_CLUSTER	0	/* cluster several segments into 1 */
 #define QUEUE_FLAG_QUEUED	1	/* uses generic tag queueing */
 #define QUEUE_FLAG_STOPPED	2	/* queue is stopped */
-#define	QUEUE_FLAG_READFULL	3	/* read queue has been filled */
-#define QUEUE_FLAG_WRITEFULL	4	/* write queue has been filled */
+#define	QUEUE_FLAG_SYNCFULL	3	/* read queue has been filled */
+#define QUEUE_FLAG_ASYNCFULL	4	/* write queue has been filled */
 #define QUEUE_FLAG_DEAD		5	/* queue being torn down */
 #define QUEUE_FLAG_REENTER	6	/* Re-entrancy avoidance */
 #define QUEUE_FLAG_PLUGGED	7	/* queue is plugged */
@@ -611,32 +620,41 @@ enum {
 #define rq_data_dir(rq)		((rq)->cmd_flags & 1)
 
 /*
- * We regard a request as sync, if it's a READ or a SYNC write.
+ * We regard a request as sync, if either a read or a sync write
  */
-#define rq_is_sync(rq)		(rq_data_dir((rq)) == READ || (rq)->cmd_flags & REQ_RW_SYNC)
+static inline bool rw_is_sync(unsigned int rw_flags)
+{
+	return !(rw_flags & REQ_RW) || (rw_flags & REQ_RW_SYNC);
+}
+
+static inline bool rq_is_sync(struct request *rq)
+{
+	return rw_is_sync(rq->cmd_flags);
+}
+
 #define rq_is_meta(rq)		((rq)->cmd_flags & REQ_RW_META)
 
-static inline int blk_queue_full(struct request_queue *q, int rw)
+static inline int blk_queue_full(struct request_queue *q, int sync)
 {
-	if (rw == READ)
-		return test_bit(QUEUE_FLAG_READFULL, &q->queue_flags);
-	return test_bit(QUEUE_FLAG_WRITEFULL, &q->queue_flags);
+	if (sync)
+		return test_bit(QUEUE_FLAG_SYNCFULL, &q->queue_flags);
+	return test_bit(QUEUE_FLAG_ASYNCFULL, &q->queue_flags);
 }
 
-static inline void blk_set_queue_full(struct request_queue *q, int rw)
+static inline void blk_set_queue_full(struct request_queue *q, int sync)
 {
-	if (rw == READ)
-		queue_flag_set(QUEUE_FLAG_READFULL, q);
+	if (sync)
+		queue_flag_set(QUEUE_FLAG_SYNCFULL, q);
 	else
-		queue_flag_set(QUEUE_FLAG_WRITEFULL, q);
+		queue_flag_set(QUEUE_FLAG_ASYNCFULL, q);
 }
 
-static inline void blk_clear_queue_full(struct request_queue *q, int rw)
+static inline void blk_clear_queue_full(struct request_queue *q, int sync)
 {
-	if (rw == READ)
-		queue_flag_clear(QUEUE_FLAG_READFULL, q);
+	if (sync)
+		queue_flag_clear(QUEUE_FLAG_SYNCFULL, q);
 	else
-		queue_flag_clear(QUEUE_FLAG_WRITEFULL, q);
+		queue_flag_clear(QUEUE_FLAG_ASYNCFULL, q);
 }