4 files changed, 366 insertions, 25 deletions

diff --git a/block/Kconfig b/block/Kconfig
index 1ab7c15c8d7a..290b219fad9c 100644
--- a/block/Kconfig
+++ b/block/Kconfig
@@ -47,6 +47,7 @@ config BLK_DEV_IO_TRACE
         depends on SYSFS
         select RELAY
         select DEBUG_FS
+        select TRACEPOINTS
         help
           Say Y here if you want to be able to trace the block layer actions
           on a given queue. Tracing allows you to see any traffic happening
diff --git a/block/blk-core.c b/block/blk-core.c
index c36aa98fafa3..561e8a1b43a4 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -28,9 +28,23 @@
 #include <linux/task_io_accounting_ops.h>
 #include <linux/blktrace_api.h>
 #include <linux/fault-inject.h>
+#include <trace/block.h>
 
 #include "blk.h"
 
+DEFINE_TRACE(block_plug);
+DEFINE_TRACE(block_unplug_io);
+DEFINE_TRACE(block_unplug_timer);
+DEFINE_TRACE(block_getrq);
+DEFINE_TRACE(block_sleeprq);
+DEFINE_TRACE(block_rq_requeue);
+DEFINE_TRACE(block_bio_backmerge);
+DEFINE_TRACE(block_bio_frontmerge);
+DEFINE_TRACE(block_bio_queue);
+DEFINE_TRACE(block_rq_complete);
+DEFINE_TRACE(block_remap);      /* Also used in drivers/md/dm.c */
+EXPORT_TRACEPOINT_SYMBOL_GPL(block_remap);
+
 static int __make_request(struct request_queue *q, struct bio *bio);
 
 /*
@@ -205,7 +219,7 @@ void blk_plug_device(struct request_queue *q)
 
         if (!queue_flag_test_and_set(QUEUE_FLAG_PLUGGED, q)) {
                 mod_timer(&q->unplug_timer, jiffies + q->unplug_delay);
-                blk_add_trace_generic(q, NULL, 0, BLK_TA_PLUG);
+                trace_block_plug(q);
         }
 }
 EXPORT_SYMBOL(blk_plug_device);
@@ -292,9 +306,7 @@ void blk_unplug_work(struct work_struct *work)
         struct request_queue *q =
                 container_of(work, struct request_queue, unplug_work);
 
-        blk_add_trace_pdu_int(q, BLK_TA_UNPLUG_IO, NULL,
-                                q->rq.count[READ] + q->rq.count[WRITE]);
-
+        trace_block_unplug_io(q);
         q->unplug_fn(q);
 }
 
@@ -302,9 +314,7 @@ void blk_unplug_timeout(unsigned long data)
 {
         struct request_queue *q = (struct request_queue *)data;
 
-        blk_add_trace_pdu_int(q, BLK_TA_UNPLUG_TIMER, NULL,
-                                q->rq.count[READ] + q->rq.count[WRITE]);
-
+        trace_block_unplug_timer(q);
         kblockd_schedule_work(q, &q->unplug_work);
 }
 
@@ -314,9 +324,7 @@ void blk_unplug(struct request_queue *q)
          * devices don't necessarily have an ->unplug_fn defined
          */
         if (q->unplug_fn) {
-                blk_add_trace_pdu_int(q, BLK_TA_UNPLUG_IO, NULL,
-                                        q->rq.count[READ] + q->rq.count[WRITE]);
-
+                trace_block_unplug_io(q);
                 q->unplug_fn(q);
         }
 }
@@ -822,7 +830,7 @@ rq_starved:
         if (ioc_batching(q, ioc))
                 ioc->nr_batch_requests--;
 
-        blk_add_trace_generic(q, bio, rw, BLK_TA_GETRQ);
+        trace_block_getrq(q, bio, rw);
 out:
         return rq;
 }
@@ -848,7 +856,7 @@ static struct request *get_request_wait(struct request_queue *q, int rw_flags,
                 prepare_to_wait_exclusive(&rl->wait[rw], &wait,
                                 TASK_UNINTERRUPTIBLE);
 
-                blk_add_trace_generic(q, bio, rw, BLK_TA_SLEEPRQ);
+                trace_block_sleeprq(q, bio, rw);
 
                 __generic_unplug_device(q);
                 spin_unlock_irq(q->queue_lock);
@@ -928,7 +936,7 @@ void blk_requeue_request(struct request_queue *q, struct request *rq)
 {
         blk_delete_timer(rq);
         blk_clear_rq_complete(rq);
-        blk_add_trace_rq(q, rq, BLK_TA_REQUEUE);
+        trace_block_rq_requeue(q, rq);
 
         if (blk_rq_tagged(rq))
                 blk_queue_end_tag(q, rq);
@@ -1167,7 +1175,7 @@ static int __make_request(struct request_queue *q, struct bio *bio)
                 if (!ll_back_merge_fn(q, req, bio))
                         break;
 
-                blk_add_trace_bio(q, bio, BLK_TA_BACKMERGE);
+                trace_block_bio_backmerge(q, bio);
 
                 req->biotail->bi_next = bio;
                 req->biotail = bio;
@@ -1186,7 +1194,7 @@ static int __make_request(struct request_queue *q, struct bio *bio)
                 if (!ll_front_merge_fn(q, req, bio))
                         break;
 
-                blk_add_trace_bio(q, bio, BLK_TA_FRONTMERGE);
+                trace_block_bio_frontmerge(q, bio);
 
                 bio->bi_next = req->bio;
                 req->bio = bio;
@@ -1269,7 +1277,7 @@ static inline void blk_partition_remap(struct bio *bio)
                 bio->bi_sector += p->start_sect;
                 bio->bi_bdev = bdev->bd_contains;
 
-                blk_add_trace_remap(bdev_get_queue(bio->bi_bdev), bio,
+                trace_block_remap(bdev_get_queue(bio->bi_bdev), bio,
                                     bdev->bd_dev, bio->bi_sector,
                                     bio->bi_sector - p->start_sect);
         }
@@ -1441,10 +1449,10 @@ end_io:
                         goto end_io;
 
                 if (old_sector != -1)
-                        blk_add_trace_remap(q, bio, old_dev, bio->bi_sector,
+                        trace_block_remap(q, bio, old_dev, bio->bi_sector,
                                             old_sector);
 
-                blk_add_trace_bio(q, bio, BLK_TA_QUEUE);
+                trace_block_bio_queue(q, bio);
 
                 old_sector = bio->bi_sector;
                 old_dev = bio->bi_bdev->bd_dev;
@@ -1678,7 +1686,7 @@ static int __end_that_request_first(struct request *req, int error,
         int total_bytes, bio_nbytes, next_idx = 0;
         struct bio *bio;
 
-        blk_add_trace_rq(req->q, req, BLK_TA_COMPLETE);
+        trace_block_rq_complete(req->q, req);
 
         /*
          * for a REQ_TYPE_BLOCK_PC request, we want to carry any eventual
diff --git a/block/blktrace.c b/block/blktrace.c
index 85049a7e7a17..b0a2cae886db 100644
--- a/block/blktrace.c
+++ b/block/blktrace.c
@@ -23,10 +23,18 @@
 #include <linux/mutex.h>
 #include <linux/debugfs.h>
 #include <linux/time.h>
+#include <trace/block.h>
 #include <asm/uaccess.h>
 
 static unsigned int blktrace_seq __read_mostly = 1;
 
+/* Global reference count of probes */
+static DEFINE_MUTEX(blk_probe_mutex);
+static atomic_t blk_probes_ref = ATOMIC_INIT(0);
+
+static int blk_register_tracepoints(void);
+static void blk_unregister_tracepoints(void);
+
 /*
  * Send out a notify message.
  */
@@ -119,7 +127,7 @@ static u32 ddir_act[2] __read_mostly = { BLK_TC_ACT(BLK_TC_READ), BLK_TC_ACT(BLK
  * The worker for the various blk_add_trace*() types. Fills out a
  * blk_io_trace structure and places it in a per-cpu subbuffer.
  */
-void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
+static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
                      int rw, u32 what, int error, int pdu_len, void *pdu_data)
 {
         struct task_struct *tsk = current;
@@ -177,8 +185,6 @@ void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
         local_irq_restore(flags);
 }
 
-EXPORT_SYMBOL_GPL(__blk_add_trace);
-
 static struct dentry *blk_tree_root;
 static DEFINE_MUTEX(blk_tree_mutex);
 static unsigned int root_users;
@@ -237,6 +243,10 @@ static void blk_trace_cleanup(struct blk_trace *bt)
         free_percpu(bt->sequence);
         free_percpu(bt->msg_data);
         kfree(bt);
+        mutex_lock(&blk_probe_mutex);
+        if (atomic_dec_and_test(&blk_probes_ref))
+                blk_unregister_tracepoints();
+        mutex_unlock(&blk_probe_mutex);
 }
 
 int blk_trace_remove(struct request_queue *q)
@@ -428,6 +438,14 @@ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
         bt->pid = buts->pid;
         bt->trace_state = Blktrace_setup;
 
+        mutex_lock(&blk_probe_mutex);
+        if (atomic_add_return(1, &blk_probes_ref) == 1) {
+                ret = blk_register_tracepoints();
+                if (ret)
+                        goto probe_err;
+        }
+        mutex_unlock(&blk_probe_mutex);
+
         ret = -EBUSY;
         old_bt = xchg(&q->blk_trace, bt);
         if (old_bt) {
@@ -436,6 +454,9 @@ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
         }
 
         return 0;
+probe_err:
+        atomic_dec(&blk_probes_ref);
+        mutex_unlock(&blk_probe_mutex);
 err:
         if (dir)
                 blk_remove_tree(dir);
@@ -562,3 +583,308 @@ void blk_trace_shutdown(struct request_queue *q)
                 blk_trace_remove(q);
         }
 }
+
+/*
+ * blktrace probes
+ */
+
+/**
+ * blk_add_trace_rq - Add a trace for a request oriented action
+ * @q:          queue the io is for
+ * @rq:         the source request
+ * @what:       the action
+ *
+ * Description:
+ *     Records an action against a request. Will log the bio offset + size.
+ *
+ **/
+static void blk_add_trace_rq(struct request_queue *q, struct request *rq,
+                                    u32 what)
+{
+        struct blk_trace *bt = q->blk_trace;
+        int rw = rq->cmd_flags & 0x03;
+
+        if (likely(!bt))
+                return;
+
+        if (blk_discard_rq(rq))
+                rw |= (1 << BIO_RW_DISCARD);
+
+        if (blk_pc_request(rq)) {
+                what |= BLK_TC_ACT(BLK_TC_PC);
+                __blk_add_trace(bt, 0, rq->data_len, rw, what, rq->errors,
+                                sizeof(rq->cmd), rq->cmd);
+        } else  {
+                what |= BLK_TC_ACT(BLK_TC_FS);
+                __blk_add_trace(bt, rq->hard_sector, rq->hard_nr_sectors << 9,
+                                rw, what, rq->errors, 0, NULL);
+        }
+}
+
+static void blk_add_trace_rq_abort(struct request_queue *q, struct request *rq)
+{
+        blk_add_trace_rq(q, rq, BLK_TA_ABORT);
+}
+
+static void blk_add_trace_rq_insert(struct request_queue *q, struct request *rq)
+{
+        blk_add_trace_rq(q, rq, BLK_TA_INSERT);
+}
+
+static void blk_add_trace_rq_issue(struct request_queue *q, struct request *rq)
+{
+        blk_add_trace_rq(q, rq, BLK_TA_ISSUE);
+}
+
+static void blk_add_trace_rq_requeue(struct request_queue *q, struct request *rq)
+{
+        blk_add_trace_rq(q, rq, BLK_TA_REQUEUE);
+}
+
+static void blk_add_trace_rq_complete(struct request_queue *q, struct request *rq)
+{
+        blk_add_trace_rq(q, rq, BLK_TA_COMPLETE);
+}
+
+/**
+ * blk_add_trace_bio - Add a trace for a bio oriented action
+ * @q:          queue the io is for
+ * @bio:        the source bio
+ * @what:       the action
+ *
+ * Description:
+ *     Records an action against a bio. Will log the bio offset + size.
+ *
+ **/
+static void blk_add_trace_bio(struct request_queue *q, struct bio *bio,
+                                     u32 what)
+{
+        struct blk_trace *bt = q->blk_trace;
+
+        if (likely(!bt))
+                return;
+
+        __blk_add_trace(bt, bio->bi_sector, bio->bi_size, bio->bi_rw, what,
+                        !bio_flagged(bio, BIO_UPTODATE), 0, NULL);
+}
+
+static void blk_add_trace_bio_bounce(struct request_queue *q, struct bio *bio)
+{
+        blk_add_trace_bio(q, bio, BLK_TA_BOUNCE);
+}
+
+static void blk_add_trace_bio_complete(struct request_queue *q, struct bio *bio)
+{
+        blk_add_trace_bio(q, bio, BLK_TA_COMPLETE);
+}
+
+static void blk_add_trace_bio_backmerge(struct request_queue *q, struct bio *bio)
+{
+        blk_add_trace_bio(q, bio, BLK_TA_BACKMERGE);
+}
+
+static void blk_add_trace_bio_frontmerge(struct request_queue *q, struct bio *bio)
+{
+        blk_add_trace_bio(q, bio, BLK_TA_FRONTMERGE);
+}
+
+static void blk_add_trace_bio_queue(struct request_queue *q, struct bio *bio)
+{
+        blk_add_trace_bio(q, bio, BLK_TA_QUEUE);
+}
+
+static void blk_add_trace_getrq(struct request_queue *q, struct bio *bio, int rw)
+{
+        if (bio)
+                blk_add_trace_bio(q, bio, BLK_TA_GETRQ);
+        else {
+                struct blk_trace *bt = q->blk_trace;
+
+                if (bt)
+                        __blk_add_trace(bt, 0, 0, rw, BLK_TA_GETRQ, 0, 0, NULL);
+        }
+}
+
+
+static void blk_add_trace_sleeprq(struct request_queue *q, struct bio *bio, int rw)
+{
+        if (bio)
+                blk_add_trace_bio(q, bio, BLK_TA_SLEEPRQ);
+        else {
+                struct blk_trace *bt = q->blk_trace;
+
+                if (bt)
+                        __blk_add_trace(bt, 0, 0, rw, BLK_TA_SLEEPRQ, 0, 0, NULL);
+        }
+}
+
+static void blk_add_trace_plug(struct request_queue *q)
+{
+        struct blk_trace *bt = q->blk_trace;
+
+        if (bt)
+                __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL);
+}
+
+static void blk_add_trace_unplug_io(struct request_queue *q)
+{
+        struct blk_trace *bt = q->blk_trace;
+
+        if (bt) {
+                unsigned int pdu = q->rq.count[READ] + q->rq.count[WRITE];
+                __be64 rpdu = cpu_to_be64(pdu);
+
+                __blk_add_trace(bt, 0, 0, 0, BLK_TA_UNPLUG_IO, 0,
+                                sizeof(rpdu), &rpdu);
+        }
+}
+
+static void blk_add_trace_unplug_timer(struct request_queue *q)
+{
+        struct blk_trace *bt = q->blk_trace;
+
+        if (bt) {
+                unsigned int pdu = q->rq.count[READ] + q->rq.count[WRITE];
+                __be64 rpdu = cpu_to_be64(pdu);
+
+                __blk_add_trace(bt, 0, 0, 0, BLK_TA_UNPLUG_TIMER, 0,
+                                sizeof(rpdu), &rpdu);
+        }
+}
+
+static void blk_add_trace_split(struct request_queue *q, struct bio *bio,
+                                unsigned int pdu)
+{
+        struct blk_trace *bt = q->blk_trace;
+
+        if (bt) {
+                __be64 rpdu = cpu_to_be64(pdu);
+
+                __blk_add_trace(bt, bio->bi_sector, bio->bi_size, bio->bi_rw,
+                                BLK_TA_SPLIT, !bio_flagged(bio, BIO_UPTODATE),
+                                sizeof(rpdu), &rpdu);
+        }
+}
+
+/**
+ * blk_add_trace_remap - Add a trace for a remap operation
+ * @q:          queue the io is for
+ * @bio:        the source bio
+ * @dev:        target device
+ * @from:       source sector
+ * @to:         target sector
+ *
+ * Description:
+ *     Device mapper or raid target sometimes need to split a bio because
+ *     it spans a stripe (or similar). Add a trace for that action.
+ *
+ **/
+static void blk_add_trace_remap(struct request_queue *q, struct bio *bio,
+                                       dev_t dev, sector_t from, sector_t to)
+{
+        struct blk_trace *bt = q->blk_trace;
+        struct blk_io_trace_remap r;
+
+        if (likely(!bt))
+                return;
+
+        r.device = cpu_to_be32(dev);
+        r.device_from = cpu_to_be32(bio->bi_bdev->bd_dev);
+        r.sector = cpu_to_be64(to);
+
+        __blk_add_trace(bt, from, bio->bi_size, bio->bi_rw, BLK_TA_REMAP,
+                        !bio_flagged(bio, BIO_UPTODATE), sizeof(r), &r);
+}
+
+/**
+ * blk_add_driver_data - Add binary message with driver-specific data
+ * @q:          queue the io is for
+ * @rq:         io request
+ * @data:       driver-specific data
+ * @len:        length of driver-specific data
+ *
+ * Description:
+ *     Some drivers might want to write driver-specific data per request.
+ *
+ **/
+void blk_add_driver_data(struct request_queue *q,
+                         struct request *rq,
+                         void *data, size_t len)
+{
+        struct blk_trace *bt = q->blk_trace;
+
+        if (likely(!bt))
+                return;
+
+        if (blk_pc_request(rq))
+                __blk_add_trace(bt, 0, rq->data_len, 0, BLK_TA_DRV_DATA,
+                                rq->errors, len, data);
+        else
+                __blk_add_trace(bt, rq->hard_sector, rq->hard_nr_sectors << 9,
+                                0, BLK_TA_DRV_DATA, rq->errors, len, data);
+}
+EXPORT_SYMBOL_GPL(blk_add_driver_data);
+
+static int blk_register_tracepoints(void)
+{
+        int ret;
+
+        ret = register_trace_block_rq_abort(blk_add_trace_rq_abort);
+        WARN_ON(ret);
+        ret = register_trace_block_rq_insert(blk_add_trace_rq_insert);
+        WARN_ON(ret);
+        ret = register_trace_block_rq_issue(blk_add_trace_rq_issue);
+        WARN_ON(ret);
+        ret = register_trace_block_rq_requeue(blk_add_trace_rq_requeue);
+        WARN_ON(ret);
+        ret = register_trace_block_rq_complete(blk_add_trace_rq_complete);
+        WARN_ON(ret);
+        ret = register_trace_block_bio_bounce(blk_add_trace_bio_bounce);
+        WARN_ON(ret);
+        ret = register_trace_block_bio_complete(blk_add_trace_bio_complete);
+        WARN_ON(ret);
+        ret = register_trace_block_bio_backmerge(blk_add_trace_bio_backmerge);
+        WARN_ON(ret);
+        ret = register_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge);
+        WARN_ON(ret);
+        ret = register_trace_block_bio_queue(blk_add_trace_bio_queue);
+        WARN_ON(ret);
+        ret = register_trace_block_getrq(blk_add_trace_getrq);
+        WARN_ON(ret);
+        ret = register_trace_block_sleeprq(blk_add_trace_sleeprq);
+        WARN_ON(ret);
+        ret = register_trace_block_plug(blk_add_trace_plug);
+        WARN_ON(ret);
+        ret = register_trace_block_unplug_timer(blk_add_trace_unplug_timer);
+        WARN_ON(ret);
+        ret = register_trace_block_unplug_io(blk_add_trace_unplug_io);
+        WARN_ON(ret);
+        ret = register_trace_block_split(blk_add_trace_split);
+        WARN_ON(ret);
+        ret = register_trace_block_remap(blk_add_trace_remap);
+        WARN_ON(ret);
+        return 0;
+}
+
+static void blk_unregister_tracepoints(void)
+{
+        unregister_trace_block_remap(blk_add_trace_remap);
+        unregister_trace_block_split(blk_add_trace_split);
+        unregister_trace_block_unplug_io(blk_add_trace_unplug_io);
+        unregister_trace_block_unplug_timer(blk_add_trace_unplug_timer);
+        unregister_trace_block_plug(blk_add_trace_plug);
+        unregister_trace_block_sleeprq(blk_add_trace_sleeprq);
+        unregister_trace_block_getrq(blk_add_trace_getrq);
+        unregister_trace_block_bio_queue(blk_add_trace_bio_queue);
+        unregister_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge);
+        unregister_trace_block_bio_backmerge(blk_add_trace_bio_backmerge);
+        unregister_trace_block_bio_complete(blk_add_trace_bio_complete);
+        unregister_trace_block_bio_bounce(blk_add_trace_bio_bounce);
+        unregister_trace_block_rq_complete(blk_add_trace_rq_complete);
+        unregister_trace_block_rq_requeue(blk_add_trace_rq_requeue);
+        unregister_trace_block_rq_issue(blk_add_trace_rq_issue);
+        unregister_trace_block_rq_insert(blk_add_trace_rq_insert);
+        unregister_trace_block_rq_abort(blk_add_trace_rq_abort);
+
+        tracepoint_synchronize_unregister();
+}
diff --git a/block/elevator.c b/block/elevator.c
index a6951f76ba0c..86836dd179c0 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -33,6 +33,7 @@
 #include <linux/compiler.h>
 #include <linux/delay.h>
 #include <linux/blktrace_api.h>
+#include <trace/block.h>
 #include <linux/hash.h>
 #include <linux/uaccess.h>
 
@@ -41,6 +42,8 @@
 static DEFINE_SPINLOCK(elv_list_lock);
 static LIST_HEAD(elv_list);
 
+DEFINE_TRACE(block_rq_abort);
+
 /*
  * Merge hash stuff.
  */
@@ -52,6 +55,9 @@ static const int elv_hash_shift = 6;
 #define rq_hash_key(rq)         ((rq)->sector + (rq)->nr_sectors)
 #define ELV_ON_HASH(rq)         (!hlist_unhashed(&(rq)->hash))
 
+DEFINE_TRACE(block_rq_insert);
+DEFINE_TRACE(block_rq_issue);
+
 /*
  * Query io scheduler to see if the current process issuing bio may be
  * merged with rq.
@@ -586,7 +592,7 @@ void elv_insert(struct request_queue *q, struct request *rq, int where)
         unsigned ordseq;
         int unplug_it = 1;
 
-        blk_add_trace_rq(q, rq, BLK_TA_INSERT);
+        trace_block_rq_insert(q, rq);
 
         rq->q = q;
 
@@ -772,7 +778,7 @@ struct request *elv_next_request(struct request_queue *q)
                          * not be passed by new incoming requests
                          */
                         rq->cmd_flags |= REQ_STARTED;
-                        blk_add_trace_rq(q, rq, BLK_TA_ISSUE);
+                        trace_block_rq_issue(q, rq);
                 }
 
                 if (!q->boundary_rq || q->boundary_rq == rq) {
@@ -914,7 +920,7 @@ void elv_abort_queue(struct request_queue *q)
         while (!list_empty(&q->queue_head)) {
                 rq = list_entry_rq(q->queue_head.next);
                 rq->cmd_flags |= REQ_QUIET;
-                blk_add_trace_rq(q, rq, BLK_TA_ABORT);
+                trace_block_rq_abort(q, rq);
                 __blk_end_request(rq, -EIO, blk_rq_bytes(rq));
         }
 }