[PATCH v4 0/2] blk-mq: fix start_time_ns and alloc_time_ns for pre-allocated rq

[PATCH v4 0/2] blk-mq: fix start_time_ns and alloc_time_ns for pre-allocated rq

[chengming.zhou]

From: Chengming Zhou <zhouchengming@bytedance.com>

Hello,

This patchset fix start_time_ns and alloc_time_ns for pre-allocated rq.

patch 1 is preparation that we move data->rq_flags initialization code
outside blk_mq_rq_ctx_init() to its caller, since we will need it to
tell if we need to get start_time_ns.

patch 2 is the fix patch that we set alloc_time_ns and start_time_ns
to now when the pre-allocated rq is actually used and we skip setting
during pre-allocation.

v4:
 - Combine the last two patches into one patch for clearer review.
 - Use blk_mq_alloc_data to pass start_time_ns instead of passing down
   yet another parameter. Thanks Christoph Hellwig.
 - [v3] https://lore.kernel.org/all/20230628124546.1056698-1-chengming.zhou@linux.dev/

v3:
 - Skip setting the alloc_time_ns and start_time_ns during pre-allocation,
   which is clearer, as suggested by Tejun.
 - [v2] https://lore.kernel.org/all/20230626050405.781253-1-chengming.zhou@linux.dev/

v2:
 - Let blk_mq_rq_ctx_init() receive start_time_ns for batched time setting.
 - Set alloc_time_ns and start_time_ns when the pre-allocated rq is actually
   used, as suggested by Tejun.
 - [v1] https://lore.kernel.org/all/20230601053919.3639954-1-chengming.zhou@linux.dev/

Chengming Zhou (2):
  blk-mq: move data->rq_flags initialization code
  blk-mq: fix start_time_ns and alloc_time_ns for pre-allocated rq

 block/blk-mq.c         | 87 ++++++++++++++++++++++++------------------
 block/blk-mq.h         |  1 +
 include/linux/blk-mq.h |  6 +--
 3 files changed, 53 insertions(+), 41 deletions(-)

-- 
2.39.2

[PATCH v4 1/2] blk-mq: move data->rq_flags initialization code

[chengming.zhou]

From: Chengming Zhou <zhouchengming@bytedance.com>

The blk_mq_rq_ctx_init() has some data->rq_flags initialization:

```
if (data->flags & BLK_MQ_REQ_PM)
	data->rq_flags |= RQF_PM;
if (blk_queue_io_stat(q))
	data->rq_flags |= RQF_IO_STAT;
rq->rq_flags = data->rq_flags;
```

We will need this data->rq_flags to tell if we need start_time_ns, so
we need to put these initialization in the callers of blk_mq_rq_ctx_init().

Now we basically have two callers, the 1st is general __blk_mq_alloc_requests(),
the 2nd is the special blk_mq_alloc_request_hctx(). So change the 2nd caller
to reuse the 1st __blk_mq_alloc_requests() to avoid code duplication.

This is safe because blk_mq_alloc_request_hctx() always has BLK_MQ_REQ_NOWAIT
flag, it won't go into the normal retry path when doesn't have free tag.
But it indeed make the general __blk_mq_alloc_requests() more complex.

Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
---
 block/blk-mq.c | 46 ++++++++++++++++++----------------------------
 1 file changed, 18 insertions(+), 28 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index decb6ab2d508..c50ef953759f 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -349,11 +349,6 @@ static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data,
 	rq->mq_ctx = ctx;
 	rq->mq_hctx = hctx;
 	rq->cmd_flags = data->cmd_flags;
-
-	if (data->flags & BLK_MQ_REQ_PM)
-		data->rq_flags |= RQF_PM;
-	if (blk_queue_io_stat(q))
-		data->rq_flags |= RQF_IO_STAT;
 	rq->rq_flags = data->rq_flags;
 
 	if (data->rq_flags & RQF_SCHED_TAGS) {
@@ -447,6 +442,15 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
 	if (data->cmd_flags & REQ_NOWAIT)
 		data->flags |= BLK_MQ_REQ_NOWAIT;
 
+	if (data->flags & BLK_MQ_REQ_RESERVED)
+		data->rq_flags |= RQF_RESV;
+
+	if (data->flags & BLK_MQ_REQ_PM)
+		data->rq_flags |= RQF_PM;
+
+	if (blk_queue_io_stat(q))
+		data->rq_flags |= RQF_IO_STAT;
+
 	if (q->elevator) {
 		/*
 		 * All requests use scheduler tags when an I/O scheduler is
@@ -471,14 +475,15 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
 	}
 
 retry:
-	data->ctx = blk_mq_get_ctx(q);
-	data->hctx = blk_mq_map_queue(q, data->cmd_flags, data->ctx);
+	/* See blk_mq_alloc_request_hctx() for details */
+	if (!data->ctx) {
+		data->ctx = blk_mq_get_ctx(q);
+		data->hctx = blk_mq_map_queue(q, data->cmd_flags, data->ctx);
+	}
+
 	if (!(data->rq_flags & RQF_SCHED_TAGS))
 		blk_mq_tag_busy(data->hctx);
 
-	if (data->flags & BLK_MQ_REQ_RESERVED)
-		data->rq_flags |= RQF_RESV;
-
 	/*
 	 * Try batched alloc if we want more than 1 tag.
 	 */
@@ -505,6 +510,7 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
 		 * is going away.
 		 */
 		msleep(3);
+		data->ctx = NULL;
 		goto retry;
 	}
 
@@ -613,16 +619,10 @@ struct request *blk_mq_alloc_request_hctx(struct request_queue *q,
 		.cmd_flags	= opf,
 		.nr_tags	= 1,
 	};
-	u64 alloc_time_ns = 0;
 	struct request *rq;
 	unsigned int cpu;
-	unsigned int tag;
 	int ret;
 
-	/* alloc_time includes depth and tag waits */
-	if (blk_queue_rq_alloc_time(q))
-		alloc_time_ns = ktime_get_ns();
-
 	/*
 	 * If the tag allocator sleeps we could get an allocation for a
 	 * different hardware context.  No need to complicate the low level
@@ -653,20 +653,10 @@ struct request *blk_mq_alloc_request_hctx(struct request_queue *q,
 		goto out_queue_exit;
 	data.ctx = __blk_mq_get_ctx(q, cpu);
 
-	if (q->elevator)
-		data.rq_flags |= RQF_SCHED_TAGS;
-	else
-		blk_mq_tag_busy(data.hctx);
-
-	if (flags & BLK_MQ_REQ_RESERVED)
-		data.rq_flags |= RQF_RESV;
-
 	ret = -EWOULDBLOCK;
-	tag = blk_mq_get_tag(&data);
-	if (tag == BLK_MQ_NO_TAG)
+	rq = __blk_mq_alloc_requests(&data);
+	if (!rq)
 		goto out_queue_exit;
-	rq = blk_mq_rq_ctx_init(&data, blk_mq_tags_from_data(&data), tag,
-					alloc_time_ns);
 	rq->__data_len = 0;
 	rq->__sector = (sector_t) -1;
 	rq->bio = rq->biotail = NULL;
-- 
2.39.2

[PATCH v4 2/2] blk-mq: fix start_time_ns and alloc_time_ns for pre-allocated rq

[chengming.zhou]

From: Chengming Zhou <zhouchengming@bytedance.com>

The iocost rely on rq start_time_ns and alloc_time_ns to tell saturation
state of the block device. Most of the time request is allocated after
rq_qos_throttle() and its alloc_time_ns or start_time_ns won't be affected.

But for plug batched allocation introduced by the commit 47c122e35d7e
("block: pre-allocate requests if plug is started and is a batch"), we can
rq_qos_throttle() after the allocation of the request. This is what the
blk_mq_get_cached_request() does.

In this case, the cached request alloc_time_ns or start_time_ns is much
ahead if blocked in any qos ->throttle().

This patch fix it by setting alloc_time_ns and start_time_ns to now
when the pre-allocated rq is actually used. And we skip setting the
alloc_time_ns and start_time_ns during pre-allocation, so just pass 0
in __blk_mq_alloc_requests_batch().

Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
---
 block/blk-mq.c         | 41 +++++++++++++++++++++++++++++++----------
 block/blk-mq.h         |  1 +
 include/linux/blk-mq.h |  6 +++---
 3 files changed, 35 insertions(+), 13 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index c50ef953759f..2849395ca4c6 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -337,6 +337,24 @@ void blk_rq_init(struct request_queue *q, struct request *rq)
 }
 EXPORT_SYMBOL(blk_rq_init);
 
+/* Set alloc and start time when pre-allocated rq is actually used */
+static inline void blk_mq_rq_time_init(struct request *rq)
+{
+	if (blk_mq_need_time_stamp(rq->rq_flags)) {
+		u64 now = ktime_get_ns();
+
+#ifdef CONFIG_BLK_RQ_ALLOC_TIME
+		/*
+		 * The alloc time is only used by iocost for now,
+		 * only possible when blk_mq_need_time_stamp().
+		 */
+		if (blk_queue_rq_alloc_time(rq->q))
+			rq->alloc_time_ns = now;
+#endif
+		rq->start_time_ns = now;
+	}
+}
+
 static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data,
 		struct blk_mq_tags *tags, unsigned int tag, u64 alloc_time_ns)
 {
@@ -360,14 +378,11 @@ static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data,
 	}
 	rq->timeout = 0;
 
-	if (blk_mq_need_time_stamp(rq))
-		rq->start_time_ns = ktime_get_ns();
-	else
-		rq->start_time_ns = 0;
 	rq->part = NULL;
 #ifdef CONFIG_BLK_RQ_ALLOC_TIME
 	rq->alloc_time_ns = alloc_time_ns;
 #endif
+	rq->start_time_ns = data->start_time_ns;
 	rq->io_start_time_ns = 0;
 	rq->stats_sectors = 0;
 	rq->nr_phys_segments = 0;
@@ -397,8 +412,7 @@ static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data,
 }
 
 static inline struct request *
-__blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data,
-		u64 alloc_time_ns)
+__blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data)
 {
 	unsigned int tag, tag_offset;
 	struct blk_mq_tags *tags;
@@ -417,7 +431,7 @@ __blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data,
 		tag = tag_offset + i;
 		prefetch(tags->static_rqs[tag]);
 		tag_mask &= ~(1UL << i);
-		rq = blk_mq_rq_ctx_init(data, tags, tag, alloc_time_ns);
+		rq = blk_mq_rq_ctx_init(data, tags, tag, 0);
 		rq_list_add(data->cached_rq, rq);
 		nr++;
 	}
@@ -488,9 +502,11 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
 	 * Try batched alloc if we want more than 1 tag.
 	 */
 	if (data->nr_tags > 1) {
-		rq = __blk_mq_alloc_requests_batch(data, alloc_time_ns);
-		if (rq)
+		rq = __blk_mq_alloc_requests_batch(data);
+		if (rq) {
+			blk_mq_rq_time_init(rq);
 			return rq;
+		}
 		data->nr_tags = 1;
 	}
 
@@ -514,6 +530,9 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
 		goto retry;
 	}
 
+	if (blk_mq_need_time_stamp(data->rq_flags))
+		data->start_time_ns = ktime_get_ns();
+
 	return blk_mq_rq_ctx_init(data, blk_mq_tags_from_data(data), tag,
 					alloc_time_ns);
 }
@@ -570,6 +589,7 @@ static struct request *blk_mq_alloc_cached_request(struct request_queue *q,
 			return NULL;
 
 		plug->cached_rq = rq_list_next(rq);
+		blk_mq_rq_time_init(rq);
 	}
 
 	rq->cmd_flags = opf;
@@ -1004,7 +1024,7 @@ static inline void __blk_mq_end_request_acct(struct request *rq, u64 now)
 
 inline void __blk_mq_end_request(struct request *rq, blk_status_t error)
 {
-	if (blk_mq_need_time_stamp(rq))
+	if (blk_mq_need_time_stamp(rq->rq_flags))
 		__blk_mq_end_request_acct(rq, ktime_get_ns());
 
 	if (rq->end_io) {
@@ -2891,6 +2911,7 @@ static inline struct request *blk_mq_get_cached_request(struct request_queue *q,
 	plug->cached_rq = rq_list_next(rq);
 	rq_qos_throttle(q, *bio);
 
+	blk_mq_rq_time_init(rq);
 	rq->cmd_flags = (*bio)->bi_opf;
 	INIT_LIST_HEAD(&rq->queuelist);
 	return rq;
diff --git a/block/blk-mq.h b/block/blk-mq.h
index 1743857e0b01..29ed4d05110e 100644
--- a/block/blk-mq.h
+++ b/block/blk-mq.h
@@ -150,6 +150,7 @@ struct blk_mq_alloc_data {
 	unsigned int shallow_depth;
 	blk_opf_t cmd_flags;
 	req_flags_t rq_flags;
+	u64 start_time_ns;
 
 	/* allocate multiple requests/tags in one go */
 	unsigned int nr_tags;
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index f401067ac03a..e8366e9c3388 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -834,9 +834,9 @@ void blk_mq_end_request_batch(struct io_comp_batch *ib);
  * Only need start/end time stamping if we have iostat or
  * blk stats enabled, or using an IO scheduler.
  */
-static inline bool blk_mq_need_time_stamp(struct request *rq)
+static inline bool blk_mq_need_time_stamp(req_flags_t rq_flags)
 {
-	return (rq->rq_flags & (RQF_IO_STAT | RQF_STATS | RQF_USE_SCHED));
+	return (rq_flags & (RQF_IO_STAT | RQF_STATS | RQF_USE_SCHED));
 }
 
 static inline bool blk_mq_is_reserved_rq(struct request *rq)
@@ -860,7 +860,7 @@ static inline bool blk_mq_add_to_batch(struct request *req,
 		iob->complete = complete;
 	else if (iob->complete != complete)
 		return false;
-	iob->need_ts |= blk_mq_need_time_stamp(req);
+	iob->need_ts |= blk_mq_need_time_stamp(req->rq_flags);
 	rq_list_add(&iob->req_list, req);
 	return true;
 }
-- 
2.39.2