Merge branch 'for-3.15/core' of git://git.kernel.dk/linux-block

Pull core block layer updates from Jens Axboe:
 "This is the pull request for the core block IO bits for the 3.15
  kernel.  It's a smaller round this time, it contains:

   - Various little blk-mq fixes and additions from Christoph and
     myself.

   - Cleanup of the IPI usage from the block layer, and associated
     helper code.  From Frederic Weisbecker and Jan Kara.

   - Duplicate code cleanup in bio-integrity from Gu Zheng.  This will
     give you a merge conflict, but that should be easy to resolve.

   - blk-mq notify spinlock fix for RT from Mike Galbraith.

   - A blktrace partial accounting bug fix from Roman Pen.

   - Missing REQ_SYNC detection fix for blk-mq from Shaohua Li"

* 'for-3.15/core' of git://git.kernel.dk/linux-block: (25 commits)
  blk-mq: add REQ_SYNC early
  rt,blk,mq: Make blk_mq_cpu_notify_lock a raw spinlock
  blk-mq: support partial I/O completions
  blk-mq: merge blk_mq_insert_request and blk_mq_run_request
  blk-mq: remove blk_mq_alloc_rq
  blk-mq: don't dump CPU -> hw queue map on driver load
  blk-mq: fix wrong usage of hctx->state vs hctx->flags
  blk-mq: allow blk_mq_init_commands() to return failure
  block: remove old blk_iopoll_enabled variable
  blktrace: fix accounting of partially completed requests
  smp: Rename __smp_call_function_single() to smp_call_function_single_async()
  smp: Remove wait argument from __smp_call_function_single()
  watchdog: Simplify a little the IPI call
  smp: Move __smp_call_function_single() below its safe version
  smp: Consolidate the various smp_call_function_single() declensions
  smp: Teach __smp_call_function_single() to check for offline cpus
  smp: Remove unused list_head from csd
  smp: Iterate functions through llist_for_each_entry_safe()
  block: Stop abusing rq->csd.list in blk-softirq
  block: Remove useless IPI struct initialization
  ...

block/blk-cgroup.c

@@ -336,7 +336,7 @@ static void blkg_destroy(struct blkcg_gq *blkg)
 	 * under queue_lock.  If it's not pointing to @blkg now, it never
 	 * will.  Hint assignment itself can race safely.
 	 */
-	if (rcu_dereference_raw(blkcg->blkg_hint) == blkg)
+	if (rcu_access_pointer(blkcg->blkg_hint) == blkg)
 		rcu_assign_pointer(blkcg->blkg_hint, NULL);
 
 	/*

block/blk-core.c

@@ -2353,7 +2353,7 @@ bool blk_update_request(struct request *req, int error, unsigned int nr_bytes)
 	if (!req->bio)
 		return false;
 
-	trace_block_rq_complete(req->q, req);
+	trace_block_rq_complete(req->q, req, nr_bytes);
 
 	/*
 	 * For fs requests, rq is just carrier of independent bio's

block/blk-ioc.c

@@ -68,7 +68,7 @@ static void ioc_destroy_icq(struct io_cq *icq)
 	 * under queue_lock.  If it's not pointing to @icq now, it never
 	 * will.  Hint assignment itself can race safely.
 	 */
-	if (rcu_dereference_raw(ioc->icq_hint) == icq)
+	if (rcu_access_pointer(ioc->icq_hint) == icq)
 		rcu_assign_pointer(ioc->icq_hint, NULL);
 
 	ioc_exit_icq(icq);

block/blk-iopoll.c

@@ -14,9 +14,6 @@
 
 #include "blk.h"
 
-int blk_iopoll_enabled = 1;
-EXPORT_SYMBOL(blk_iopoll_enabled);
-
 static unsigned int blk_iopoll_budget __read_mostly = 256;
 
 static DEFINE_PER_CPU(struct list_head, blk_cpu_iopoll);

block/blk-mq-cpumap.c

@@ -9,15 +9,6 @@
 #include "blk.h"
 #include "blk-mq.h"
 
-static void show_map(unsigned int *map, unsigned int nr)
-{
-	int i;
-
-	pr_info("blk-mq: CPU -> queue map\n");
-	for_each_online_cpu(i)
-		pr_info("  CPU%2u -> Queue %u\n", i, map[i]);
-}
-
 static int cpu_to_queue_index(unsigned int nr_cpus, unsigned int nr_queues,
 			      const int cpu)
 {
@@ -85,7 +76,6 @@ int blk_mq_update_queue_map(unsigned int *map, unsigned int nr_queues)
 			map[i] = map[first_sibling];
 	}
 
-	show_map(map, nr_cpus);
 	free_cpumask_var(cpus);
 	return 0;
 }

block/blk-mq-sysfs.c

@@ -244,6 +244,32 @@ static ssize_t blk_mq_hw_sysfs_tags_show(struct blk_mq_hw_ctx *hctx, char *page)
 	return blk_mq_tag_sysfs_show(hctx->tags, page);
 }
 
+static ssize_t blk_mq_hw_sysfs_cpus_show(struct blk_mq_hw_ctx *hctx, char *page)
+{
+	unsigned int i, queue_num, first = 1;
+	ssize_t ret = 0;
+
+	blk_mq_disable_hotplug();
+
+	for_each_online_cpu(i) {
+		queue_num = hctx->queue->mq_map[i];
+		if (queue_num != hctx->queue_num)
+			continue;
+
+		if (first)
+			ret += sprintf(ret + page, "%u", i);
+		else
+			ret += sprintf(ret + page, ", %u", i);
+
+		first = 0;
+	}
+
+	blk_mq_enable_hotplug();
+
+	ret += sprintf(ret + page, "\n");
+	return ret;
+}
+
 static struct blk_mq_ctx_sysfs_entry blk_mq_sysfs_dispatched = {
 	.attr = {.name = "dispatched", .mode = S_IRUGO },
 	.show = blk_mq_sysfs_dispatched_show,
@@ -294,6 +320,10 @@ static struct blk_mq_hw_ctx_sysfs_entry blk_mq_hw_sysfs_tags = {
 	.attr = {.name = "tags", .mode = S_IRUGO },
 	.show = blk_mq_hw_sysfs_tags_show,
 };
+static struct blk_mq_hw_ctx_sysfs_entry blk_mq_hw_sysfs_cpus = {
+	.attr = {.name = "cpu_list", .mode = S_IRUGO },
+	.show = blk_mq_hw_sysfs_cpus_show,
+};
 
 static struct attribute *default_hw_ctx_attrs[] = {
 	&blk_mq_hw_sysfs_queued.attr,
@@ -302,6 +332,7 @@ static struct attribute *default_hw_ctx_attrs[] = {
 	&blk_mq_hw_sysfs_pending.attr,
 	&blk_mq_hw_sysfs_ipi.attr,
 	&blk_mq_hw_sysfs_tags.attr,
+	&blk_mq_hw_sysfs_cpus.attr,
 	NULL,
 };
 

block/blk-mq.c

@@ -320,7 +320,7 @@ void __blk_mq_complete_request(struct request *rq)
 		rq->csd.func = __blk_mq_complete_request_remote;
 		rq->csd.info = rq;
 		rq->csd.flags = 0;
-		__smp_call_function_single(ctx->cpu, &rq->csd, 0);
+		smp_call_function_single_async(ctx->cpu, &rq->csd);
 	} else {
 		rq->q->softirq_done_fn(rq);
 	}
@@ -514,7 +514,7 @@ static void __blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx)
 	LIST_HEAD(rq_list);
 	int bit, queued;
 
-	if (unlikely(test_bit(BLK_MQ_S_STOPPED, &hctx->flags)))
+	if (unlikely(test_bit(BLK_MQ_S_STOPPED, &hctx->state)))
 		return;
 
 	hctx->run++;
@@ -603,7 +603,7 @@ static void __blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx)
 
 void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async)
 {
-	if (unlikely(test_bit(BLK_MQ_S_STOPPED, &hctx->flags)))
+	if (unlikely(test_bit(BLK_MQ_S_STOPPED, &hctx->state)))
 		return;
 
 	if (!async)
@@ -623,7 +623,7 @@ void blk_mq_run_queues(struct request_queue *q, bool async)
 	queue_for_each_hw_ctx(q, hctx, i) {
 		if ((!blk_mq_hctx_has_pending(hctx) &&
 		    list_empty_careful(&hctx->dispatch)) ||
-		    test_bit(BLK_MQ_S_STOPPED, &hctx->flags))
+		    test_bit(BLK_MQ_S_STOPPED, &hctx->state))
 			continue;
 
 		blk_mq_run_hw_queue(hctx, async);
@@ -994,8 +994,46 @@ static void blk_mq_hctx_notify(void *data, unsigned long action,
 	blk_mq_put_ctx(ctx);
 }
 
-static void blk_mq_init_hw_commands(struct blk_mq_hw_ctx *hctx,
-				    void (*init)(void *, struct blk_mq_hw_ctx *,
+static int blk_mq_init_hw_commands(struct blk_mq_hw_ctx *hctx,
+				   int (*init)(void *, struct blk_mq_hw_ctx *,
+					struct request *, unsigned int),
+				   void *data)
+{
+	unsigned int i;
+	int ret = 0;
+
+	for (i = 0; i < hctx->queue_depth; i++) {
+		struct request *rq = hctx->rqs[i];
+
+		ret = init(data, hctx, rq, i);
+		if (ret)
+			break;
+	}
+
+	return ret;
+}
+
+int blk_mq_init_commands(struct request_queue *q,
+			 int (*init)(void *, struct blk_mq_hw_ctx *,
+					struct request *, unsigned int),
+			 void *data)
+{
+	struct blk_mq_hw_ctx *hctx;
+	unsigned int i;
+	int ret = 0;
+
+	queue_for_each_hw_ctx(q, hctx, i) {
+		ret = blk_mq_init_hw_commands(hctx, init, data);
+		if (ret)
+			break;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(blk_mq_init_commands);
+
+static void blk_mq_free_hw_commands(struct blk_mq_hw_ctx *hctx,
+				    void (*free)(void *, struct blk_mq_hw_ctx *,
 					struct request *, unsigned int),
 				    void *data)
 {
@@ -1004,12 +1042,12 @@ static void blk_mq_init_hw_commands(struct blk_mq_hw_ctx *hctx,
 	for (i = 0; i < hctx->queue_depth; i++) {
 		struct request *rq = hctx->rqs[i];
 
-		init(data, hctx, rq, i);
+		free(data, hctx, rq, i);
 	}
 }
 
-void blk_mq_init_commands(struct request_queue *q,
-			  void (*init)(void *, struct blk_mq_hw_ctx *,
+void blk_mq_free_commands(struct request_queue *q,
+			  void (*free)(void *, struct blk_mq_hw_ctx *,
 					struct request *, unsigned int),
 			  void *data)
 {
@@ -1017,9 +1055,9 @@ void blk_mq_init_commands(struct request_queue *q,
 	unsigned int i;
 
 	queue_for_each_hw_ctx(q, hctx, i)
-		blk_mq_init_hw_commands(hctx, init, data);
+		blk_mq_free_hw_commands(hctx, free, data);
 }
-EXPORT_SYMBOL(blk_mq_init_commands);
+EXPORT_SYMBOL(blk_mq_free_commands);
 
 static void blk_mq_free_rq_map(struct blk_mq_hw_ctx *hctx)
 {
@@ -1430,6 +1468,16 @@ static int blk_mq_queue_reinit_notify(struct notifier_block *nb,
 	return NOTIFY_OK;
 }
 
+void blk_mq_disable_hotplug(void)
+{
+	mutex_lock(&all_q_mutex);
+}
+
+void blk_mq_enable_hotplug(void)
+{
+	mutex_unlock(&all_q_mutex);
+}
+
 static int __init blk_mq_init(void)
 {
 	blk_mq_cpu_init();

block/blk-mq.h

@@ -39,6 +39,8 @@ void blk_mq_init_cpu_notifier(struct blk_mq_cpu_notifier *notifier,
 void blk_mq_register_cpu_notifier(struct blk_mq_cpu_notifier *notifier);
 void blk_mq_unregister_cpu_notifier(struct blk_mq_cpu_notifier *notifier);
 void blk_mq_cpu_init(void);
+void blk_mq_enable_hotplug(void);
+void blk_mq_disable_hotplug(void);
 
 /*
  * CPU -> queue mappings

block/blk-softirq.c

@@ -30,8 +30,8 @@ static void blk_done_softirq(struct softirq_action *h)
 	while (!list_empty(&local_list)) {
 		struct request *rq;
 
-		rq = list_entry(local_list.next, struct request, csd.list);
-		list_del_init(&rq->csd.list);
+		rq = list_entry(local_list.next, struct request, queuelist);
+		list_del_init(&rq->queuelist);
 		rq->q->softirq_done_fn(rq);
 	}
 }
@@ -45,9 +45,14 @@ static void trigger_softirq(void *data)
 
 	local_irq_save(flags);
 	list = this_cpu_ptr(&blk_cpu_done);
-	list_add_tail(&rq->csd.list, list);
+	/*
+	 * We reuse queuelist for a list of requests to process. Since the
+	 * queuelist is used by the block layer only for requests waiting to be
+	 * submitted to the device it is unused now.
+	 */
+	list_add_tail(&rq->queuelist, list);
 
-	if (list->next == &rq->csd.list)
+	if (list->next == &rq->queuelist)
 		raise_softirq_irqoff(BLOCK_SOFTIRQ);
 
 	local_irq_restore(flags);
@@ -65,7 +70,7 @@ static int raise_blk_irq(int cpu, struct request *rq)
 		data->info = rq;
 		data->flags = 0;
 
-		__smp_call_function_single(cpu, data, 0);
+		smp_call_function_single_async(cpu, data);
 		return 0;
 	}
 
@@ -136,7 +141,7 @@ void __blk_complete_request(struct request *req)
 		struct list_head *list;
 do_local:
 		list = this_cpu_ptr(&blk_cpu_done);
-		list_add_tail(&req->csd.list, list);
+		list_add_tail(&req->queuelist, list);
 
 		/*
 		 * if the list only contains our just added request,
@@ -144,7 +149,7 @@ void __blk_complete_request(struct request *req)
 		 * entries there, someone already raised the irq but it
 		 * hasn't run yet.
 		 */
-		if (list->next == &req->csd.list)
+		if (list->next == &req->queuelist)
 			raise_softirq_irqoff(BLOCK_SOFTIRQ);
 	} else if (raise_blk_irq(ccpu, req))
 		goto do_local;

block/cfq-iosched.c

@@ -2367,10 +2367,10 @@ cfq_merged_requests(struct request_queue *q, struct request *rq,
 	 * reposition in fifo if next is older than rq
 	 */
 	if (!list_empty(&rq->queuelist) && !list_empty(&next->queuelist) &&
-	    time_before(rq_fifo_time(next), rq_fifo_time(rq)) &&
+	    time_before(next->fifo_time, rq->fifo_time) &&
 	    cfqq == RQ_CFQQ(next)) {
 		list_move(&rq->queuelist, &next->queuelist);
-		rq_set_fifo_time(rq, rq_fifo_time(next));
+		rq->fifo_time = next->fifo_time;
 	}
 
 	if (cfqq->next_rq == next)
@@ -2814,7 +2814,7 @@ static struct request *cfq_check_fifo(struct cfq_queue *cfqq)
 		return NULL;
 
 	rq = rq_entry_fifo(cfqq->fifo.next);
-	if (time_before(jiffies, rq_fifo_time(rq)))
+	if (time_before(jiffies, rq->fifo_time))
 		rq = NULL;
 
 	cfq_log_cfqq(cfqq->cfqd, cfqq, "fifo=%p", rq);
@@ -3927,7 +3927,7 @@ static void cfq_insert_request(struct request_queue *q, struct request *rq)
 	cfq_log_cfqq(cfqd, cfqq, "insert_request");
 	cfq_init_prio_data(cfqq, RQ_CIC(rq));
 
-	rq_set_fifo_time(rq, jiffies + cfqd->cfq_fifo_expire[rq_is_sync(rq)]);
+	rq->fifo_time = jiffies + cfqd->cfq_fifo_expire[rq_is_sync(rq)];
 	list_add_tail(&rq->queuelist, &cfqq->fifo);
 	cfq_add_rq_rb(rq);
 	cfqg_stats_update_io_add(RQ_CFQG(rq), cfqd->serving_group,

block/deadline-iosched.c

@@ -106,7 +106,7 @@ deadline_add_request(struct request_queue *q, struct request *rq)
 	/*
 	 * set expire time and add to fifo list
 	 */
-	rq_set_fifo_time(rq, jiffies + dd->fifo_expire[data_dir]);
+	rq->fifo_time = jiffies + dd->fifo_expire[data_dir];
 	list_add_tail(&rq->queuelist, &dd->fifo_list[data_dir]);
 }
 
@@ -174,9 +174,9 @@ deadline_merged_requests(struct request_queue *q, struct request *req,
 	 * and move into next position (next will be deleted) in fifo
 	 */
 	if (!list_empty(&req->queuelist) && !list_empty(&next->queuelist)) {
-		if (time_before(rq_fifo_time(next), rq_fifo_time(req))) {
+		if (time_before(next->fifo_time, req->fifo_time)) {
 			list_move(&req->queuelist, &next->queuelist);
-			rq_set_fifo_time(req, rq_fifo_time(next));
+			req->fifo_time = next->fifo_time;
 		}
 	}
 
@@ -230,7 +230,7 @@ static inline int deadline_check_fifo(struct deadline_data *dd, int ddir)
 	/*
 	 * rq is expired!
 	 */
-	if (time_after_eq(jiffies, rq_fifo_time(rq)))
+	if (time_after_eq(jiffies, rq->fifo_time))
 		return 1;
 
 	return 0;

block/partitions/atari.h

@@ -11,6 +11,8 @@
  * by Guenther Kelleter (guenther@pool.informatik.rwth-aachen.de)
  */
 
+#include <linux/compiler.h>
+
 struct partition_info
 {
   u8 flg;			/* bit 0: active; bit 7: bootable */
@@ -29,6 +31,6 @@ struct rootsector
   u32 bsl_st;			/* start of bad sector list */
   u32 bsl_cnt;			/* length of bad sector list */
   u16 checksum;			/* checksum for bootable disks */
-} __attribute__((__packed__));
+} __packed;
 
 int atari_partition(struct parsed_partitions *state);

block/partitions/efi.h

@@ -32,6 +32,7 @@
 #include <linux/major.h>
 #include <linux/string.h>
 #include <linux/efi.h>
+#include <linux/compiler.h>
 
 #define MSDOS_MBR_SIGNATURE 0xaa55
 #define EFI_PMBR_OSTYPE_EFI 0xEF
@@ -87,13 +88,13 @@ typedef struct _gpt_header {
 	 *
 	 * uint8_t		reserved2[ BlockSize - 92 ];
 	 */
-} __attribute__ ((packed)) gpt_header;
+} __packed gpt_header;
 
 typedef struct _gpt_entry_attributes {
 	u64 required_to_function:1;
 	u64 reserved:47;
         u64 type_guid_specific:16;
-} __attribute__ ((packed)) gpt_entry_attributes;
+} __packed gpt_entry_attributes;
 
 typedef struct _gpt_entry {
 	efi_guid_t partition_type_guid;
@@ -102,7 +103,7 @@ typedef struct _gpt_entry {
 	__le64 ending_lba;
 	gpt_entry_attributes attributes;
 	efi_char16_t partition_name[72 / sizeof (efi_char16_t)];
-} __attribute__ ((packed)) gpt_entry;
+} __packed gpt_entry;
 
 typedef struct _gpt_mbr_record {
 	u8	boot_indicator; /* unused by EFI, set to 0x80 for bootable */
@@ -124,7 +125,7 @@ typedef struct _legacy_mbr {
 	__le16 unknown;
 	gpt_mbr_record partition_record[4];
 	__le16 signature;
-} __attribute__ ((packed)) legacy_mbr;
+} __packed legacy_mbr;
 
 /* Functions */
 extern int efi_partition(struct parsed_partitions *state);

block/partitions/karma.c

@@ -8,6 +8,7 @@
 
 #include "check.h"
 #include "karma.h"
+#include <linux/compiler.h>
 
 int karma_partition(struct parsed_partitions *state)
 {
@@ -26,7 +27,7 @@ int karma_partition(struct parsed_partitions *state)
 		} d_partitions[2];
 		u8 d_blank[208];
 		__le16 d_magic;
-	} __attribute__((packed)) *label;
+	} __packed *label;
 	struct d_partition *p;
 
 	data = read_part_sector(state, 0, &sect);

drivers/block/virtio_blk.c

@@ -490,13 +490,14 @@ static struct blk_mq_reg virtio_mq_reg = {
 	.flags		= BLK_MQ_F_SHOULD_MERGE,
 };
 
-static void virtblk_init_vbr(void *data, struct blk_mq_hw_ctx *hctx,
+static int virtblk_init_vbr(void *data, struct blk_mq_hw_ctx *hctx,
 			     struct request *rq, unsigned int nr)
 {
 	struct virtio_blk *vblk = data;
 	struct virtblk_req *vbr = rq->special;
 
 	sg_init_table(vbr->sg, vblk->sg_elems);
+	return 0;
 }
 
 static int virtblk_probe(struct virtio_device *vdev)

drivers/cpuidle/coupled.c

@@ -323,7 +323,7 @@ static void cpuidle_coupled_poke(int cpu)
 	struct call_single_data *csd = &per_cpu(cpuidle_coupled_poke_cb, cpu);
 
 	if (!cpumask_test_and_set_cpu(cpu, &cpuidle_coupled_poke_pending))
-		__smp_call_function_single(cpu, csd, 0);
+		smp_call_function_single_async(cpu, csd);
 }
 
 /**

drivers/scsi/be2iscsi/be_main.c

@@ -898,7 +898,6 @@ static irqreturn_t be_isr_msix(int irq, void *dev_id)
 	struct be_queue_info *cq;
 	unsigned int num_eq_processed;
 	struct be_eq_obj *pbe_eq;
-	unsigned long flags;
 
 	pbe_eq = dev_id;
 	eq = &pbe_eq->q;
@@ -907,31 +906,15 @@ static irqreturn_t be_isr_msix(int irq, void *dev_id)
 
 	phba = pbe_eq->phba;
 	num_eq_processed = 0;
-	if (blk_iopoll_enabled) {
-		while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
-					& EQE_VALID_MASK) {
-			if (!blk_iopoll_sched_prep(&pbe_eq->iopoll))
-				blk_iopoll_sched(&pbe_eq->iopoll);
-
-			AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
-			queue_tail_inc(eq);
-			eqe = queue_tail_node(eq);
-			num_eq_processed++;
-		}
-	} else {
-		while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
-						& EQE_VALID_MASK) {
-			spin_lock_irqsave(&phba->isr_lock, flags);
-			pbe_eq->todo_cq = true;
-			spin_unlock_irqrestore(&phba->isr_lock, flags);
-			AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
-			queue_tail_inc(eq);
-			eqe = queue_tail_node(eq);
-			num_eq_processed++;
-		}
+	while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
+				& EQE_VALID_MASK) {
+		if (!blk_iopoll_sched_prep(&pbe_eq->iopoll))
+			blk_iopoll_sched(&pbe_eq->iopoll);
 
-		if (pbe_eq->todo_cq)
-			queue_work(phba->wq, &pbe_eq->work_cqs);
+		AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
+		queue_tail_inc(eq);
+		eqe = queue_tail_node(eq);
+		num_eq_processed++;
 	}
 
 	if (num_eq_processed)
@@ -952,7 +935,6 @@ static irqreturn_t be_isr(int irq, void *dev_id)
 	struct hwi_context_memory *phwi_context;
 	struct be_eq_entry *eqe = NULL;
 	struct be_queue_info *eq;
-	struct be_queue_info *cq;
 	struct be_queue_info *mcc;
 	unsigned long flags, index;
 	unsigned int num_mcceq_processed, num_ioeq_processed;
@@ -978,72 +960,40 @@ static irqreturn_t be_isr(int irq, void *dev_id)
 
 	num_ioeq_processed = 0;
 	num_mcceq_processed = 0;
-	if (blk_iopoll_enabled) {
-		while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
-					& EQE_VALID_MASK) {
-			if (((eqe->dw[offsetof(struct amap_eq_entry,
-			     resource_id) / 32] &
-			     EQE_RESID_MASK) >> 16) == mcc->id) {
-				spin_lock_irqsave(&phba->isr_lock, flags);
-				pbe_eq->todo_mcc_cq = true;
-				spin_unlock_irqrestore(&phba->isr_lock, flags);
-				num_mcceq_processed++;
-			} else {
-				if (!blk_iopoll_sched_prep(&pbe_eq->iopoll))
-					blk_iopoll_sched(&pbe_eq->iopoll);
-				num_ioeq_processed++;
-			}
-			AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
-			queue_tail_inc(eq);
-			eqe = queue_tail_node(eq);
-		}
-		if (num_ioeq_processed || num_mcceq_processed) {
-			if (pbe_eq->todo_mcc_cq)
-				queue_work(phba->wq, &pbe_eq->work_cqs);
-
-			if ((num_mcceq_processed) && (!num_ioeq_processed))
-				hwi_ring_eq_db(phba, eq->id, 0,
-					      (num_ioeq_processed +
-					       num_mcceq_processed) , 1, 1);
-			else
-				hwi_ring_eq_db(phba, eq->id, 0,
-					       (num_ioeq_processed +
-						num_mcceq_processed), 0, 1);
-
-			return IRQ_HANDLED;
-		} else
-			return IRQ_NONE;
-	} else {
-		cq = &phwi_context->be_cq[0];
-		while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
-						& EQE_VALID_MASK) {
-
-			if (((eqe->dw[offsetof(struct amap_eq_entry,
-			     resource_id) / 32] &
-			     EQE_RESID_MASK) >> 16) != cq->id) {
-				spin_lock_irqsave(&phba->isr_lock, flags);
-				pbe_eq->todo_mcc_cq = true;
-				spin_unlock_irqrestore(&phba->isr_lock, flags);
-			} else {
-				spin_lock_irqsave(&phba->isr_lock, flags);
-				pbe_eq->todo_cq = true;
-				spin_unlock_irqrestore(&phba->isr_lock, flags);
-			}
-			AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
-			queue_tail_inc(eq);
-			eqe = queue_tail_node(eq);
+	while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
+				& EQE_VALID_MASK) {
+		if (((eqe->dw[offsetof(struct amap_eq_entry,
+		     resource_id) / 32] &
+		     EQE_RESID_MASK) >> 16) == mcc->id) {
+			spin_lock_irqsave(&phba->isr_lock, flags);
+			pbe_eq->todo_mcc_cq = true;
+			spin_unlock_irqrestore(&phba->isr_lock, flags);
+			num_mcceq_processed++;
+		} else {
+			if (!blk_iopoll_sched_prep(&pbe_eq->iopoll))
+				blk_iopoll_sched(&pbe_eq->iopoll);
 			num_ioeq_processed++;
 		}
-		if (pbe_eq->todo_cq || pbe_eq->todo_mcc_cq)
+		AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
+		queue_tail_inc(eq);
+		eqe = queue_tail_node(eq);
+	}
+	if (num_ioeq_processed || num_mcceq_processed) {
+		if (pbe_eq->todo_mcc_cq)
 			queue_work(phba->wq, &pbe_eq->work_cqs);
 
-		if (num_ioeq_processed) {
+		if ((num_mcceq_processed) && (!num_ioeq_processed))
 			hwi_ring_eq_db(phba, eq->id, 0,
-				       num_ioeq_processed, 1, 1);
-			return IRQ_HANDLED;
-		} else
-			return IRQ_NONE;
-	}
+				      (num_ioeq_processed +
+				       num_mcceq_processed) , 1, 1);
+		else
+			hwi_ring_eq_db(phba, eq->id, 0,
+				       (num_ioeq_processed +
+					num_mcceq_processed), 0, 1);
+
+		return IRQ_HANDLED;
+	} else
+		return IRQ_NONE;
 }
 
 static int beiscsi_init_irqs(struct beiscsi_hba *phba)
@@ -5263,11 +5213,10 @@ static void beiscsi_quiesce(struct beiscsi_hba *phba,
 		}
 	pci_disable_msix(phba->pcidev);
 
-	if (blk_iopoll_enabled)
-		for (i = 0; i < phba->num_cpus; i++) {
-			pbe_eq = &phwi_context->be_eq[i];
-			blk_iopoll_disable(&pbe_eq->iopoll);
-		}
+	for (i = 0; i < phba->num_cpus; i++) {
+		pbe_eq = &phwi_context->be_eq[i];
+		blk_iopoll_disable(&pbe_eq->iopoll);
+	}
 
 	if (unload_state == BEISCSI_CLEAN_UNLOAD) {
 		destroy_workqueue(phba->wq);
@@ -5478,32 +5427,18 @@ static void beiscsi_eeh_resume(struct pci_dev *pdev)
 	phwi_ctrlr = phba->phwi_ctrlr;
 	phwi_context = phwi_ctrlr->phwi_ctxt;
 
-	if (blk_iopoll_enabled) {
-		for (i = 0; i < phba->num_cpus; i++) {
-			pbe_eq = &phwi_context->be_eq[i];
-			blk_iopoll_init(&pbe_eq->iopoll, be_iopoll_budget,
-					be_iopoll);
-			blk_iopoll_enable(&pbe_eq->iopoll);
-		}
-
-		i = (phba->msix_enabled) ? i : 0;
-		/* Work item for MCC handling */
+	for (i = 0; i < phba->num_cpus; i++) {
 		pbe_eq = &phwi_context->be_eq[i];
-		INIT_WORK(&pbe_eq->work_cqs, beiscsi_process_all_cqs);
-	} else {
-		if (phba->msix_enabled) {
-			for (i = 0; i <= phba->num_cpus; i++) {
-				pbe_eq = &phwi_context->be_eq[i];
-				INIT_WORK(&pbe_eq->work_cqs,
-					  beiscsi_process_all_cqs);
-			}
-		} else {
-			pbe_eq = &phwi_context->be_eq[0];
-			INIT_WORK(&pbe_eq->work_cqs,
-				  beiscsi_process_all_cqs);
-		}
+		blk_iopoll_init(&pbe_eq->iopoll, be_iopoll_budget,
+				be_iopoll);
+		blk_iopoll_enable(&pbe_eq->iopoll);
 	}
 
+	i = (phba->msix_enabled) ? i : 0;
+	/* Work item for MCC handling */
+	pbe_eq = &phwi_context->be_eq[i];
+	INIT_WORK(&pbe_eq->work_cqs, beiscsi_process_all_cqs);
+
 	ret = beiscsi_init_irqs(phba);
 	if (ret < 0) {
 		beiscsi_log(phba, KERN_ERR, BEISCSI_LOG_INIT,
@@ -5665,32 +5600,18 @@ static int beiscsi_dev_probe(struct pci_dev *pcidev,
 	phwi_ctrlr = phba->phwi_ctrlr;
 	phwi_context = phwi_ctrlr->phwi_ctxt;
 
-	if (blk_iopoll_enabled) {
-		for (i = 0; i < phba->num_cpus; i++) {
-			pbe_eq = &phwi_context->be_eq[i];
-			blk_iopoll_init(&pbe_eq->iopoll, be_iopoll_budget,
-					be_iopoll);
-			blk_iopoll_enable(&pbe_eq->iopoll);
-		}
-
-		i = (phba->msix_enabled) ? i : 0;
-		/* Work item for MCC handling */
+	for (i = 0; i < phba->num_cpus; i++) {
 		pbe_eq = &phwi_context->be_eq[i];
-		INIT_WORK(&pbe_eq->work_cqs, beiscsi_process_all_cqs);
-	} else {
-		if (phba->msix_enabled) {
-			for (i = 0; i <= phba->num_cpus; i++) {
-				pbe_eq = &phwi_context->be_eq[i];
-				INIT_WORK(&pbe_eq->work_cqs,
-					  beiscsi_process_all_cqs);
-			}
-		} else {
-				pbe_eq = &phwi_context->be_eq[0];
-				INIT_WORK(&pbe_eq->work_cqs,
-					  beiscsi_process_all_cqs);
-			}
+		blk_iopoll_init(&pbe_eq->iopoll, be_iopoll_budget,
+				be_iopoll);
+		blk_iopoll_enable(&pbe_eq->iopoll);
 	}
 
+	i = (phba->msix_enabled) ? i : 0;
+	/* Work item for MCC handling */
+	pbe_eq = &phwi_context->be_eq[i];
+	INIT_WORK(&pbe_eq->work_cqs, beiscsi_process_all_cqs);
+
 	ret = beiscsi_init_irqs(phba);
 	if (ret < 0) {
 		beiscsi_log(phba, KERN_ERR, BEISCSI_LOG_INIT,
@@ -5719,11 +5640,10 @@ static int beiscsi_dev_probe(struct pci_dev *pcidev,
 
 free_blkenbld:
 	destroy_workqueue(phba->wq);
-	if (blk_iopoll_enabled)
-		for (i = 0; i < phba->num_cpus; i++) {
-			pbe_eq = &phwi_context->be_eq[i];
-			blk_iopoll_disable(&pbe_eq->iopoll);
-		}
+	for (i = 0; i < phba->num_cpus; i++) {
+		pbe_eq = &phwi_context->be_eq[i];
+		blk_iopoll_disable(&pbe_eq->iopoll);
+	}
 free_twq:
 	beiscsi_clean_port(phba);
 	beiscsi_free_mem(phba);

drivers/scsi/ipr.c

@@ -3670,16 +3670,14 @@ static ssize_t ipr_store_iopoll_weight(struct device *dev,
 		return strlen(buf);
 	}
 
-	if (blk_iopoll_enabled && ioa_cfg->iopoll_weight &&
-			ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
+	if (ioa_cfg->iopoll_weight && ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
 		for (i = 1; i < ioa_cfg->hrrq_num; i++)
 			blk_iopoll_disable(&ioa_cfg->hrrq[i].iopoll);
 	}
 
 	spin_lock_irqsave(shost->host_lock, lock_flags);
 	ioa_cfg->iopoll_weight = user_iopoll_weight;
-	if (blk_iopoll_enabled && ioa_cfg->iopoll_weight &&
-			ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
+	if (ioa_cfg->iopoll_weight && ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
 		for (i = 1; i < ioa_cfg->hrrq_num; i++) {
 			blk_iopoll_init(&ioa_cfg->hrrq[i].iopoll,
 					ioa_cfg->iopoll_weight, ipr_iopoll);
@@ -5525,8 +5523,7 @@ static irqreturn_t ipr_isr_mhrrq(int irq, void *devp)
 		return IRQ_NONE;
 	}
 
-	if (blk_iopoll_enabled && ioa_cfg->iopoll_weight &&
-			ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
+	if (ioa_cfg->iopoll_weight && ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
 		if ((be32_to_cpu(*hrrq->hrrq_curr) & IPR_HRRQ_TOGGLE_BIT) ==
 		       hrrq->toggle_bit) {
 			if (!blk_iopoll_sched_prep(&hrrq->iopoll))
@@ -9975,8 +9972,7 @@ static int ipr_probe(struct pci_dev *pdev, const struct pci_device_id *dev_id)
 	ioa_cfg->host->max_channel = IPR_VSET_BUS;
 	ioa_cfg->iopoll_weight = ioa_cfg->chip_cfg->iopoll_weight;
 
-	if (blk_iopoll_enabled && ioa_cfg->iopoll_weight &&
-			ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
+	if (ioa_cfg->iopoll_weight && ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
 		for (i = 1; i < ioa_cfg->hrrq_num; i++) {
 			blk_iopoll_init(&ioa_cfg->hrrq[i].iopoll,
 					ioa_cfg->iopoll_weight, ipr_iopoll);
@@ -10005,8 +10001,7 @@ static void ipr_shutdown(struct pci_dev *pdev)
 	int i;
 
 	spin_lock_irqsave(ioa_cfg->host->host_lock, lock_flags);
-	if (blk_iopoll_enabled && ioa_cfg->iopoll_weight &&
-			ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
+	if (ioa_cfg->iopoll_weight && ioa_cfg->sis64 && ioa_cfg->nvectors > 1) {
 		ioa_cfg->iopoll_weight = 0;
 		for (i = 1; i < ioa_cfg->hrrq_num; i++)
 			blk_iopoll_disable(&ioa_cfg->hrrq[i].iopoll);

fs/bio-integrity.c

@@ -301,25 +301,25 @@ int bio_integrity_get_tag(struct bio *bio, void *tag_buf, unsigned int len)
 EXPORT_SYMBOL(bio_integrity_get_tag);
 
 /**
- * bio_integrity_generate - Generate integrity metadata for a bio
- * @bio:	bio to generate integrity metadata for
- *
- * Description: Generates integrity metadata for a bio by calling the
- * block device's generation callback function.  The bio must have a
- * bip attached with enough room to accommodate the generated
- * integrity metadata.
+ * bio_integrity_generate_verify - Generate/verify integrity metadata for a bio
+ * @bio:	bio to generate/verify integrity metadata for
+ * @operate:	operate number, 1 for generate, 0 for verify
  */
-static void bio_integrity_generate(struct bio *bio)
+static int bio_integrity_generate_verify(struct bio *bio, int operate)
 {
 	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);
 	struct blk_integrity_exchg bix;
 	struct bio_vec bv;
 	struct bvec_iter iter;
-	sector_t sector = bio->bi_iter.bi_sector;
-	unsigned int sectors, total;
+	sector_t sector;
+	unsigned int sectors, ret = 0;
 	void *prot_buf = bio->bi_integrity->bip_buf;
 
-	total = 0;
+	if (operate)
+		sector = bio->bi_iter.bi_sector;
+	else
+		sector = bio->bi_integrity->bip_iter.bi_sector;
+
 	bix.disk_name = bio->bi_bdev->bd_disk->disk_name;
 	bix.sector_size = bi->sector_size;
 
@@ -330,16 +330,37 @@ static void bio_integrity_generate(struct bio *bio)
 		bix.prot_buf = prot_buf;
 		bix.sector = sector;
 
-		bi->generate_fn(&bix);
+		if (operate) {
+			bi->generate_fn(&bix);
+		} else {
+			ret = bi->verify_fn(&bix);
+			if (ret) {
+				kunmap_atomic(kaddr);
+				return ret;
+			}
+		}
 
 		sectors = bv.bv_len / bi->sector_size;
 		sector += sectors;
 		prot_buf += sectors * bi->tuple_size;
-		total += sectors * bi->tuple_size;
-		BUG_ON(total > bio->bi_integrity->bip_iter.bi_size);
 
 		kunmap_atomic(kaddr);
 	}
+	return ret;
+}
+
+/**
+ * bio_integrity_generate - Generate integrity metadata for a bio
+ * @bio:	bio to generate integrity metadata for
+ *
+ * Description: Generates integrity metadata for a bio by calling the
+ * block device's generation callback function.  The bio must have a
+ * bip attached with enough room to accommodate the generated
+ * integrity metadata.
+ */
+static void bio_integrity_generate(struct bio *bio)
+{
+	bio_integrity_generate_verify(bio, 1);
 }
 
 static inline unsigned short blk_integrity_tuple_size(struct blk_integrity *bi)
@@ -454,40 +475,7 @@ EXPORT_SYMBOL(bio_integrity_prep);
  */
 static int bio_integrity_verify(struct bio *bio)
 {
-	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);
-	struct blk_integrity_exchg bix;
-	struct bio_vec *bv;
-	sector_t sector = bio->bi_integrity->bip_iter.bi_sector;
-	unsigned int sectors, ret = 0;
-	void *prot_buf = bio->bi_integrity->bip_buf;
-	int i;
-
-	bix.disk_name = bio->bi_bdev->bd_disk->disk_name;
-	bix.sector_size = bi->sector_size;
-
-	bio_for_each_segment_all(bv, bio, i) {
-		void *kaddr = kmap_atomic(bv->bv_page);
-
-		bix.data_buf = kaddr + bv->bv_offset;
-		bix.data_size = bv->bv_len;
-		bix.prot_buf = prot_buf;
-		bix.sector = sector;
-
-		ret = bi->verify_fn(&bix);
-
-		if (ret) {
-			kunmap_atomic(kaddr);
-			return ret;
-		}
-
-		sectors = bv->bv_len / bi->sector_size;
-		sector += sectors;
-		prot_buf += sectors * bi->tuple_size;
-
-		kunmap_atomic(kaddr);
-	}
-
-	return ret;
+	return bio_integrity_generate_verify(bio, 0);
 }
 
 /**

fs/bio.c

@@ -116,7 +116,6 @@ static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size)
 	if (!slab)
 		goto out_unlock;
 
-	printk(KERN_INFO "bio: create slab <%s> at %d\n", bslab->name, entry);
 	bslab->slab = slab;
 	bslab->slab_ref = 1;
 	bslab->slab_size = sz;

include/linux/blk-iopoll.h

@@ -43,6 +43,4 @@ extern void __blk_iopoll_complete(struct blk_iopoll *);
 extern void blk_iopoll_enable(struct blk_iopoll *);
 extern void blk_iopoll_disable(struct blk_iopoll *);
 
-extern int blk_iopoll_enabled;
-
 #endif

include/linux/blk-mq.h

@@ -109,7 +109,7 @@ enum {
 	BLK_MQ_F_SHOULD_SORT	= 1 << 1,
 	BLK_MQ_F_SHOULD_IPI	= 1 << 2,
 
-	BLK_MQ_S_STOPPED	= 1 << 0,
+	BLK_MQ_S_STOPPED	= 0,
 
 	BLK_MQ_MAX_DEPTH	= 2048,
 };
@@ -117,7 +117,8 @@ enum {
 struct request_queue *blk_mq_init_queue(struct blk_mq_reg *, void *);
 int blk_mq_register_disk(struct gendisk *);
 void blk_mq_unregister_disk(struct gendisk *);
-void blk_mq_init_commands(struct request_queue *, void (*init)(void *data, struct blk_mq_hw_ctx *, struct request *, unsigned int), void *data);
+int blk_mq_init_commands(struct request_queue *, int (*init)(void *data, struct blk_mq_hw_ctx *, struct request *, unsigned int), void *data);
+void blk_mq_free_commands(struct request_queue *, void (*free)(void *data, struct blk_mq_hw_ctx *, struct request *, unsigned int), void *data);
 
 void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule);
 

include/linux/blkdev.h

@@ -99,6 +99,7 @@ struct request {
 	union {
 		struct call_single_data csd;
 		struct work_struct mq_flush_work;
+		unsigned long fifo_time;
 	};
 
 	struct request_queue *q;

include/linux/elevator.h

@@ -202,17 +202,8 @@ enum {
 #define rq_end_sector(rq)	(blk_rq_pos(rq) + blk_rq_sectors(rq))
 #define rb_entry_rq(node)	rb_entry((node), struct request, rb_node)
 
-/*
- * Hack to reuse the csd.list list_head as the fifo time holder while
- * the request is in the io scheduler. Saves an unsigned long in rq.
- */
-#define rq_fifo_time(rq)	((unsigned long) (rq)->csd.list.next)
-#define rq_set_fifo_time(rq,exp)	((rq)->csd.list.next = (void *) (exp))
 #define rq_entry_fifo(ptr)	list_entry((ptr), struct request, queuelist)
-#define rq_fifo_clear(rq)	do {		\
-	list_del_init(&(rq)->queuelist);	\
-	INIT_LIST_HEAD(&(rq)->csd.list);	\
-	} while (0)
+#define rq_fifo_clear(rq)	list_del_init(&(rq)->queuelist)
 
 #else /* CONFIG_BLOCK */
 

include/linux/smp.h

@@ -17,10 +17,7 @@ extern void cpu_idle(void);
 
 typedef void (*smp_call_func_t)(void *info);
 struct call_single_data {
-	union {
-		struct list_head list;
-		struct llist_node llist;
-	};
+	struct llist_node llist;
 	smp_call_func_t func;
 	void *info;
 	u16 flags;
@@ -53,8 +50,7 @@ void on_each_cpu_cond(bool (*cond_func)(int cpu, void *info),
 		smp_call_func_t func, void *info, bool wait,
 		gfp_t gfp_flags);
 
-void __smp_call_function_single(int cpuid, struct call_single_data *data,
-				int wait);
+int smp_call_function_single_async(int cpu, struct call_single_data *csd);
 
 #ifdef CONFIG_SMP
 

include/trace/events/block.h

@@ -132,6 +132,7 @@ DEFINE_EVENT(block_rq_with_error, block_rq_requeue,
  * block_rq_complete - block IO operation completed by device driver
  * @q: queue containing the block operation request
  * @rq: block operations request
+ * @nr_bytes: number of completed bytes
  *
  * The block_rq_complete tracepoint event indicates that some portion
  * of operation request has been completed by the device driver.  If
@@ -139,11 +140,37 @@ DEFINE_EVENT(block_rq_with_error, block_rq_requeue,
  * do for the request. If @rq->bio is non-NULL then there is
  * additional work required to complete the request.
  */
-DEFINE_EVENT(block_rq_with_error, block_rq_complete,
+TRACE_EVENT(block_rq_complete,
 
-	TP_PROTO(struct request_queue *q, struct request *rq),
+	TP_PROTO(struct request_queue *q, struct request *rq,
+		 unsigned int nr_bytes),
 
-	TP_ARGS(q, rq)
+	TP_ARGS(q, rq, nr_bytes),
+
+	TP_STRUCT__entry(
+		__field(  dev_t,	dev			)
+		__field(  sector_t,	sector			)
+		__field(  unsigned int,	nr_sector		)
+		__field(  int,		errors			)
+		__array(  char,		rwbs,	RWBS_LEN	)
+		__dynamic_array( char,	cmd,	blk_cmd_buf_len(rq)	)
+	),
+
+	TP_fast_assign(
+		__entry->dev	   = rq->rq_disk ? disk_devt(rq->rq_disk) : 0;
+		__entry->sector    = blk_rq_pos(rq);
+		__entry->nr_sector = nr_bytes >> 9;
+		__entry->errors    = rq->errors;
+
+		blk_fill_rwbs(__entry->rwbs, rq->cmd_flags, nr_bytes);
+		blk_dump_cmd(__get_str(cmd), rq);
+	),
+
+	TP_printk("%d,%d %s (%s) %llu + %u [%d]",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->rwbs, __get_str(cmd),
+		  (unsigned long long)__entry->sector,
+		  __entry->nr_sector, __entry->errors)
 );
 
 DECLARE_EVENT_CLASS(block_rq,

kernel/sched/core.c

@@ -432,7 +432,7 @@ void hrtick_start(struct rq *rq, u64 delay)
 	if (rq == this_rq()) {
 		__hrtick_restart(rq);
 	} else if (!rq->hrtick_csd_pending) {
-		__smp_call_function_single(cpu_of(rq), &rq->hrtick_csd, 0);
+		smp_call_function_single_async(cpu_of(rq), &rq->hrtick_csd);
 		rq->hrtick_csd_pending = 1;
 	}
 }

kernel/smp.c

@@ -117,13 +117,43 @@ static void csd_unlock(struct call_single_data *csd)
 	csd->flags &= ~CSD_FLAG_LOCK;
 }
 
+static DEFINE_PER_CPU_SHARED_ALIGNED(struct call_single_data, csd_data);
+
 /*
  * Insert a previously allocated call_single_data element
  * for execution on the given CPU. data must already have
  * ->func, ->info, and ->flags set.
  */
-static void generic_exec_single(int cpu, struct call_single_data *csd, int wait)
+static int generic_exec_single(int cpu, struct call_single_data *csd,
+			       smp_call_func_t func, void *info, int wait)
 {
+	struct call_single_data csd_stack = { .flags = 0 };
+	unsigned long flags;
+
+
+	if (cpu == smp_processor_id()) {
+		local_irq_save(flags);
+		func(info);
+		local_irq_restore(flags);
+		return 0;
+	}
+
+
+	if ((unsigned)cpu >= nr_cpu_ids || !cpu_online(cpu))
+		return -ENXIO;
+
+
+	if (!csd) {
+		csd = &csd_stack;
+		if (!wait)
+			csd = &__get_cpu_var(csd_data);
+	}
+
+	csd_lock(csd);
+
+	csd->func = func;
+	csd->info = info;
+
 	if (wait)
 		csd->flags |= CSD_FLAG_WAIT;
 
@@ -143,6 +173,8 @@ static void generic_exec_single(int cpu, struct call_single_data *csd, int wait)
 
 	if (wait)
 		csd_lock_wait(csd);
+
+	return 0;
 }
 
 /*
@@ -151,7 +183,8 @@ static void generic_exec_single(int cpu, struct call_single_data *csd, int wait)
  */
 void generic_smp_call_function_single_interrupt(void)
 {
-	struct llist_node *entry, *next;
+	struct llist_node *entry;
+	struct call_single_data *csd, *csd_next;
 
 	/*
 	 * Shouldn't receive this interrupt on a cpu that is not yet online.
@@ -161,21 +194,12 @@ void generic_smp_call_function_single_interrupt(void)
 	entry = llist_del_all(&__get_cpu_var(call_single_queue));
 	entry = llist_reverse_order(entry);
 
-	while (entry) {
-		struct call_single_data *csd;
-
-		next = entry->next;
-
-		csd = llist_entry(entry, struct call_single_data, llist);
+	llist_for_each_entry_safe(csd, csd_next, entry, llist) {
 		csd->func(csd->info);
 		csd_unlock(csd);
-
-		entry = next;
 	}
 }
 
-static DEFINE_PER_CPU_SHARED_ALIGNED(struct call_single_data, csd_data);
-
 /*
  * smp_call_function_single - Run a function on a specific CPU
  * @func: The function to run. This must be fast and non-blocking.
@@ -187,12 +211,8 @@ static DEFINE_PER_CPU_SHARED_ALIGNED(struct call_single_data, csd_data);
 int smp_call_function_single(int cpu, smp_call_func_t func, void *info,
 			     int wait)
 {
-	struct call_single_data d = {
-		.flags = 0,
-	};
-	unsigned long flags;
 	int this_cpu;
-	int err = 0;
+	int err;
 
 	/*
 	 * prevent preemption and reschedule on another processor,
@@ -209,32 +229,41 @@ int smp_call_function_single(int cpu, smp_call_func_t func, void *info,
 	WARN_ON_ONCE(cpu_online(this_cpu) && irqs_disabled()
 		     && !oops_in_progress);
 
-	if (cpu == this_cpu) {
-		local_irq_save(flags);
-		func(info);
-		local_irq_restore(flags);
-	} else {
-		if ((unsigned)cpu < nr_cpu_ids && cpu_online(cpu)) {
-			struct call_single_data *csd = &d;
+	err = generic_exec_single(cpu, NULL, func, info, wait);
 
-			if (!wait)
-				csd = &__get_cpu_var(csd_data);
+	put_cpu();
 
-			csd_lock(csd);
+	return err;
+}
+EXPORT_SYMBOL(smp_call_function_single);
 
-			csd->func = func;
-			csd->info = info;
-			generic_exec_single(cpu, csd, wait);
-		} else {
-			err = -ENXIO;	/* CPU not online */
-		}
-	}
+/**
+ * smp_call_function_single_async(): Run an asynchronous function on a
+ * 			         specific CPU.
+ * @cpu: The CPU to run on.
+ * @csd: Pre-allocated and setup data structure
+ *
+ * Like smp_call_function_single(), but the call is asynchonous and
+ * can thus be done from contexts with disabled interrupts.
+ *
+ * The caller passes his own pre-allocated data structure
+ * (ie: embedded in an object) and is responsible for synchronizing it
+ * such that the IPIs performed on the @csd are strictly serialized.
+ *
+ * NOTE: Be careful, there is unfortunately no current debugging facility to
+ * validate the correctness of this serialization.
+ */
+int smp_call_function_single_async(int cpu, struct call_single_data *csd)
+{
+	int err = 0;
 
-	put_cpu();
+	preempt_disable();
+	err = generic_exec_single(cpu, csd, csd->func, csd->info, 0);
+	preempt_enable();
 
 	return err;
 }
-EXPORT_SYMBOL(smp_call_function_single);
+EXPORT_SYMBOL_GPL(smp_call_function_single_async);
 
 /*
  * smp_call_function_any - Run a function on any of the given cpus
@@ -279,44 +308,6 @@ int smp_call_function_any(const struct cpumask *mask,
 }
 EXPORT_SYMBOL_GPL(smp_call_function_any);
 
-/**
- * __smp_call_function_single(): Run a function on a specific CPU
- * @cpu: The CPU to run on.
- * @data: Pre-allocated and setup data structure
- * @wait: If true, wait until function has completed on specified CPU.
- *
- * Like smp_call_function_single(), but allow caller to pass in a
- * pre-allocated data structure. Useful for embedding @data inside
- * other structures, for instance.
- */
-void __smp_call_function_single(int cpu, struct call_single_data *csd,
-				int wait)
-{
-	unsigned int this_cpu;
-	unsigned long flags;
-
-	this_cpu = get_cpu();
-	/*
-	 * Can deadlock when called with interrupts disabled.
-	 * We allow cpu's that are not yet online though, as no one else can
-	 * send smp call function interrupt to this cpu and as such deadlocks
-	 * can't happen.
-	 */
-	WARN_ON_ONCE(cpu_online(smp_processor_id()) && wait && irqs_disabled()
-		     && !oops_in_progress);
-
-	if (cpu == this_cpu) {
-		local_irq_save(flags);
-		csd->func(csd->info);
-		local_irq_restore(flags);
-	} else {
-		csd_lock(csd);
-		generic_exec_single(cpu, csd, wait);
-	}
-	put_cpu();
-}
-EXPORT_SYMBOL_GPL(__smp_call_function_single);
-
 /**
  * smp_call_function_many(): Run a function on a set of other CPUs.
  * @mask: The set of cpus to run on (only runs on online subset).

kernel/sysctl.c

@@ -112,9 +112,6 @@ extern int sysctl_nr_open_min, sysctl_nr_open_max;
 #ifndef CONFIG_MMU
 extern int sysctl_nr_trim_pages;
 #endif
-#ifdef CONFIG_BLOCK
-extern int blk_iopoll_enabled;
-#endif
 
 /* Constants used for minimum and  maximum */
 #ifdef CONFIG_LOCKUP_DETECTOR
@@ -1086,15 +1083,6 @@ static struct ctl_table kern_table[] = {
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec,
 	},
-#endif
-#ifdef CONFIG_BLOCK
-	{
-		.procname	= "blk_iopoll",
-		.data		= &blk_iopoll_enabled,
-		.maxlen		= sizeof(int),
-		.mode		= 0644,
-		.proc_handler	= proc_dointvec,
-	},
 #endif
 	{ }
 };

kernel/trace/blktrace.c

@@ -702,6 +702,7 @@ void blk_trace_shutdown(struct request_queue *q)
  * blk_add_trace_rq - Add a trace for a request oriented action
  * @q:		queue the io is for
  * @rq:		the source request
+ * @nr_bytes:	number of completed bytes
  * @what:	the action
  *
  * Description:
@@ -709,7 +710,7 @@ void blk_trace_shutdown(struct request_queue *q)
  *
  **/
 static void blk_add_trace_rq(struct request_queue *q, struct request *rq,
-			     u32 what)
+			     unsigned int nr_bytes, u32 what)
 {
 	struct blk_trace *bt = q->blk_trace;
 
@@ -718,11 +719,11 @@ static void blk_add_trace_rq(struct request_queue *q, struct request *rq,
 
 	if (rq->cmd_type == REQ_TYPE_BLOCK_PC) {
 		what |= BLK_TC_ACT(BLK_TC_PC);
-		__blk_add_trace(bt, 0, blk_rq_bytes(rq), rq->cmd_flags,
+		__blk_add_trace(bt, 0, nr_bytes, rq->cmd_flags,
 				what, rq->errors, rq->cmd_len, rq->cmd);
 	} else  {
 		what |= BLK_TC_ACT(BLK_TC_FS);
-		__blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq),
+		__blk_add_trace(bt, blk_rq_pos(rq), nr_bytes,
 				rq->cmd_flags, what, rq->errors, 0, NULL);
 	}
 }
@@ -730,33 +731,34 @@ static void blk_add_trace_rq(struct request_queue *q, struct request *rq,
 static void blk_add_trace_rq_abort(void *ignore,
 				   struct request_queue *q, struct request *rq)
 {
-	blk_add_trace_rq(q, rq, BLK_TA_ABORT);
+	blk_add_trace_rq(q, rq, blk_rq_bytes(rq), BLK_TA_ABORT);
 }
 
 static void blk_add_trace_rq_insert(void *ignore,
 				    struct request_queue *q, struct request *rq)
 {
-	blk_add_trace_rq(q, rq, BLK_TA_INSERT);
+	blk_add_trace_rq(q, rq, blk_rq_bytes(rq), BLK_TA_INSERT);
 }
 
 static void blk_add_trace_rq_issue(void *ignore,
 				   struct request_queue *q, struct request *rq)
 {
-	blk_add_trace_rq(q, rq, BLK_TA_ISSUE);
+	blk_add_trace_rq(q, rq, blk_rq_bytes(rq), BLK_TA_ISSUE);
 }
 
 static void blk_add_trace_rq_requeue(void *ignore,
 				     struct request_queue *q,
 				     struct request *rq)
 {
-	blk_add_trace_rq(q, rq, BLK_TA_REQUEUE);
+	blk_add_trace_rq(q, rq, blk_rq_bytes(rq), BLK_TA_REQUEUE);
 }
 
 static void blk_add_trace_rq_complete(void *ignore,
 				      struct request_queue *q,
-				      struct request *rq)
+				      struct request *rq,
+				      unsigned int nr_bytes)
 {
-	blk_add_trace_rq(q, rq, BLK_TA_COMPLETE);
+	blk_add_trace_rq(q, rq, nr_bytes, BLK_TA_COMPLETE);
 }
 
 /**

kernel/up.c

@@ -22,16 +22,16 @@ int smp_call_function_single(int cpu, void (*func) (void *info), void *info,
 }
 EXPORT_SYMBOL(smp_call_function_single);
 
-void __smp_call_function_single(int cpu, struct call_single_data *csd,
-				int wait)
+int smp_call_function_single_async(int cpu, struct call_single_data *csd)
 {
 	unsigned long flags;
 
 	local_irq_save(flags);
 	csd->func(csd->info);
 	local_irq_restore(flags);
+	return 0;
 }
-EXPORT_SYMBOL(__smp_call_function_single);
+EXPORT_SYMBOL(smp_call_function_single_async);
 
 int on_each_cpu(smp_call_func_t func, void *info, int wait)
 {

kernel/watchdog.c

@@ -505,7 +505,6 @@ static void restart_watchdog_hrtimer(void *info)
 
 static void update_timers(int cpu)
 {
-	struct call_single_data data = {.func = restart_watchdog_hrtimer};
 	/*
 	 * Make sure that perf event counter will adopt to a new
 	 * sampling period. Updating the sampling period directly would
@@ -515,7 +514,7 @@ static void update_timers(int cpu)
 	 * might be late already so we have to restart the timer as well.
 	 */
 	watchdog_nmi_disable(cpu);
-	__smp_call_function_single(cpu, &data, 1);
+	smp_call_function_single(cpu, restart_watchdog_hrtimer, NULL, 1);
 	watchdog_nmi_enable(cpu);
 }
 

net/core/dev.c

@@ -4135,8 +4135,8 @@ static void net_rps_action_and_irq_enable(struct softnet_data *sd)
 			struct softnet_data *next = remsd->rps_ipi_next;
 
 			if (cpu_online(remsd->cpu))
-				__smp_call_function_single(remsd->cpu,
-							   &remsd->csd, 0);
+				smp_call_function_single_async(remsd->cpu,
+							   &remsd->csd);
 			remsd = next;
 		}
 	} else