diff --git a/block/blk-core.c b/block/blk-core.c
index d6128d9a..e0a5ee3 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -56,25 +56,26 @@
 {
 	struct hd_struct *part;
 	int rw = rq_data_dir(rq);
+	int cpu;
 
 	if (!blk_fs_request(rq) || !rq->rq_disk)
 		return;
 
-	rcu_read_lock();
-
+	cpu = disk_stat_lock();
 	part = disk_map_sector_rcu(rq->rq_disk, rq->sector);
+
 	if (!new_io)
-		__all_stat_inc(rq->rq_disk, part, merges[rw], rq->sector);
+		all_stat_inc(cpu, rq->rq_disk, part, merges[rw], rq->sector);
 	else {
-		disk_round_stats(rq->rq_disk);
+		disk_round_stats(cpu, rq->rq_disk);
 		rq->rq_disk->in_flight++;
 		if (part) {
-			part_round_stats(part);
+			part_round_stats(cpu, part);
 			part->in_flight++;
 		}
 	}
 
-	rcu_read_unlock();
+	disk_stat_unlock();
 }
 
 void blk_queue_congestion_threshold(struct request_queue *q)
@@ -997,7 +998,7 @@
  * /proc/diskstats.  This accounts immediately for all queue usage up to
  * the current jiffies and restarts the counters again.
  */
-void disk_round_stats(struct gendisk *disk)
+void disk_round_stats(int cpu, struct gendisk *disk)
 {
 	unsigned long now = jiffies;
 
@@ -1005,15 +1006,15 @@
 		return;
 
 	if (disk->in_flight) {
-		__disk_stat_add(disk, time_in_queue,
-				disk->in_flight * (now - disk->stamp));
-		__disk_stat_add(disk, io_ticks, (now - disk->stamp));
+		disk_stat_add(cpu, disk, time_in_queue,
+			      disk->in_flight * (now - disk->stamp));
+		disk_stat_add(cpu, disk, io_ticks, (now - disk->stamp));
 	}
 	disk->stamp = now;
 }
 EXPORT_SYMBOL_GPL(disk_round_stats);
 
-void part_round_stats(struct hd_struct *part)
+void part_round_stats(int cpu, struct hd_struct *part)
 {
 	unsigned long now = jiffies;
 
@@ -1021,9 +1022,9 @@
 		return;
 
 	if (part->in_flight) {
-		__part_stat_add(part, time_in_queue,
-				part->in_flight * (now - part->stamp));
-		__part_stat_add(part, io_ticks, (now - part->stamp));
+		part_stat_add(cpu, part, time_in_queue,
+			      part->in_flight * (now - part->stamp));
+		part_stat_add(cpu, part, io_ticks, (now - part->stamp));
 	}
 	part->stamp = now;
 }
@@ -1563,12 +1564,13 @@
 	if (blk_fs_request(req) && req->rq_disk) {
 		const int rw = rq_data_dir(req);
 		struct hd_struct *part;
+		int cpu;
 
-		rcu_read_lock();
+		cpu = disk_stat_lock();
 		part = disk_map_sector_rcu(req->rq_disk, req->sector);
-		all_stat_add(req->rq_disk, part, sectors[rw],
-				nr_bytes >> 9, req->sector);
-		rcu_read_unlock();
+		all_stat_add(cpu, req->rq_disk, part, sectors[rw],
+			     nr_bytes >> 9, req->sector);
+		disk_stat_unlock();
 	}
 
 	total_bytes = bio_nbytes = 0;
@@ -1753,21 +1755,21 @@
 		unsigned long duration = jiffies - req->start_time;
 		const int rw = rq_data_dir(req);
 		struct hd_struct *part;
+		int cpu;
 
-		rcu_read_lock();
-
+		cpu = disk_stat_lock();
 		part = disk_map_sector_rcu(disk, req->sector);
 
-		__all_stat_inc(disk, part, ios[rw], req->sector);
-		__all_stat_add(disk, part, ticks[rw], duration, req->sector);
-		disk_round_stats(disk);
+		all_stat_inc(cpu, disk, part, ios[rw], req->sector);
+		all_stat_add(cpu, disk, part, ticks[rw], duration, req->sector);
+		disk_round_stats(cpu, disk);
 		disk->in_flight--;
 		if (part) {
-			part_round_stats(part);
+			part_round_stats(cpu, part);
 			part->in_flight--;
 		}
 
-		rcu_read_unlock();
+		disk_stat_unlock();
 	}
 
 	if (req->end_io)
diff --git a/block/blk-merge.c b/block/blk-merge.c
index eb2a3ca..d926a24 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -388,18 +388,19 @@
 
 	if (req->rq_disk) {
 		struct hd_struct *part;
+		int cpu;
 
-		rcu_read_lock();
-
+		cpu = disk_stat_lock();
 		part = disk_map_sector_rcu(req->rq_disk, req->sector);
-		disk_round_stats(req->rq_disk);
+
+		disk_round_stats(cpu, req->rq_disk);
 		req->rq_disk->in_flight--;
 		if (part) {
-			part_round_stats(part);
+			part_round_stats(cpu, part);
 			part->in_flight--;
 		}
 
-		rcu_read_unlock();
+		disk_stat_unlock();
 	}
 
 	req->ioprio = ioprio_best(req->ioprio, next->ioprio);
diff --git a/block/genhd.c b/block/genhd.c
index b431d65..430626e 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -633,10 +633,11 @@
 			      struct device_attribute *attr, char *buf)
 {
 	struct gendisk *disk = dev_to_disk(dev);
+	int cpu;
 
-	preempt_disable();
-	disk_round_stats(disk);
-	preempt_enable();
+	cpu = disk_stat_lock();
+	disk_round_stats(cpu, disk);
+	disk_stat_unlock();
 	return sprintf(buf,
 		"%8lu %8lu %8llu %8u "
 		"%8lu %8lu %8llu %8u "
@@ -749,6 +750,7 @@
 	struct disk_part_iter piter;
 	struct hd_struct *hd;
 	char buf[BDEVNAME_SIZE];
+	int cpu;
 
 	/*
 	if (&gp->dev.kobj.entry == block_class.devices.next)
@@ -758,9 +760,9 @@
 				"\n\n");
 	*/
  
-	preempt_disable();
-	disk_round_stats(gp);
-	preempt_enable();
+	cpu = disk_stat_lock();
+	disk_round_stats(cpu, gp);
+	disk_stat_unlock();
 	seq_printf(seqf, "%4d %4d %s %lu %lu %llu %u %lu %lu %llu %u %u %u %u\n",
 		MAJOR(disk_devt(gp)), MINOR(disk_devt(gp)),
 		disk_name(gp, 0, buf),
@@ -777,9 +779,9 @@
 	/* now show all non-0 size partitions of it */
 	disk_part_iter_init(&piter, gp, 0);
 	while ((hd = disk_part_iter_next(&piter))) {
-		preempt_disable();
-		part_round_stats(hd);
-		preempt_enable();
+		cpu = disk_stat_lock();
+		part_round_stats(cpu, hd);
+		disk_stat_unlock();
 		seq_printf(seqf, "%4d %4d %s %lu %lu %llu "
 			   "%u %lu %lu %llu %u %u %u %u\n",
 			   MAJOR(part_devt(hd)), MINOR(part_devt(hd)),
diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c
index 84c03d6..17eed8c 100644
--- a/drivers/block/aoe/aoecmd.c
+++ b/drivers/block/aoe/aoecmd.c
@@ -756,16 +756,17 @@
 	unsigned long n_sect = bio->bi_size >> 9;
 	const int rw = bio_data_dir(bio);
 	struct hd_struct *part;
+	int cpu;
 
-	rcu_read_lock();
-
+	cpu = disk_stat_lock();
 	part = disk_map_sector_rcu(disk, sector);
-	all_stat_inc(disk, part, ios[rw], sector);
-	all_stat_add(disk, part, ticks[rw], duration, sector);
-	all_stat_add(disk, part, sectors[rw], n_sect, sector);
-	all_stat_add(disk, part, io_ticks, duration, sector);
 
-	rcu_read_unlock();
+	all_stat_inc(cpu, disk, part, ios[rw], sector);
+	all_stat_add(cpu, disk, part, ticks[rw], duration, sector);
+	all_stat_add(cpu, disk, part, sectors[rw], n_sect, sector);
+	all_stat_add(cpu, disk, part, io_ticks, duration, sector);
+
+	disk_stat_unlock();
 }
 
 void
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index a78caad..6536247 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -377,12 +377,13 @@
 static void start_io_acct(struct dm_io *io)
 {
 	struct mapped_device *md = io->md;
+	int cpu;
 
 	io->start_time = jiffies;
 
-	preempt_disable();
-	disk_round_stats(dm_disk(md));
-	preempt_enable();
+	cpu = disk_stat_lock();
+	disk_round_stats(cpu, dm_disk(md));
+	disk_stat_unlock();
 	dm_disk(md)->in_flight = atomic_inc_return(&md->pending);
 }
 
@@ -391,15 +392,15 @@
 	struct mapped_device *md = io->md;
 	struct bio *bio = io->bio;
 	unsigned long duration = jiffies - io->start_time;
-	int pending;
+	int pending, cpu;
 	int rw = bio_data_dir(bio);
 
-	preempt_disable();
-	disk_round_stats(dm_disk(md));
-	preempt_enable();
-	dm_disk(md)->in_flight = pending = atomic_dec_return(&md->pending);
+	cpu = disk_stat_lock();
+	disk_round_stats(cpu, dm_disk(md));
+	disk_stat_add(cpu, dm_disk(md), ticks[rw], duration);
+	disk_stat_unlock();
 
-	disk_stat_add(dm_disk(md), ticks[rw], duration);
+	dm_disk(md)->in_flight = pending = atomic_dec_return(&md->pending);
 
 	return !pending;
 }
@@ -885,6 +886,7 @@
 	int r = -EIO;
 	int rw = bio_data_dir(bio);
 	struct mapped_device *md = q->queuedata;
+	int cpu;
 
 	/*
 	 * There is no use in forwarding any barrier request since we can't
@@ -897,8 +899,10 @@
 
 	down_read(&md->io_lock);
 
-	disk_stat_inc(dm_disk(md), ios[rw]);
-	disk_stat_add(dm_disk(md), sectors[rw], bio_sectors(bio));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, dm_disk(md), ios[rw]);
+	disk_stat_add(cpu, dm_disk(md), sectors[rw], bio_sectors(bio));
+	disk_stat_unlock();
 
 	/*
 	 * If we're suspended we have to queue
diff --git a/drivers/md/linear.c b/drivers/md/linear.c
index b1eebf8..00cbc8e 100644
--- a/drivers/md/linear.c
+++ b/drivers/md/linear.c
@@ -318,14 +318,17 @@
 	mddev_t *mddev = q->queuedata;
 	dev_info_t *tmp_dev;
 	sector_t block;
+	int cpu;
 
 	if (unlikely(bio_barrier(bio))) {
 		bio_endio(bio, -EOPNOTSUPP);
 		return 0;
 	}
 
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, mddev->gendisk, ios[rw]);
+	disk_stat_add(cpu, mddev->gendisk, sectors[rw], bio_sectors(bio));
+	disk_stat_unlock();
 
 	tmp_dev = which_dev(mddev, bio->bi_sector);
 	block = bio->bi_sector >> 1;
diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index c4779cc..182f5a9 100644
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -147,6 +147,7 @@
 	struct multipath_bh * mp_bh;
 	struct multipath_info *multipath;
 	const int rw = bio_data_dir(bio);
+	int cpu;
 
 	if (unlikely(bio_barrier(bio))) {
 		bio_endio(bio, -EOPNOTSUPP);
@@ -158,8 +159,10 @@
 	mp_bh->master_bio = bio;
 	mp_bh->mddev = mddev;
 
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, mddev->gendisk, ios[rw]);
+	disk_stat_add(cpu, mddev->gendisk, sectors[rw], bio_sectors(bio));
+	disk_stat_unlock();
 
 	mp_bh->path = multipath_map(conf);
 	if (mp_bh->path < 0) {
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index 1836106..e26030f 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -399,14 +399,17 @@
 	sector_t chunk;
 	sector_t block, rsect;
 	const int rw = bio_data_dir(bio);
+	int cpu;
 
 	if (unlikely(bio_barrier(bio))) {
 		bio_endio(bio, -EOPNOTSUPP);
 		return 0;
 	}
 
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, mddev->gendisk, ios[rw]);
+	disk_stat_add(cpu, mddev->gendisk, sectors[rw], bio_sectors(bio));
+	disk_stat_unlock();
 
 	chunk_size = mddev->chunk_size >> 10;
 	chunk_sects = mddev->chunk_size >> 9;
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 0b82030..babb130 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -779,7 +779,7 @@
 	struct page **behind_pages = NULL;
 	const int rw = bio_data_dir(bio);
 	const int do_sync = bio_sync(bio);
-	int do_barriers;
+	int cpu, do_barriers;
 	mdk_rdev_t *blocked_rdev;
 
 	/*
@@ -804,8 +804,10 @@
 
 	bitmap = mddev->bitmap;
 
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, mddev->gendisk, ios[rw]);
+	disk_stat_add(cpu, mddev->gendisk, sectors[rw], bio_sectors(bio));
+	disk_stat_unlock();
 
 	/*
 	 * make_request() can abort the operation when READA is being
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index d3b9aa0..5ec80da 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -789,6 +789,7 @@
 	mirror_info_t *mirror;
 	r10bio_t *r10_bio;
 	struct bio *read_bio;
+	int cpu;
 	int i;
 	int chunk_sects = conf->chunk_mask + 1;
 	const int rw = bio_data_dir(bio);
@@ -843,8 +844,10 @@
 	 */
 	wait_barrier(conf);
 
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bio));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, mddev->gendisk, ios[rw]);
+	disk_stat_add(cpu, mddev->gendisk, sectors[rw], bio_sectors(bio));
+	disk_stat_unlock();
 
 	r10_bio = mempool_alloc(conf->r10bio_pool, GFP_NOIO);
 
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 37e5465..5899f21 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -3387,7 +3387,7 @@
 	sector_t logical_sector, last_sector;
 	struct stripe_head *sh;
 	const int rw = bio_data_dir(bi);
-	int remaining;
+	int cpu, remaining;
 
 	if (unlikely(bio_barrier(bi))) {
 		bio_endio(bi, -EOPNOTSUPP);
@@ -3396,8 +3396,10 @@
 
 	md_write_start(mddev, bi);
 
-	disk_stat_inc(mddev->gendisk, ios[rw]);
-	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
+	cpu = disk_stat_lock();
+	disk_stat_inc(cpu, mddev->gendisk, ios[rw]);
+	disk_stat_add(cpu, mddev->gendisk, sectors[rw], bio_sectors(bi));
+	disk_stat_unlock();
 
 	if (rw == READ &&
 	     mddev->reshape_position == MaxSector &&
diff --git a/fs/partitions/check.c b/fs/partitions/check.c
index 96c8bf4..c442f0a 100644
--- a/fs/partitions/check.c
+++ b/fs/partitions/check.c
@@ -219,10 +219,11 @@
 			      struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
+	int cpu;
 
-	preempt_disable();
-	part_round_stats(p);
-	preempt_enable();
+	cpu = disk_stat_lock();
+	part_round_stats(cpu, p);
+	disk_stat_unlock();
 	return sprintf(buf,
 		"%8lu %8lu %8llu %8u "
 		"%8lu %8lu %8llu %8u "
diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index 7fbba19..ac8a901 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -209,16 +209,24 @@
 extern struct hd_struct *disk_map_sector_rcu(struct gendisk *disk,
 					     sector_t sector);
 
-/* 
+/*
  * Macros to operate on percpu disk statistics:
  *
- * The __ variants should only be called in critical sections. The full
- * variants disable/enable preemption.
+ * {disk|part|all}_stat_{add|sub|inc|dec}() modify the stat counters
+ * and should be called between disk_stat_lock() and
+ * disk_stat_unlock().
+ *
+ * part_stat_read() can be called at any time.
+ *
+ * part_stat_{add|set_all}() and {init|free}_part_stats are for
+ * internal use only.
  */
-
 #ifdef	CONFIG_SMP
-#define __disk_stat_add(gendiskp, field, addnd) 	\
-	(per_cpu_ptr(gendiskp->dkstats, smp_processor_id())->field += addnd)
+#define disk_stat_lock()	({ rcu_read_lock(); get_cpu(); })
+#define disk_stat_unlock()	do { put_cpu(); rcu_read_unlock(); } while (0)
+
+#define disk_stat_add(cpu, gendiskp, field, addnd)			\
+	(per_cpu_ptr(gendiskp->dkstats, cpu)->field += addnd)
 
 #define disk_stat_read(gendiskp, field)					\
 ({									\
@@ -229,7 +237,8 @@
 	res;								\
 })
 
-static inline void disk_stat_set_all(struct gendisk *gendiskp, int value)	{
+static inline void disk_stat_set_all(struct gendisk *gendiskp, int value)
+{
 	int i;
 
 	for_each_possible_cpu(i)
@@ -237,14 +246,14 @@
 				sizeof(struct disk_stats));
 }		
 
-#define __part_stat_add(part, field, addnd)				\
-	(per_cpu_ptr(part->dkstats, smp_processor_id())->field += addnd)
+#define part_stat_add(cpu, part, field, addnd)				\
+	(per_cpu_ptr(part->dkstats, cpu)->field += addnd)
 
-#define __all_stat_add(gendiskp, part, field, addnd, sector)	\
-({								\
-	if (part)						\
-		__part_stat_add(part, field, addnd);		\
-	__disk_stat_add(gendiskp, field, addnd);		\
+#define all_stat_add(cpu, gendiskp, part, field, addnd, sector)		\
+({									\
+	if (part)							\
+		part_stat_add(cpu, part, field, addnd);			\
+	disk_stat_add(cpu, gendiskp, field, addnd);			\
 })
 
 #define part_stat_read(part, field)					\
@@ -264,10 +273,13 @@
 		memset(per_cpu_ptr(part->dkstats, i), value,
 				sizeof(struct disk_stats));
 }
-				
+
 #else /* !CONFIG_SMP */
-#define __disk_stat_add(gendiskp, field, addnd) \
-				(gendiskp->dkstats.field += addnd)
+#define disk_stat_lock()	({ rcu_read_lock(); 0; })
+#define disk_stat_unlock()	rcu_read_unlock()
+
+#define disk_stat_add(cpu, gendiskp, field, addnd)			\
+	(gendiskp->dkstats.field += addnd)
 #define disk_stat_read(gendiskp, field)	(gendiskp->dkstats.field)
 
 static inline void disk_stat_set_all(struct gendisk *gendiskp, int value)
@@ -275,14 +287,14 @@
 	memset(&gendiskp->dkstats, value, sizeof (struct disk_stats));
 }
 
-#define __part_stat_add(part, field, addnd) \
+#define part_stat_add(cpu, part, field, addnd)				\
 	(part->dkstats.field += addnd)
 
-#define __all_stat_add(gendiskp, part, field, addnd, sector)	\
-({								\
-	if (part)						\
-		part->dkstats.field += addnd;			\
-	__disk_stat_add(gendiskp, field, addnd);		\
+#define all_stat_add(cpu, gendiskp, part, field, addnd, sector)		\
+({									\
+	if (part)							\
+		part_stat_add(cpu, part, field, addnd);			\
+	disk_stat_add(cpu, gendiskp, field, addnd);			\
 })
 
 #define part_stat_read(part, field)	(part->dkstats.field)
@@ -294,63 +306,26 @@
 
 #endif /* CONFIG_SMP */
 
-#define disk_stat_add(gendiskp, field, addnd)			\
-	do {							\
-		preempt_disable();				\
-		__disk_stat_add(gendiskp, field, addnd);	\
-		preempt_enable();				\
-	} while (0)
+#define disk_stat_dec(cpu, gendiskp, field)				\
+	disk_stat_add(cpu, gendiskp, field, -1)
+#define disk_stat_inc(cpu, gendiskp, field)				\
+	disk_stat_add(cpu, gendiskp, field, 1)
+#define disk_stat_sub(cpu, gendiskp, field, subnd)			\
+	disk_stat_add(cpu, gendiskp, field, -subnd)
 
-#define __disk_stat_dec(gendiskp, field) __disk_stat_add(gendiskp, field, -1)
-#define disk_stat_dec(gendiskp, field) disk_stat_add(gendiskp, field, -1)
+#define part_stat_dec(cpu, gendiskp, field)				\
+	part_stat_add(cpu, gendiskp, field, -1)
+#define part_stat_inc(cpu, gendiskp, field)				\
+	part_stat_add(cpu, gendiskp, field, 1)
+#define part_stat_sub(cpu, gendiskp, field, subnd)			\
+	part_stat_add(cpu, gendiskp, field, -subnd)
 
-#define __disk_stat_inc(gendiskp, field) __disk_stat_add(gendiskp, field, 1)
-#define disk_stat_inc(gendiskp, field) disk_stat_add(gendiskp, field, 1)
-
-#define __disk_stat_sub(gendiskp, field, subnd) \
-		__disk_stat_add(gendiskp, field, -subnd)
-#define disk_stat_sub(gendiskp, field, subnd) \
-		disk_stat_add(gendiskp, field, -subnd)
-
-#define part_stat_add(gendiskp, field, addnd)		\
-	do {						\
-		preempt_disable();			\
-		__part_stat_add(gendiskp, field, addnd);\
-		preempt_enable();			\
-	} while (0)
-
-#define __part_stat_dec(gendiskp, field) __part_stat_add(gendiskp, field, -1)
-#define part_stat_dec(gendiskp, field) part_stat_add(gendiskp, field, -1)
-
-#define __part_stat_inc(gendiskp, field) __part_stat_add(gendiskp, field, 1)
-#define part_stat_inc(gendiskp, field) part_stat_add(gendiskp, field, 1)
-
-#define __part_stat_sub(gendiskp, field, subnd) \
-		__part_stat_add(gendiskp, field, -subnd)
-#define part_stat_sub(gendiskp, field, subnd) \
-		part_stat_add(gendiskp, field, -subnd)
-
-#define all_stat_add(gendiskp, part, field, addnd, sector)	\
-	do {							\
-		preempt_disable();				\
-		__all_stat_add(gendiskp, part, field, addnd, sector);	\
-		preempt_enable();				\
-	} while (0)
-
-#define __all_stat_dec(gendiskp, field, sector) \
-		__all_stat_add(gendiskp, field, -1, sector)
-#define all_stat_dec(gendiskp, field, sector) \
-		all_stat_add(gendiskp, field, -1, sector)
-
-#define __all_stat_inc(gendiskp, part, field, sector) \
-		__all_stat_add(gendiskp, part, field, 1, sector)
-#define all_stat_inc(gendiskp, part, field, sector) \
-		all_stat_add(gendiskp, part, field, 1, sector)
-
-#define __all_stat_sub(gendiskp, part, field, subnd, sector) \
-		__all_stat_add(gendiskp, part, field, -subnd, sector)
-#define all_stat_sub(gendiskp, part, field, subnd, sector) \
-		all_stat_add(gendiskp, part, field, -subnd, sector)
+#define all_stat_dec(cpu, gendiskp, field, sector)			\
+	all_stat_add(cpu, gendiskp, field, -1, sector)
+#define all_stat_inc(cpu, gendiskp, part, field, sector)		\
+	all_stat_add(cpu, gendiskp, part, field, 1, sector)
+#define all_stat_sub(cpu, gendiskp, part, field, subnd, sector)		\
+	all_stat_add(cpu, gendiskp, part, field, -subnd, sector)
 
 /* Inlines to alloc and free disk stats in struct gendisk */
 #ifdef  CONFIG_SMP
@@ -401,8 +376,8 @@
 #endif	/* CONFIG_SMP */
 
 /* drivers/block/ll_rw_blk.c */
-extern void disk_round_stats(struct gendisk *disk);
-extern void part_round_stats(struct hd_struct *part);
+extern void disk_round_stats(int cpu, struct gendisk *disk);
+extern void part_round_stats(int cpu, struct hd_struct *part);
 
 /* drivers/block/genhd.c */
 extern int get_blkdev_list(char *, int);