Merge branch 'for-linus' of ../linux-2.6-block into block-for-2.6.39/core

This merge creates two set of conflicts.  One is simple context
conflicts caused by removal of throtl_scheduled_delayed_work() in
for-linus and removal of throtl_shutdown_timer_wq() in
for-2.6.39/core.

The other is caused by commit 255bb490c8 (block: blk-flush shouldn't
call directly into q->request_fn() __blk_run_queue()) in for-linus
crashing with FLUSH reimplementation in for-2.6.39/core.  The conflict
isn't trivial but the resolution is straight-forward.

* __blk_run_queue() calls in flush_end_io() and flush_data_end_io()
  should be called with @force_kblockd set to %true.

* elv_insert() in blk_kick_flush() should use
  %ELEVATOR_INSERT_REQUEUE.

Both changes are to avoid invoking ->request_fn() directly from
request completion path and closely match the changes in the commit
255bb490c8.

Signed-off-by: Tejun Heo <tj@kernel.org>

8 files changed, 49 insertions, 38 deletions

diff --git a/block/blk-core.c b/block/blk-core.c
index accff29ad67..74d496ccf4d 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -342,7 +342,7 @@ void blk_start_queue(struct request_queue *q)
 	WARN_ON(!irqs_disabled());
 
 	queue_flag_clear(QUEUE_FLAG_STOPPED, q);
-	__blk_run_queue(q);
+	__blk_run_queue(q, false);
 }
 EXPORT_SYMBOL(blk_start_queue);
 
@@ -396,13 +396,14 @@ EXPORT_SYMBOL(blk_sync_queue);
 /**
  * __blk_run_queue - run a single device queue
  * @q:	The queue to run
+ * @force_kblockd: Don't run @q->request_fn directly.  Use kblockd.
  *
  * Description:
  *    See @blk_run_queue. This variant must be called with the queue lock
  *    held and interrupts disabled.
  *
  */
-void __blk_run_queue(struct request_queue *q)
+void __blk_run_queue(struct request_queue *q, bool force_kblockd)
 {
 	blk_remove_plug(q);
 
@@ -416,7 +417,7 @@ void __blk_run_queue(struct request_queue *q)
 	 * Only recurse once to avoid overrunning the stack, let the unplug
 	 * handling reinvoke the handler shortly if we already got there.
 	 */
-	if (!queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) {
+	if (!force_kblockd && !queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) {
 		q->request_fn(q);
 		queue_flag_clear(QUEUE_FLAG_REENTER, q);
 	} else {
@@ -439,7 +440,7 @@ void blk_run_queue(struct request_queue *q)
 	unsigned long flags;
 
 	spin_lock_irqsave(q->queue_lock, flags);
-	__blk_run_queue(q);
+	__blk_run_queue(q, false);
 	spin_unlock_irqrestore(q->queue_lock, flags);
 }
 EXPORT_SYMBOL(blk_run_queue);
@@ -1085,7 +1086,7 @@ void blk_insert_request(struct request_queue *q, struct request *rq,
 
 	drive_stat_acct(rq, 1);
 	__elv_add_request(q, rq, where, 0);
-	__blk_run_queue(q);
+	__blk_run_queue(q, false);
 	spin_unlock_irqrestore(q->queue_lock, flags);
 }
 EXPORT_SYMBOL(blk_insert_request);
@@ -2642,13 +2643,6 @@ int kblockd_schedule_work(struct request_queue *q, struct work_struct *work)
 }
 EXPORT_SYMBOL(kblockd_schedule_work);
 
-int kblockd_schedule_delayed_work(struct request_queue *q,
-			struct delayed_work *dwork, unsigned long delay)
-{
-	return queue_delayed_work(kblockd_workqueue, dwork, delay);
-}
-EXPORT_SYMBOL(kblockd_schedule_delayed_work);
-
 int __init blk_dev_init(void)
 {
 	BUILD_BUG_ON(__REQ_NR_BITS > 8 *
diff --git a/block/blk-flush.c b/block/blk-flush.c
index a867e3f524f..0bd8c9c5d6e 100644
--- a/block/blk-flush.c
+++ b/block/blk-flush.c
@@ -212,9 +212,14 @@ static void flush_end_io(struct request *flush_rq, int error)
 		queued |= blk_flush_complete_seq(rq, seq, error);
 	}
 
-	/* after populating an empty queue, kick it to avoid stall */
+	/*
+	 * Moving a request silently to empty queue_head may stall the
+	 * queue.  Kick the queue in those cases.  This function is called
+	 * from request completion path and calling directly into
+	 * request_fn may confuse the driver.  Always use kblockd.
+	 */
 	if (queued && was_empty)
-		__blk_run_queue(q);
+		__blk_run_queue(q, true);
 }
 
 /**
@@ -257,7 +262,7 @@ static bool blk_kick_flush(struct request_queue *q)
 	q->flush_rq.end_io = flush_end_io;
 
 	q->flush_pending_idx ^= 1;
-	elv_insert(q, &q->flush_rq, ELEVATOR_INSERT_FRONT);
+	elv_insert(q, &q->flush_rq, ELEVATOR_INSERT_REQUEUE);
 	return true;
 }
 
@@ -266,9 +271,12 @@ static void flush_data_end_io(struct request *rq, int error)
 	struct request_queue *q = rq->q;
 	bool was_empty = elv_queue_empty(q);
 
-	/* after populating an empty queue, kick it to avoid stall */
+	/*
+	 * After populating an empty queue, kick it to avoid stall.  Read
+	 * the comment in flush_end_io().
+	 */
 	if (blk_flush_complete_seq(rq, REQ_FSEQ_DATA, error) && was_empty)
-		__blk_run_queue(q);
+		__blk_run_queue(q, true);
 }
 
 /**
diff --git a/block/blk-lib.c b/block/blk-lib.c
index 1a320d2406b..eec78becb35 100644
--- a/block/blk-lib.c
+++ b/block/blk-lib.c
@@ -132,7 +132,7 @@ static void bio_batch_end_io(struct bio *bio, int err)
 }
 
 /**
- * blkdev_issue_zeroout generate number of zero filed write bios
+ * blkdev_issue_zeroout - generate number of zero filed write bios
  * @bdev:	blockdev to issue
  * @sector:	start sector
  * @nr_sects:	number of sectors to write
diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index c0f62374216..061dee66e2a 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -20,6 +20,11 @@ static int throtl_quantum = 32;
 /* Throttling is performed over 100ms slice and after that slice is renewed */
 static unsigned long throtl_slice = HZ/10;	/* 100 ms */
 
+/* A workqueue to queue throttle related work */
+static struct workqueue_struct *kthrotld_workqueue;
+static void throtl_schedule_delayed_work(struct throtl_data *td,
+				unsigned long delay);
+
 struct throtl_rb_root {
 	struct rb_root rb;
 	struct rb_node *left;
@@ -345,10 +350,9 @@ static void throtl_schedule_next_dispatch(struct throtl_data *td)
 	update_min_dispatch_time(st);
 
 	if (time_before_eq(st->min_disptime, jiffies))
-		throtl_schedule_delayed_work(td->queue, 0);
+		throtl_schedule_delayed_work(td, 0);
 	else
-		throtl_schedule_delayed_work(td->queue,
-				(st->min_disptime - jiffies));
+		throtl_schedule_delayed_work(td, (st->min_disptime - jiffies));
 }
 
 static inline void
@@ -815,10 +819,10 @@ void blk_throtl_work(struct work_struct *work)
 }
 
 /* Call with queue lock held */
-void throtl_schedule_delayed_work(struct request_queue *q, unsigned long delay)
+static void
+throtl_schedule_delayed_work(struct throtl_data *td, unsigned long delay)
 {
 
-	struct throtl_data *td = q->td;
 	struct delayed_work *dwork = &td->throtl_work;
 
 	if (total_nr_queued(td) > 0) {
@@ -827,12 +831,11 @@ void throtl_schedule_delayed_work(struct request_queue *q, unsigned long delay)
 		 * Cancel that and schedule a new one.
 		 */
 		__cancel_delayed_work(dwork);
-		kblockd_schedule_delayed_work(q, dwork, delay);
+		queue_delayed_work(kthrotld_workqueue, dwork, delay);
 		throtl_log(td, "schedule work. delay=%lu jiffies=%lu",
 				delay, jiffies);
 	}
 }
-EXPORT_SYMBOL(throtl_schedule_delayed_work);
 
 static void
 throtl_destroy_tg(struct throtl_data *td, struct throtl_grp *tg)
@@ -920,7 +923,7 @@ static void throtl_update_blkio_group_read_bps(void *key,
 	smp_mb__after_atomic_inc();
 
 	/* Schedule a work now to process the limit change */
-	throtl_schedule_delayed_work(td->queue, 0);
+	throtl_schedule_delayed_work(td, 0);
 }
 
 static void throtl_update_blkio_group_write_bps(void *key,
@@ -934,7 +937,7 @@ static void throtl_update_blkio_group_write_bps(void *key,
 	smp_mb__before_atomic_inc();
 	atomic_inc(&td->limits_changed);
 	smp_mb__after_atomic_inc();
-	throtl_schedule_delayed_work(td->queue, 0);
+	throtl_schedule_delayed_work(td, 0);
 }
 
 static void throtl_update_blkio_group_read_iops(void *key,
@@ -948,7 +951,7 @@ static void throtl_update_blkio_group_read_iops(void *key,
 	smp_mb__before_atomic_inc();
 	atomic_inc(&td->limits_changed);
 	smp_mb__after_atomic_inc();
-	throtl_schedule_delayed_work(td->queue, 0);
+	throtl_schedule_delayed_work(td, 0);
 }
 
 static void throtl_update_blkio_group_write_iops(void *key,
@@ -962,7 +965,7 @@ static void throtl_update_blkio_group_write_iops(void *key,
 	smp_mb__before_atomic_inc();
 	atomic_inc(&td->limits_changed);
 	smp_mb__after_atomic_inc();
-	throtl_schedule_delayed_work(td->queue, 0);
+	throtl_schedule_delayed_work(td, 0);
 }
 
 static void throtl_shutdown_wq(struct request_queue *q)
@@ -1135,6 +1138,10 @@ void blk_throtl_exit(struct request_queue *q)
 
 static int __init throtl_init(void)
 {
+	kthrotld_workqueue = alloc_workqueue("kthrotld", WQ_MEM_RECLAIM, 0);
+	if (!kthrotld_workqueue)
+		panic("Failed to create kthrotld\n");
+
 	blkio_policy_register(&blkio_policy_throtl);
 	return 0;
 }
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index 3202c7e87fb..89dc745c7d9 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -3324,7 +3324,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq,
 			    cfqd->busy_queues > 1) {
 				cfq_del_timer(cfqd, cfqq);
 				cfq_clear_cfqq_wait_request(cfqq);
-				__blk_run_queue(cfqd->queue);
+				__blk_run_queue(cfqd->queue, false);
 			} else {
 				cfq_blkiocg_update_idle_time_stats(
 						&cfqq->cfqg->blkg);
@@ -3339,7 +3339,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq,
 		 * this new queue is RT and the current one is BE
 		 */
 		cfq_preempt_queue(cfqd, cfqq);
-		__blk_run_queue(cfqd->queue);
+		__blk_run_queue(cfqd->queue, false);
 	}
 }
 
@@ -3700,7 +3700,7 @@ static void cfq_kick_queue(struct work_struct *work)
 	struct request_queue *q = cfqd->queue;
 
 	spin_lock_irq(q->queue_lock);
-	__blk_run_queue(cfqd->queue);
+	__blk_run_queue(cfqd->queue, false);
 	spin_unlock_irq(q->queue_lock);
 }
 
diff --git a/block/elevator.c b/block/elevator.c
index f98e92edc93..fabf3675c91 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -602,7 +602,7 @@ void elv_quiesce_start(struct request_queue *q)
 	 */
 	elv_drain_elevator(q);
 	while (q->rq.elvpriv) {
-		__blk_run_queue(q);
+		__blk_run_queue(q, false);
 		spin_unlock_irq(q->queue_lock);
 		msleep(10);
 		spin_lock_irq(q->queue_lock);
@@ -651,7 +651,7 @@ void elv_insert(struct request_queue *q, struct request *rq, int where)
 		 *   with anything.  There's no point in delaying queue
 		 *   processing.
 		 */
-		__blk_run_queue(q);
+		__blk_run_queue(q, false);
 		break;
 
 	case ELEVATOR_INSERT_SORT:
diff --git a/block/genhd.c b/block/genhd.c
index 73d85a8e3c8..3e2b57b55e3 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -1355,7 +1355,7 @@ int invalidate_partition(struct gendisk *disk, int partno)
 	struct block_device *bdev = bdget_disk(disk, partno);
 	if (bdev) {
 		fsync_bdev(bdev);
-		res = __invalidate_device(bdev);
+		res = __invalidate_device(bdev, true);
 		bdput(bdev);
 	}
 	return res;
diff --git a/block/ioctl.c b/block/ioctl.c
index 9049d460fa8..1124cd29726 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -294,9 +294,11 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd,
 			return -EINVAL;
 		if (get_user(n, (int __user *) arg))
 			return -EFAULT;
-		if (!(mode & FMODE_EXCL) &&
-		    blkdev_get(bdev, mode | FMODE_EXCL, &bdev) < 0)
-			return -EBUSY;
+		if (!(mode & FMODE_EXCL)) {
+			bdgrab(bdev);
+			if (blkdev_get(bdev, mode | FMODE_EXCL, &bdev) < 0)
+				return -EBUSY;
+		}
 		ret = set_blocksize(bdev, n);
 		if (!(mode & FMODE_EXCL))
 			blkdev_put(bdev, mode | FMODE_EXCL);