[PATCHSET v3 0/5] Alloc batch fixes

[PATCHSET v3 0/5] Alloc batch fixes

[Jens Axboe]

Hi,

Since the new feedback on v1 arrived post v2, here's a v3 that takes
it all into account.

A few fixes related to the batched allocations:

- Have the requests hold a queue reference, flush them on schedule
  unplug as well.

- Make sure the queue matches, could be a mismatch if we're driving
  multiple devices.

Since v2:
- Split both alloc handlers into helpers
- Retained blk_try_enter_queue() the way that it was, don't want to
  inline all of bio_queue_enter(). Added justification for that in the
  commit message.
- Add fops based submit_bio helper
Since v1:
- Reshuffle series to do plug rq alloc helper before enter changes
- Protect submit_bio_checks() by queue enter reference as well

-- 
Jens Axboe

[PATCH 1/5] block: have plug stored requests hold references to the queue

[Jens Axboe]

Requests that were stored in the cache deliberately didn't hold an enter
reference to the queue, instead we grabbed one every time we pulled a
request out of there. That made for awkward logic on freeing the remainder
of the cached list, if needed, where we had to artificially raise the
queue usage count before each free.

Grab references up front for cached plug requests. That's safer, and also
more efficient.

Fixes: 47c122e35d7e ("block: pre-allocate requests if plug is started and is a batch")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-core.c | 2 +-
 block/blk-mq.c   | 7 ++++---
 2 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index fd389a16013c..c2d267b6f910 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1643,7 +1643,7 @@ void blk_flush_plug(struct blk_plug *plug, bool from_schedule)
 		flush_plug_callbacks(plug, from_schedule);
 	if (!rq_list_empty(plug->mq_list))
 		blk_mq_flush_plug_list(plug, from_schedule);
-	if (unlikely(!from_schedule && plug->cached_rq))
+	if (unlikely(!rq_list_empty(plug->cached_rq)))
 		blk_mq_free_plug_rqs(plug);
 }
 
diff --git a/block/blk-mq.c b/block/blk-mq.c
index c68aa0a332e1..5498454c2164 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -410,7 +410,10 @@ __blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data,
 		tag_mask &= ~(1UL << i);
 		rq = blk_mq_rq_ctx_init(data, tags, tag, alloc_time_ns);
 		rq_list_add(data->cached_rq, rq);
+		nr++;
 	}
+	/* caller already holds a reference, add for remainder */
+	percpu_ref_get_many(&data->q->q_usage_counter, nr - 1);
 	data->nr_tags -= nr;
 
 	return rq_list_pop(data->cached_rq);
@@ -630,10 +633,8 @@ void blk_mq_free_plug_rqs(struct blk_plug *plug)
 {
 	struct request *rq;
 
-	while ((rq = rq_list_pop(&plug->cached_rq)) != NULL) {
-		percpu_ref_get(&rq->q->q_usage_counter);
+	while ((rq = rq_list_pop(&plug->cached_rq)) != NULL)
 		blk_mq_free_request(rq);
-	}
 }
 
 static void req_bio_endio(struct request *rq, struct bio *bio,
-- 
2.33.1

Re: [PATCH 1/5] block: have plug stored requests hold references to the queue

[Christoph Hellwig]

On Thu, Nov 04, 2021 at 12:21:57PM -0600, Jens Axboe wrote:
> Requests that were stored in the cache deliberately didn't hold an enter
> reference to the queue, instead we grabbed one every time we pulled a
> request out of there. That made for awkward logic on freeing the remainder
> of the cached list, if needed, where we had to artificially raise the
> queue usage count before each free.
> 
> Grab references up front for cached plug requests. That's safer, and also
> more efficient.

I think it would be useful to add zour explanation why the cached
requests should be flushed at schedule time now here.

Otherwise this looks good:

Reviewed-by: Christoph Hellwig <hch@lst.de>

Re: [PATCH 1/5] block: have plug stored requests hold references to the queue

[Jens Axboe]

On 11/4/21 12:34 PM, Christoph Hellwig wrote:
> On Thu, Nov 04, 2021 at 12:21:57PM -0600, Jens Axboe wrote:
>> Requests that were stored in the cache deliberately didn't hold an enter
>> reference to the queue, instead we grabbed one every time we pulled a
>> request out of there. That made for awkward logic on freeing the remainder
>> of the cached list, if needed, where we had to artificially raise the
>> queue usage count before each free.
>>
>> Grab references up front for cached plug requests. That's safer, and also
>> more efficient.
> 
> I think it would be useful to add zour explanation why the cached
> requests should be flushed at schedule time now here.
> 
> Otherwise this looks good:
> 
> Reviewed-by: Christoph Hellwig <hch@lst.de>

Thanks, I'll add a comment when amending with your reviewed-by.

-- 
Jens Axboe

[PATCH 2/5] block: split request allocation components into helpers

[Jens Axboe]

This is in preparation for a fix, but serves as a cleanup as well moving
the cached vs regular alloc logic out of blk_mq_submit_bio().

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-mq.c | 71 ++++++++++++++++++++++++++++++++++----------------
 1 file changed, 48 insertions(+), 23 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 5498454c2164..dcb413297a96 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2478,6 +2478,51 @@ static inline unsigned short blk_plug_max_rq_count(struct blk_plug *plug)
 	return BLK_MAX_REQUEST_COUNT;
 }
 
+static struct request *blk_mq_get_new_requests(struct request_queue *q,
+					       struct blk_plug *plug,
+					       struct bio *bio)
+{
+	struct blk_mq_alloc_data data = {
+		.q		= q,
+		.nr_tags	= 1,
+		.cmd_flags	= bio->bi_opf,
+	};
+	struct request *rq;
+
+	if (plug) {
+		data.nr_tags = plug->nr_ios;
+		plug->nr_ios = 1;
+		data.cached_rq = &plug->cached_rq;
+	}
+
+	rq = __blk_mq_alloc_requests(&data);
+	if (rq)
+		return rq;
+
+	rq_qos_cleanup(q, bio);
+	if (bio->bi_opf & REQ_NOWAIT)
+		bio_wouldblock_error(bio);
+	return NULL;
+}
+
+static inline struct request *blk_mq_get_request(struct request_queue *q,
+						 struct blk_plug *plug,
+						 struct bio *bio)
+{
+	if (plug) {
+		struct request *rq;
+
+		rq = rq_list_peek(&plug->cached_rq);
+		if (rq) {
+			plug->cached_rq = rq_list_next(rq);
+			INIT_LIST_HEAD(&rq->queuelist);
+			return rq;
+		}
+	}
+
+	return blk_mq_get_new_requests(q, plug, bio);
+}
+
 /**
  * blk_mq_submit_bio - Create and send a request to block device.
  * @bio: Bio pointer.
@@ -2518,29 +2563,9 @@ void blk_mq_submit_bio(struct bio *bio)
 	rq_qos_throttle(q, bio);
 
 	plug = blk_mq_plug(q, bio);
-	if (plug && plug->cached_rq) {
-		rq = rq_list_pop(&plug->cached_rq);
-		INIT_LIST_HEAD(&rq->queuelist);
-	} else {
-		struct blk_mq_alloc_data data = {
-			.q		= q,
-			.nr_tags	= 1,
-			.cmd_flags	= bio->bi_opf,
-		};
-
-		if (plug) {
-			data.nr_tags = plug->nr_ios;
-			plug->nr_ios = 1;
-			data.cached_rq = &plug->cached_rq;
-		}
-		rq = __blk_mq_alloc_requests(&data);
-		if (unlikely(!rq)) {
-			rq_qos_cleanup(q, bio);
-			if (bio->bi_opf & REQ_NOWAIT)
-				bio_wouldblock_error(bio);
-			goto queue_exit;
-		}
-	}
+	rq = blk_mq_get_request(q, plug, bio);
+	if (unlikely(!rq))
+		goto queue_exit;
 
 	trace_block_getrq(bio);
 
-- 
2.33.1

Re: [PATCH 2/5] block: split request allocation components into helpers

[Christoph Hellwig]

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 3/5] block: make blk_try_enter_queue() available for blk-mq

[Jens Axboe]

Just a prep patch for shifting the queue enter logic. This moves the
expected fast path inline, and leaves bio_queue_enter() as an
out-of-line function call. We don't want to inline the latter, as it's
mostly slow path code.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-core.c | 26 +-------------------------
 block/blk.h      | 25 +++++++++++++++++++++++++
 2 files changed, 26 insertions(+), 25 deletions(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index c2d267b6f910..e00f5a2287cc 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -386,30 +386,6 @@ void blk_cleanup_queue(struct request_queue *q)
 }
 EXPORT_SYMBOL(blk_cleanup_queue);
 
-static bool blk_try_enter_queue(struct request_queue *q, bool pm)
-{
-	rcu_read_lock();
-	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
-		goto fail;
-
-	/*
-	 * The code that increments the pm_only counter must ensure that the
-	 * counter is globally visible before the queue is unfrozen.
-	 */
-	if (blk_queue_pm_only(q) &&
-	    (!pm || queue_rpm_status(q) == RPM_SUSPENDED))
-		goto fail_put;
-
-	rcu_read_unlock();
-	return true;
-
-fail_put:
-	blk_queue_exit(q);
-fail:
-	rcu_read_unlock();
-	return false;
-}
-
 /**
  * blk_queue_enter() - try to increase q->q_usage_counter
  * @q: request queue pointer
@@ -442,7 +418,7 @@ int blk_queue_enter(struct request_queue *q, blk_mq_req_flags_t flags)
 	return 0;
 }
 
-static inline int bio_queue_enter(struct bio *bio)
+int bio_queue_enter(struct bio *bio)
 {
 	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
 
diff --git a/block/blk.h b/block/blk.h
index 7afffd548daf..f7371d3b1522 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -55,6 +55,31 @@ void blk_free_flush_queue(struct blk_flush_queue *q);
 void blk_freeze_queue(struct request_queue *q);
 void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
 void blk_queue_start_drain(struct request_queue *q);
+int bio_queue_enter(struct bio *bio);
+
+static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
+{
+	rcu_read_lock();
+	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
+		goto fail;
+
+	/*
+	 * The code that increments the pm_only counter must ensure that the
+	 * counter is globally visible before the queue is unfrozen.
+	 */
+	if (blk_queue_pm_only(q) &&
+	    (!pm || queue_rpm_status(q) == RPM_SUSPENDED))
+		goto fail_put;
+
+	rcu_read_unlock();
+	return true;
+
+fail_put:
+	blk_queue_exit(q);
+fail:
+	rcu_read_unlock();
+	return false;
+}
 
 #define BIO_INLINE_VECS 4
 struct bio_vec *bvec_alloc(mempool_t *pool, unsigned short *nr_vecs,
-- 
2.33.1

[PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

Retain the old logic for the fops based submit, but for our internal
blk_mq_submit_bio(), move the queue entering logic into the core
function itself.

We need to be a bit careful if going into the scheduler, as a scheduler
or queue mappings can arbitrarily change before we have entered the queue.
Have the bio scheduler mapping do that separately, it's a very cheap
operation compared to actually doing merging locking and lookups.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-core.c     | 25 +++++++++++++------------
 block/blk-mq-sched.c | 13 ++++++++++---
 block/blk-mq.c       | 36 ++++++++++++++++++++++++++----------
 block/blk.h          |  1 +
 4 files changed, 50 insertions(+), 25 deletions(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index e00f5a2287cc..70cfac1d7fe1 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -746,7 +746,7 @@ static inline blk_status_t blk_check_zone_append(struct request_queue *q,
 	return BLK_STS_OK;
 }
 
-static noinline_for_stack bool submit_bio_checks(struct bio *bio)
+noinline_for_stack bool submit_bio_checks(struct bio *bio)
 {
 	struct block_device *bdev = bio->bi_bdev;
 	struct request_queue *q = bdev_get_queue(bdev);
@@ -864,22 +864,23 @@ static noinline_for_stack bool submit_bio_checks(struct bio *bio)
 	return false;
 }
 
-static void __submit_bio(struct bio *bio)
+static void __submit_bio_fops(struct gendisk *disk, struct bio *bio)
 {
-	struct gendisk *disk = bio->bi_bdev->bd_disk;
-
 	if (unlikely(bio_queue_enter(bio) != 0))
 		return;
+	if (submit_bio_checks(bio) && blk_crypto_bio_prep(&bio))
+		disk->fops->submit_bio(bio);
+	blk_queue_exit(disk->queue);
+}
 
-	if (!submit_bio_checks(bio) || !blk_crypto_bio_prep(&bio))
-		goto queue_exit;
-	if (!disk->fops->submit_bio) {
+static void __submit_bio(struct bio *bio)
+{
+	struct gendisk *disk = bio->bi_bdev->bd_disk;
+
+	if (!disk->fops->submit_bio)
 		blk_mq_submit_bio(bio);
-		return;
-	}
-	disk->fops->submit_bio(bio);
-queue_exit:
-	blk_queue_exit(disk->queue);
+	else
+		__submit_bio_fops(disk, bio);
 }
 
 /*
diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c
index 4a6789e4398b..4be652fa38e7 100644
--- a/block/blk-mq-sched.c
+++ b/block/blk-mq-sched.c
@@ -370,15 +370,20 @@ bool blk_mq_sched_bio_merge(struct request_queue *q, struct bio *bio,
 	bool ret = false;
 	enum hctx_type type;
 
-	if (e && e->type->ops.bio_merge)
-		return e->type->ops.bio_merge(q, bio, nr_segs);
+	if (bio_queue_enter(bio))
+		return false;
+
+	if (e && e->type->ops.bio_merge) {
+		ret = e->type->ops.bio_merge(q, bio, nr_segs);
+		goto out_put;
+	}
 
 	ctx = blk_mq_get_ctx(q);
 	hctx = blk_mq_map_queue(q, bio->bi_opf, ctx);
 	type = hctx->type;
 	if (!(hctx->flags & BLK_MQ_F_SHOULD_MERGE) ||
 	    list_empty_careful(&ctx->rq_lists[type]))
-		return false;
+		goto out_put;
 
 	/* default per sw-queue merge */
 	spin_lock(&ctx->lock);
@@ -391,6 +396,8 @@ bool blk_mq_sched_bio_merge(struct request_queue *q, struct bio *bio,
 		ret = true;
 
 	spin_unlock(&ctx->lock);
+out_put:
+	blk_queue_exit(q);
 	return ret;
 }
 
diff --git a/block/blk-mq.c b/block/blk-mq.c
index dcb413297a96..875bd0c04409 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2478,6 +2478,13 @@ static inline unsigned short blk_plug_max_rq_count(struct blk_plug *plug)
 	return BLK_MAX_REQUEST_COUNT;
 }
 
+static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
+{
+	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
+		return false;
+	return true;
+}
+
 static struct request *blk_mq_get_new_requests(struct request_queue *q,
 					       struct blk_plug *plug,
 					       struct bio *bio)
@@ -2489,6 +2496,13 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 	};
 	struct request *rq;
 
+	if (unlikely(!blk_mq_queue_enter(q, bio)))
+		return NULL;
+	if (unlikely(!submit_bio_checks(bio)))
+		goto put_exit;
+
+	rq_qos_throttle(q, bio);
+
 	if (plug) {
 		data.nr_tags = plug->nr_ios;
 		plug->nr_ios = 1;
@@ -2502,6 +2516,8 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 	rq_qos_cleanup(q, bio);
 	if (bio->bi_opf & REQ_NOWAIT)
 		bio_wouldblock_error(bio);
+put_exit:
+	blk_queue_exit(q);
 	return NULL;
 }
 
@@ -2514,8 +2530,11 @@ static inline struct request *blk_mq_get_request(struct request_queue *q,
 
 		rq = rq_list_peek(&plug->cached_rq);
 		if (rq) {
+			if (unlikely(!submit_bio_checks(bio)))
+				return NULL;
 			plug->cached_rq = rq_list_next(rq);
 			INIT_LIST_HEAD(&rq->queuelist);
+			rq_qos_throttle(q, bio);
 			return rq;
 		}
 	}
@@ -2546,26 +2565,27 @@ void blk_mq_submit_bio(struct bio *bio)
 	unsigned int nr_segs = 1;
 	blk_status_t ret;
 
+	if (unlikely(!blk_crypto_bio_prep(&bio)))
+		return;
+
 	blk_queue_bounce(q, &bio);
 	if (blk_may_split(q, bio))
 		__blk_queue_split(q, &bio, &nr_segs);
 
 	if (!bio_integrity_prep(bio))
-		goto queue_exit;
+		return;
 
 	if (!blk_queue_nomerges(q) && bio_mergeable(bio)) {
 		if (blk_attempt_plug_merge(q, bio, nr_segs, &same_queue_rq))
-			goto queue_exit;
+			return;
 		if (blk_mq_sched_bio_merge(q, bio, nr_segs))
-			goto queue_exit;
+			return;
 	}
 
-	rq_qos_throttle(q, bio);
-
 	plug = blk_mq_plug(q, bio);
 	rq = blk_mq_get_request(q, plug, bio);
 	if (unlikely(!rq))
-		goto queue_exit;
+		return;
 
 	trace_block_getrq(bio);
 
@@ -2646,10 +2666,6 @@ void blk_mq_submit_bio(struct bio *bio)
 		/* Default case. */
 		blk_mq_sched_insert_request(rq, false, true, true);
 	}
-
-	return;
-queue_exit:
-	blk_queue_exit(q);
 }
 
 static size_t order_to_size(unsigned int order)
diff --git a/block/blk.h b/block/blk.h
index f7371d3b1522..79c98ced59c8 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -56,6 +56,7 @@ void blk_freeze_queue(struct request_queue *q);
 void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
 void blk_queue_start_drain(struct request_queue *q);
 int bio_queue_enter(struct bio *bio);
+bool submit_bio_checks(struct bio *bio);
 
 static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
 {
-- 
2.33.1

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Christoph Hellwig]

> +static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
> +{
> +	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
> +		return false;
> +	return true;
> +}

Didn't we just agree on splitting bio_queue_enter into an inline helper
and an out of line slowpath instead?

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

On 11/4/21 12:36 PM, Christoph Hellwig wrote:
>> +static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
>> +{
>> +	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
>> +		return false;
>> +	return true;
>> +}
> 
> Didn't we just agree on splitting bio_queue_enter into an inline helper
> and an out of line slowpath instead?

See cover letter, and I also added to the commit message of this one. I do
think this approach is better, as bio_queue_enter() itself is just slow
path and there's no point polluting the code with 90% of what's in there.

Hence I kept it as-is.

-- 
Jens Axboe

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Christoph Hellwig]

On Thu, Nov 04, 2021 at 12:37:25PM -0600, Jens Axboe wrote:
> On 11/4/21 12:36 PM, Christoph Hellwig wrote:
> >> +static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
> >> +{
> >> +	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
> >> +		return false;
> >> +	return true;
> >> +}
> > 
> > Didn't we just agree on splitting bio_queue_enter into an inline helper
> > and an out of line slowpath instead?
> 
> See cover letter, and I also added to the commit message of this one. I do
> think this approach is better, as bio_queue_enter() itself is just slow
> path and there's no point polluting the code with 90% of what's in there.
> 
> Hence I kept it as-is.

Well, let me reword this then:  why do you think the above is
blk-mq secific and should not be used by every other caller of
bio_queue_enter as well?  In other words, why not rename
bio_queue_enter __bio_queue_enter and make the above the public
bio_queue_enter interface then?

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

On 11/4/21 12:39 PM, Christoph Hellwig wrote:
> On Thu, Nov 04, 2021 at 12:37:25PM -0600, Jens Axboe wrote:
>> On 11/4/21 12:36 PM, Christoph Hellwig wrote:
>>>> +static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
>>>> +{
>>>> +	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
>>>> +		return false;
>>>> +	return true;
>>>> +}
>>>
>>> Didn't we just agree on splitting bio_queue_enter into an inline helper
>>> and an out of line slowpath instead?
>>
>> See cover letter, and I also added to the commit message of this one. I do
>> think this approach is better, as bio_queue_enter() itself is just slow
>> path and there's no point polluting the code with 90% of what's in there.
>>
>> Hence I kept it as-is.
> 
> Well, let me reword this then:  why do you think the above is
> blk-mq secific and should not be used by every other caller of
> bio_queue_enter as well?  In other words, why not rename
> bio_queue_enter __bio_queue_enter and make the above the public
> bio_queue_enter interface then?

OK, that I can agree too. I'll respin it as such. Gets the job done as
well.

-- 
Jens Axboe

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

On 11/4/21 12:40 PM, Jens Axboe wrote:
> On 11/4/21 12:39 PM, Christoph Hellwig wrote:
>> On Thu, Nov 04, 2021 at 12:37:25PM -0600, Jens Axboe wrote:
>>> On 11/4/21 12:36 PM, Christoph Hellwig wrote:
>>>>> +static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
>>>>> +{
>>>>> +	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
>>>>> +		return false;
>>>>> +	return true;
>>>>> +}
>>>>
>>>> Didn't we just agree on splitting bio_queue_enter into an inline helper
>>>> and an out of line slowpath instead?
>>>
>>> See cover letter, and I also added to the commit message of this one. I do
>>> think this approach is better, as bio_queue_enter() itself is just slow
>>> path and there's no point polluting the code with 90% of what's in there.
>>>
>>> Hence I kept it as-is.
>>
>> Well, let me reword this then:  why do you think the above is
>> blk-mq secific and should not be used by every other caller of
>> bio_queue_enter as well?  In other words, why not rename
>> bio_queue_enter __bio_queue_enter and make the above the public
>> bio_queue_enter interface then?
> 
> OK, that I can agree too. I'll respin it as such. Gets the job done as
> well.

Ala:


diff --git a/block/blk-core.c b/block/blk-core.c
index c2d267b6f910..0084067949d8 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -386,30 +386,6 @@ void blk_cleanup_queue(struct request_queue *q)
 }
 EXPORT_SYMBOL(blk_cleanup_queue);
 
-static bool blk_try_enter_queue(struct request_queue *q, bool pm)
-{
-	rcu_read_lock();
-	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
-		goto fail;
-
-	/*
-	 * The code that increments the pm_only counter must ensure that the
-	 * counter is globally visible before the queue is unfrozen.
-	 */
-	if (blk_queue_pm_only(q) &&
-	    (!pm || queue_rpm_status(q) == RPM_SUSPENDED))
-		goto fail_put;
-
-	rcu_read_unlock();
-	return true;
-
-fail_put:
-	blk_queue_exit(q);
-fail:
-	rcu_read_unlock();
-	return false;
-}
-
 /**
  * blk_queue_enter() - try to increase q->q_usage_counter
  * @q: request queue pointer
@@ -442,10 +418,8 @@ int blk_queue_enter(struct request_queue *q, blk_mq_req_flags_t flags)
 	return 0;
 }
 
-static inline int bio_queue_enter(struct bio *bio)
+int __bio_queue_enter(struct request_queue *q, struct bio *bio)
 {
-	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
-
 	while (!blk_try_enter_queue(q, false)) {
 		struct gendisk *disk = bio->bi_bdev->bd_disk;
 
diff --git a/block/blk.h b/block/blk.h
index 7afffd548daf..814d9632d43e 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -55,6 +55,40 @@ void blk_free_flush_queue(struct blk_flush_queue *q);
 void blk_freeze_queue(struct request_queue *q);
 void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
 void blk_queue_start_drain(struct request_queue *q);
+int __bio_queue_enter(struct request_queue *q, struct bio *bio);
+
+static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
+{
+	rcu_read_lock();
+	if (!percpu_ref_tryget_live_rcu(&q->q_usage_counter))
+		goto fail;
+
+	/*
+	 * The code that increments the pm_only counter must ensure that the
+	 * counter is globally visible before the queue is unfrozen.
+	 */
+	if (blk_queue_pm_only(q) &&
+	    (!pm || queue_rpm_status(q) == RPM_SUSPENDED))
+		goto fail_put;
+
+	rcu_read_unlock();
+	return true;
+
+fail_put:
+	blk_queue_exit(q);
+fail:
+	rcu_read_unlock();
+	return false;
+}
+
+static inline int bio_queue_enter(struct bio *bio)
+{
+	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
+
+	if (blk_try_enter_queue(q, false))
+		return 0;
+	return __bio_queue_enter(q, bio);
+}
 
 #define BIO_INLINE_VECS 4
 struct bio_vec *bvec_alloc(mempool_t *pool, unsigned short *nr_vecs,

-- 
Jens Axboe

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Christoph Hellwig]

Looks good:

Reviewed-by: Christoph Hellwig <hch@lst.de>

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> Looks good:
> 
> Reviewed-by: Christoph Hellwig <hch@lst.de>

So these two are now:

https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96

which is the one I sent here, and then the next one gets cleaned up to
remove that queue enter helper:

https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b

Can I add your reviewed-by to this last one as well? Only change is the
removal of blk_mq_enter_queue() and the weird construct there, it's just
bio_queue_enter() now.

-- 
Jens Axboe

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Christoph Hellwig]

On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
> On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> > Looks good:
> > 
> > Reviewed-by: Christoph Hellwig <hch@lst.de>
> 
> So these two are now:
> 
> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
> 
> which is the one I sent here, and then the next one gets cleaned up to
> remove that queue enter helper:
> 
> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
> 
> Can I add your reviewed-by to this last one as well? Only change is the
> removal of blk_mq_enter_queue() and the weird construct there, it's just
> bio_queue_enter() now.

Sure.

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

On 11/4/21 1:04 PM, Christoph Hellwig wrote:
> On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
>> On 11/4/21 12:52 PM, Christoph Hellwig wrote:
>>> Looks good:
>>>
>>> Reviewed-by: Christoph Hellwig <hch@lst.de>
>>
>> So these two are now:
>>
>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
>>
>> which is the one I sent here, and then the next one gets cleaned up to
>> remove that queue enter helper:
>>
>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
>>
>> Can I add your reviewed-by to this last one as well? Only change is the
>> removal of blk_mq_enter_queue() and the weird construct there, it's just
>> bio_queue_enter() now.
> 
> Sure.

Thanks, prematurely already done, as you could tell :-)

Thanks for the reviews.

-- 
Jens Axboe

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Geert Uytterhoeven]

 	Hi Jens,

On Thu, 4 Nov 2021, Jens Axboe wrote:
> On 11/4/21 1:04 PM, Christoph Hellwig wrote:
>> On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
>>> On 11/4/21 12:52 PM, Christoph Hellwig wrote:
>>>> Looks good:
>>>>
>>>> Reviewed-by: Christoph Hellwig <hch@lst.de>
>>>
>>> So these two are now:
>>>
>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
>>>
>>> which is the one I sent here, and then the next one gets cleaned up to
>>> remove that queue enter helper:
>>>
>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
>>>
>>> Can I add your reviewed-by to this last one as well? Only change is the
>>> removal of blk_mq_enter_queue() and the weird construct there, it's just
>>> bio_queue_enter() now.
>>
>> Sure.
>
> Thanks, prematurely already done, as you could tell :-)

The updated version is now commit 900e080752025f00 ("block: move queue
enter logic into blk_mq_submit_bio()") in Linus' tree.

I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
root device) to this commit, e.g.:

     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
     Buffer I/O error on dev sda1, logical block 0, lost sync page write

     EXT4-fs (sda1): I/O error while writing superblock
     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
     Buffer I/O error on dev sda1, logical block 0, lost sync page write
     EXT4-fs (sda1): I/O error while writing superblock

This may happen either when mounting the root file system (leading to an
unable to mount root fs panic), or later (leading to a read-only
rootfs).

Gr{oetje,eeting}s,

 						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
 							    -- Linus Torvalds

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Martin K. Petersen]

Geert,

> sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> Buffer I/O error on dev sda1, logical block 0, lost sync page write

Peculiar. That write command looks OK to me. I wonder if it's the FUA
bit that trips it?

What does:

# dmesg | grep FUA

say?

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Geert Uytterhoeven]

Hi Martin,

On Thu, Nov 11, 2021 at 2:19 PM Martin K. Petersen
<martin.petersen@oracle.com> wrote:
> > sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> > sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> > sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> > sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> > critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> > Buffer I/O error on dev sda1, logical block 0, lost sync page write
>
> Peculiar. That write command looks OK to me. I wonder if it's the FUA
> bit that trips it?
>
> What does:
>
> # dmesg | grep FUA
>
> say?

sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
support DPO or FUA

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Martin K. Petersen]

Geert,

>> > sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
                                               ^^
                                               FUA

> sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
> support DPO or FUA

The device is correct in rejecting the command, then. But I'm not sure
how the commit in question would cause us to inadvertently set that
flag...

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Michael Schmitz]

Hi Geert,

how easy is that to reproduce?

sd_setup_read_write_cmnd() does not validate the request's FUA flag 
against sdkp->DPOFUA (not suggesting that it should ...). I'd like to 
try and trace when such a mismatch happens.

Cheers,

	Michael

On 12/11/21 03:48, Geert Uytterhoeven wrote:
> Hi Martin,
>
> On Thu, Nov 11, 2021 at 2:19 PM Martin K. Petersen
> <martin.petersen@oracle.com> wrote:
>>> sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>>> sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>>> sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>>> sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>>> critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>>> Buffer I/O error on dev sda1, logical block 0, lost sync page write
>>
>> Peculiar. That write command looks OK to me. I wonder if it's the FUA
>> bit that trips it?
>>
>> What does:
>>
>> # dmesg | grep FUA
>>
>> say?
>
> sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
> support DPO or FUA
>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds
>

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Geert Uytterhoeven]

Hi Michael,

On Thu, Nov 11, 2021 at 10:35 PM Michael Schmitz <schmitzmic@gmail.com> wrote:
> how easy is that to reproduce?

Fairly easy: it happens either on mounting, or after a few seconds booting
into my old Debian userspace.

> sd_setup_read_write_cmnd() does not validate the request's FUA flag
> against sdkp->DPOFUA (not suggesting that it should ...). I'd like to
> try and trace when such a mismatch happens.

Thanks!

> On 12/11/21 03:48, Geert Uytterhoeven wrote:
> > On Thu, Nov 11, 2021 at 2:19 PM Martin K. Petersen
> > <martin.petersen@oracle.com> wrote:
> >>> sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> >>> sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> >>> sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> >>> sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> >>> critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> >>> Buffer I/O error on dev sda1, logical block 0, lost sync page write
> >>
> >> Peculiar. That write command looks OK to me. I wonder if it's the FUA
> >> bit that trips it?
> >>
> >> What does:
> >>
> >> # dmesg | grep FUA
> >>
> >> say?
> >
> > sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
> > support DPO or FUA

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Michael Schmitz]

Hi Geert,

On 12/11/21 20:37, Geert Uytterhoeven wrote:
> Hi Michael,
>
> On Thu, Nov 11, 2021 at 10:35 PM Michael Schmitz <schmitzmic@gmail.com> wrote:
>> how easy is that to reproduce?
>
> Fairly easy: it happens either on mounting, or after a few seconds booting
> into my old Debian userspace.

I must be too thick for this:

EXT4-fs (sda1): Cannot load crc32c driver.
VFS: Cannot open root device "sda1" or unknown-block(8,1): error -80

(linux-block 5833291ab6de, from kernel.org)

Same config on m68k 5.15 works just fine.

But this looks resolved now, so I'll punt for now ...

Cheers,

	Michael

>
>> sd_setup_read_write_cmnd() does not validate the request's FUA flag
>> against sdkp->DPOFUA (not suggesting that it should ...). I'd like to
>> try and trace when such a mismatch happens.
>
> Thanks!
>
>> On 12/11/21 03:48, Geert Uytterhoeven wrote:
>>> On Thu, Nov 11, 2021 at 2:19 PM Martin K. Petersen
>>> <martin.petersen@oracle.com> wrote:
>>>>> sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>>>>> sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>>>>> sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>>>>> sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>>>>> critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>>>>> Buffer I/O error on dev sda1, logical block 0, lost sync page write
>>>>
>>>> Peculiar. That write command looks OK to me. I wonder if it's the FUA
>>>> bit that trips it?
>>>>
>>>> What does:
>>>>
>>>> # dmesg | grep FUA
>>>>
>>>> say?
>>>
>>> sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
>>> support DPO or FUA
>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds
>

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Michael Schmitz]

Hi Geert, Jens,

On 13/11/21 11:34, Michael Schmitz wrote:
> Hi Geert,
>
> On 12/11/21 20:37, Geert Uytterhoeven wrote:
>> Hi Michael,
>>
>> On Thu, Nov 11, 2021 at 10:35 PM Michael Schmitz
>> <schmitzmic@gmail.com> wrote:
>>> how easy is that to reproduce?
>>
>> Fairly easy: it happens either on mounting, or after a few seconds
>> booting
>> into my old Debian userspace.
>
> I must be too thick for this:
>
> EXT4-fs (sda1): Cannot load crc32c driver.
> VFS: Cannot open root device "sda1" or unknown-block(8,1): error -80
>
> (linux-block 5833291ab6de, from kernel.org)
>
> Same config on m68k 5.15 works just fine.

FWIW - I had to revert cad439fc040efe5f4381e3a7d583c5c200dbc186 to 
overcome this, and reproduce Geert's finding.

Cheers,

	Michael

>
> But this looks resolved now, so I'll punt for now ...
>
> Cheers,
>
>     Michael
>
>>
>>> sd_setup_read_write_cmnd() does not validate the request's FUA flag
>>> against sdkp->DPOFUA (not suggesting that it should ...). I'd like to
>>> try and trace when such a mismatch happens.
>>
>> Thanks!
>>
>>> On 12/11/21 03:48, Geert Uytterhoeven wrote:
>>>> On Thu, Nov 11, 2021 at 2:19 PM Martin K. Petersen
>>>> <martin.petersen@oracle.com> wrote:
>>>>>> sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK
>>>>>> driverbyte=DRIVER_OK cmd_age=0s
>>>>>> sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>>>>>> sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>>>>>> sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>>>>>> critical target error, dev sda, sector 1 op 0x1:(WRITE) flags
>>>>>> 0x20800 phys_seg 1 prio class 0
>>>>>> Buffer I/O error on dev sda1, logical block 0, lost sync page write
>>>>>
>>>>> Peculiar. That write command looks OK to me. I wonder if it's the FUA
>>>>> bit that trips it?
>>>>>
>>>>> What does:
>>>>>
>>>>> # dmesg | grep FUA
>>>>>
>>>>> say?
>>>>
>>>> sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
>>>> support DPO or FUA
>>
>> Gr{oetje,eeting}s,
>>
>>                         Geert
>>
>> --
>> Geert Uytterhoeven -- There's lots of Linux beyond ia32 --
>> geert@linux-m68k.org
>>
>> In personal conversations with technical people, I call myself a
>> hacker. But
>> when I'm talking to journalists I just say "programmer" or something
>> like that.
>>                                 -- Linus Torvalds
>>

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Geert Uytterhoeven]

On Fri, Nov 12, 2021 at 11:34 PM Michael Schmitz <schmitzmic@gmail.com> wrote:
> On 12/11/21 20:37, Geert Uytterhoeven wrote:
> > On Thu, Nov 11, 2021 at 10:35 PM Michael Schmitz <schmitzmic@gmail.com> wrote:
> >> how easy is that to reproduce?
> >
> > Fairly easy: it happens either on mounting, or after a few seconds booting
> > into my old Debian userspace.
>
> I must be too thick for this:
>
> EXT4-fs (sda1): Cannot load crc32c driver.
> VFS: Cannot open root device "sda1" or unknown-block(8,1): error -80

I know ;-) (or  :-(
Please cherry-pick commit beaaaa37c664e9af ("crypto: api - Fix boot-up
crash when crypto manager is disabled"), which is now upstream.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Michael Schmitz]

Hi Geert, Martin,

On 12/11/21 20:37, Geert Uytterhoeven wrote:
> Hi Michael,
>
> On Thu, Nov 11, 2021 at 10:35 PM Michael Schmitz <schmitzmic@gmail.com> wrote:
>> how easy is that to reproduce?
>
> Fairly easy: it happens either on mounting, or after a few seconds booting
> into my old Debian userspace.
>
>> sd_setup_read_write_cmnd() does not validate the request's FUA flag
>> against sdkp->DPOFUA (not suggesting that it should ...). I'd like to
>> try and trace when such a mismatch happens.
>
> Thanks!

Also happens with the pata-falcon driver (which stats it doesnt support 
DPO or FUA).

Seen twice, once on boot, once on reboot (FUA bit not set in the SCSI 
command descriptor when the disk doesn't support that, to avoid a panic, 
otherwise still 5833291ab6de). Each time when a remount operation happens:

EXT4-fs (sda1): re-mounted. Opts: . Quota mode: disabled.
------------[ cut here ]------------
WARNING: CPU: 0 PID: 695 at drivers/scsi/sd.c:1271 
sd_init_command+0x6aa/0x886
Modules linked in:
CPU: 0 PID: 695 Comm: mount Not tainted 
5.15.0-atari-fpuemu-clean-signalfixes-remove-set_fs-ataflopfix4+ #88
Stack from 00be1c10:
         00be1c10 0035cc02 0035cc02 002cac7a 00379b57 000004f7 00000009 
00ad80a4
         002cacd4 00379b57 000004f7 0020ac5e 00000009 00000000 00000000 
00209d00
         00000008 00000000 00000001 00000000 00000001 00ab6400 0020ac5e 
00379b57
         000004f7 00000009 00000000 00adc800 00add9d0 00ab6400 00000001 
001269c4
         00003a98 00ad8000 00adc800 00ad80a4 00ab6400 00adc800 00000008 
00020801
         00000000 00000008 00000000 00000001 00000001 00201876 00ad80a4 
00ad80a4
Call Trace: [<002cac7a>] __warn+0x9e/0xb6
  [<002cacd4>] warn_slowpath_fmt+0x42/0x62
  [<0020ac5e>] sd_init_command+0x6aa/0x886
  [<00209d00>] max_write_same_blocks_store+0x6c/0x88
  [<0020ac5e>] sd_init_command+0x6aa/0x886
  [<001269c4>] ext4_get_group_desc+0x0/0xb0
  [<00003a98>] setup_frame+0xba/0x20a
  [<00020801>] _I_CALL_TOP+0xf21/0x1900
  [<00201876>] scsi_queue_rq+0x4f8/0x67e
  [<00194aaa>] __blk_mq_try_issue_directly+0x8a/0xf8
  [<00194adc>] __blk_mq_try_issue_directly+0xbc/0xf8
  [<00194fd0>] blk_mq_try_issue_directly+0x28/0x7a
  [<0019584c>] blk_mq_submit_bio+0x3fc/0x422
  [<0018c6f6>] __submit_bio+0x0/0x78
  [<0018b5cc>] bio_list_pop+0x0/0x1a
  [<00092800>] btf_check_type_match+0x274/0x2e4
  [<00020801>] _I_CALL_TOP+0xf21/0x1900
  [<00002000>] _start+0x0/0x8
  [<0018c7dc>] submit_bio_noacct+0x6e/0x152
  [<00020801>] _I_CALL_TOP+0xf21/0x1900
  [<000f21fc>] test_bit+0x0/0x12
  [<00051f68>] ktime_get_real_seconds+0x0/0x30
  [<000f21fc>] test_bit+0x0/0x12
  [<00051f68>] ktime_get_real_seconds+0x0/0x30
  [<000f329c>] submit_bh_wbc.isra.54+0x178/0x182
  [<000f21fc>] test_bit+0x0/0x12
  [<000f32b8>] submit_bh+0x12/0x18
  [<00020800>] _I_CALL_TOP+0xf20/0x1900
  [<000f362e>] __sync_dirty_buffer+0x78/0xaa
  [<00020800>] _I_CALL_TOP+0xf20/0x1900
  [<00158684>] test_bit+0x0/0x12
  [<0015a508>] ext4_commit_super+0xa8/0xf8
  [<00020800>] _I_CALL_TOP+0xf20/0x1900
  [<0015be82>] ext4_setup_super+0x132/0x1ac
  [<0015c970>] ext4_remount+0x49c/0x57e
  [<00004003>] do_rt_sigreturn+0xe1/0x1b8
  [<000dbe3e>] dentry_lru_isolate_shrink+0x0/0x1c
  [<000efaf0>] legacy_reconfigure+0x30/0x42
  [<000ce664>] reconfigure_super+0xae/0x178
  [<000e4a38>] path_mount+0x274/0x61c
  [<000e4e20>] do_mount+0x40/0x5c
  [<000e50c4>] sys_mount+0xf8/0x10e
  [<00002922>] syscall+0x8/0xc
  [<0018c00c>] blk_cleanup_queue+0x5a/0xc6

---[ end trace b0cd975622ec08c4 ]---
EXT4-fs (sda1): re-mounted. Opts: errors=remount-ro. Quota mode: disabled.
ext4 filesystem being remounted at / supports timestamps until 2038 
(0x7fffffff)

------------[ cut here ]------------
WARNING: CPU: 0 PID: 1174 at drivers/scsi/sd.c:1271 
sd_init_command+0x6aa/0x886
Modules linked in:
CPU: 0 PID: 1174 Comm: umount Tainted: G        W 
5.15.0-atari-fpuemu-clean-signalfixes-remove-set_fs-ataflopfix4+ #88
Stack from 00b6dc70:
         00b6dc70 0035cc02 0035cc02 002cac7a 00379b57 000004f7 00000009 
00ad80a4
         002cacd4 00379b57 000004f7 0020ac5e 00000009 00000000 00000000 
00209d00
         00000008 00000000 00000001 00000000 00000001 00ab6400 0020ac5e 
00379b57
         000004f7 00000009 00000000 00adc800 00add9d0 00ab6400 00000001 
00000000
         00003a98 00ad8000 00adc800 00ad80a4 00ab6400 00adc800 00000008 
00023801
         00000000 00000008 00000000 00000001 00000001 00201876 00ad80a4 
00ad80a4
Call Trace: [<002cac7a>] __warn+0x9e/0xb6
  [<002cacd4>] warn_slowpath_fmt+0x42/0x62
  [<0020ac5e>] sd_init_command+0x6aa/0x886
  [<00209d00>] max_write_same_blocks_store+0x6c/0x88
  [<0020ac5e>] sd_init_command+0x6aa/0x886
  [<00003a98>] setup_frame+0xba/0x20a
  [<00023801>] fp_fadd+0x1d5/0x1e4
  [<00201876>] scsi_queue_rq+0x4f8/0x67e
  [<00194aaa>] __blk_mq_try_issue_directly+0x8a/0xf8
  [<00194adc>] __blk_mq_try_issue_directly+0xbc/0xf8
  [<00194fd0>] blk_mq_try_issue_directly+0x28/0x7a
  [<0019584c>] blk_mq_submit_bio+0x3fc/0x422
  [<0018c6f6>] __submit_bio+0x0/0x78
  [<0018b5cc>] bio_list_pop+0x0/0x1a
  [<00092800>] btf_check_type_match+0x274/0x2e4
  [<00023801>] fp_fadd+0x1d5/0x1e4
  [<00002000>] _start+0x0/0x8
  [<0018c7dc>] submit_bio_noacct+0x6e/0x152
  [<00023801>] fp_fadd+0x1d5/0x1e4
  [<000f21fc>] test_bit+0x0/0x12
  [<000f21fc>] test_bit+0x0/0x12
  [<000f329c>] submit_bh_wbc.isra.54+0x178/0x182
  [<00158684>] test_bit+0x0/0x12
  [<000f21fc>] test_bit+0x0/0x12
  [<000f32b8>] submit_bh+0x12/0x18
  [<00020800>] _I_CALL_TOP+0xf20/0x1900
  [<000f362e>] __sync_dirty_buffer+0x78/0xaa
  [<00020800>] _I_CALL_TOP+0xf20/0x1900
  [<00158684>] test_bit+0x0/0x12
  [<00158684>] test_bit+0x0/0x12
  [<0015a508>] ext4_commit_super+0xa8/0xf8
  [<00020800>] _I_CALL_TOP+0xf20/0x1900
  [<00158684>] test_bit+0x0/0x12
  [<0015c80e>] ext4_remount+0x33a/0x57e
  [<00004003>] do_rt_sigreturn+0xe1/0x1b8
  [<000dbe3e>] dentry_lru_isolate_shrink+0x0/0x1c
  [<000efaf0>] legacy_reconfigure+0x30/0x42
  [<000ce664>] reconfigure_super+0xae/0x178
  [<000e3ab0>] path_umount+0x17c/0x2e8
  [<000e3c62>] ksys_umount+0x46/0x54
  [<000e3c8a>] sys_oldumount+0xa/0xe
  [<00002922>] syscall+0x8/0xc
  [<0018c00c>] blk_cleanup_queue+0x5a/0xc6

---[ end trace b0cd975622ec08c5 ]---
EXT4-fs (sda1): re-mounted. Opts: (null). Quota mode: disabled.
EXT4-fs (sda1): re-mounted. Opts: errors=remount-ro. Quota mode: disabled.

Not sure this helps much.

Cheers,

	Michael

>
>> On 12/11/21 03:48, Geert Uytterhoeven wrote:
>>> On Thu, Nov 11, 2021 at 2:19 PM Martin K. Petersen
>>> <martin.petersen@oracle.com> wrote:
>>>>> sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>>>>> sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>>>>> sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>>>>> sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>>>>> critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>>>>> Buffer I/O error on dev sda1, logical block 0, lost sync page write
>>>>
>>>> Peculiar. That write command looks OK to me. I wonder if it's the FUA
>>>> bit that trips it?
>>>>
>>>> What does:
>>>>
>>>> # dmesg | grep FUA
>>>>
>>>> say?
>>>
>>> sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
>>> support DPO or FUA
>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds
>

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Ming Lei]

On Thu, Nov 11, 2021 at 01:58:38PM +0100, Geert Uytterhoeven wrote:
> 	Hi Jens,
> 
> On Thu, 4 Nov 2021, Jens Axboe wrote:
> > On 11/4/21 1:04 PM, Christoph Hellwig wrote:
> > > On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
> > > > On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> > > > > Looks good:
> > > > > 
> > > > > Reviewed-by: Christoph Hellwig <hch@lst.de>
> > > > 
> > > > So these two are now:
> > > > 
> > > > https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
> > > > 
> > > > which is the one I sent here, and then the next one gets cleaned up to
> > > > remove that queue enter helper:
> > > > 
> > > > https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
> > > > 
> > > > Can I add your reviewed-by to this last one as well? Only change is the
> > > > removal of blk_mq_enter_queue() and the weird construct there, it's just
> > > > bio_queue_enter() now.
> > > 
> > > Sure.
> > 
> > Thanks, prematurely already done, as you could tell :-)
> 
> The updated version is now commit 900e080752025f00 ("block: move queue
> enter logic into blk_mq_submit_bio()") in Linus' tree.
> 
> I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
> root device) to this commit, e.g.:
> 
>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> 
>     EXT4-fs (sda1): I/O error while writing superblock
>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
>     EXT4-fs (sda1): I/O error while writing superblock
> 
> This may happen either when mounting the root file system (leading to an
> unable to mount root fs panic), or later (leading to a read-only
> rootfs).

BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
caused by commit 900e080752025f00, and the following patch can fix it:

- blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge

https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28

Maybe you can try the above patch.

Thanks,
Ming

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Geert Uytterhoeven]

Hi Ming,

On Thu, Nov 11, 2021 at 2:45 PM Ming Lei <ming.lei@redhat.com> wrote:
> On Thu, Nov 11, 2021 at 01:58:38PM +0100, Geert Uytterhoeven wrote:
> > On Thu, 4 Nov 2021, Jens Axboe wrote:
> > > On 11/4/21 1:04 PM, Christoph Hellwig wrote:
> > > > On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
> > > > > On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> > > > > > Looks good:
> > > > > >
> > > > > > Reviewed-by: Christoph Hellwig <hch@lst.de>
> > > > >
> > > > > So these two are now:
> > > > >
> > > > > https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
> > > > >
> > > > > which is the one I sent here, and then the next one gets cleaned up to
> > > > > remove that queue enter helper:
> > > > >
> > > > > https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
> > > > >
> > > > > Can I add your reviewed-by to this last one as well? Only change is the
> > > > > removal of blk_mq_enter_queue() and the weird construct there, it's just
> > > > > bio_queue_enter() now.
> > > >
> > > > Sure.
> > >
> > > Thanks, prematurely already done, as you could tell :-)
> >
> > The updated version is now commit 900e080752025f00 ("block: move queue
> > enter logic into blk_mq_submit_bio()") in Linus' tree.
> >
> > I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
> > root device) to this commit, e.g.:
> >
> >     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> >     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> >     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> >     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> >     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> >     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> >
> >     EXT4-fs (sda1): I/O error while writing superblock
> >     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> >     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> >     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> >     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> >     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> >     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> >     EXT4-fs (sda1): I/O error while writing superblock
> >
> > This may happen either when mounting the root file system (leading to an
> > unable to mount root fs panic), or later (leading to a read-only
> > rootfs).
>
> BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
> caused by commit 900e080752025f00, and the following patch can fix it:
>
> - blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
>
> https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28
>
> Maybe you can try the above patch.

Thanks! I have applied both patches, but it doesn't make a difference.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Ming Lei]

Hi Geert,

On Thu, Nov 11, 2021 at 03:51:28PM +0100, Geert Uytterhoeven wrote:
> Hi Ming,
> 
> On Thu, Nov 11, 2021 at 2:45 PM Ming Lei <ming.lei@redhat.com> wrote:
> > On Thu, Nov 11, 2021 at 01:58:38PM +0100, Geert Uytterhoeven wrote:
> > > On Thu, 4 Nov 2021, Jens Axboe wrote:
> > > > On 11/4/21 1:04 PM, Christoph Hellwig wrote:
> > > > > On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
> > > > > > On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> > > > > > > Looks good:
> > > > > > >
> > > > > > > Reviewed-by: Christoph Hellwig <hch@lst.de>
> > > > > >
> > > > > > So these two are now:
> > > > > >
> > > > > > https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
> > > > > >
> > > > > > which is the one I sent here, and then the next one gets cleaned up to
> > > > > > remove that queue enter helper:
> > > > > >
> > > > > > https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
> > > > > >
> > > > > > Can I add your reviewed-by to this last one as well? Only change is the
> > > > > > removal of blk_mq_enter_queue() and the weird construct there, it's just
> > > > > > bio_queue_enter() now.
> > > > >
> > > > > Sure.
> > > >
> > > > Thanks, prematurely already done, as you could tell :-)
> > >
> > > The updated version is now commit 900e080752025f00 ("block: move queue
> > > enter logic into blk_mq_submit_bio()") in Linus' tree.
> > >
> > > I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
> > > root device) to this commit, e.g.:
> > >
> > >     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> > >     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> > >     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> > >     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> > >     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> > >     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> > >
> > >     EXT4-fs (sda1): I/O error while writing superblock
> > >     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> > >     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> > >     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> > >     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> > >     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> > >     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> > >     EXT4-fs (sda1): I/O error while writing superblock
> > >
> > > This may happen either when mounting the root file system (leading to an
> > > unable to mount root fs panic), or later (leading to a read-only
> > > rootfs).
> >
> > BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
> > caused by commit 900e080752025f00, and the following patch can fix it:
> >
> > - blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
> >
> > https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28
> >
> > Maybe you can try the above patch.
> 
> Thanks! I have applied both patches, but it doesn't make a difference.

Thanks for your test!

Can you try the following patch?


diff --git a/block/blk-mq.c b/block/blk-mq.c
index f511db395c7f..a5ab2f2e9f67 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2517,7 +2517,6 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 	struct blk_mq_alloc_data data = {
 		.q		= q,
 		.nr_tags	= 1,
-		.cmd_flags	= bio->bi_opf,
 	};
 	struct request *rq;
 
@@ -2525,6 +2524,7 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 		return NULL;
 	if (unlikely(!submit_bio_checks(bio)))
 		goto put_exit;
+	data.cmd_flags	= bio->bi_opf;
 	if (blk_mq_attempt_bio_merge(q, bio, nsegs, same_queue_rq))
 		goto put_exit;
 

-- 
Ming

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

On 11/11/21 8:23 AM, Ming Lei wrote:
> Hi Geert,
> 
> On Thu, Nov 11, 2021 at 03:51:28PM +0100, Geert Uytterhoeven wrote:
>> Hi Ming,
>>
>> On Thu, Nov 11, 2021 at 2:45 PM Ming Lei <ming.lei@redhat.com> wrote:
>>> On Thu, Nov 11, 2021 at 01:58:38PM +0100, Geert Uytterhoeven wrote:
>>>> On Thu, 4 Nov 2021, Jens Axboe wrote:
>>>>> On 11/4/21 1:04 PM, Christoph Hellwig wrote:
>>>>>> On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
>>>>>>> On 11/4/21 12:52 PM, Christoph Hellwig wrote:
>>>>>>>> Looks good:
>>>>>>>>
>>>>>>>> Reviewed-by: Christoph Hellwig <hch@lst.de>
>>>>>>>
>>>>>>> So these two are now:
>>>>>>>
>>>>>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
>>>>>>>
>>>>>>> which is the one I sent here, and then the next one gets cleaned up to
>>>>>>> remove that queue enter helper:
>>>>>>>
>>>>>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
>>>>>>>
>>>>>>> Can I add your reviewed-by to this last one as well? Only change is the
>>>>>>> removal of blk_mq_enter_queue() and the weird construct there, it's just
>>>>>>> bio_queue_enter() now.
>>>>>>
>>>>>> Sure.
>>>>>
>>>>> Thanks, prematurely already done, as you could tell :-)
>>>>
>>>> The updated version is now commit 900e080752025f00 ("block: move queue
>>>> enter logic into blk_mq_submit_bio()") in Linus' tree.
>>>>
>>>> I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
>>>> root device) to this commit, e.g.:
>>>>
>>>>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>>>>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>>>>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>>>>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>>>>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>>>>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
>>>>
>>>>     EXT4-fs (sda1): I/O error while writing superblock
>>>>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
>>>>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
>>>>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
>>>>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
>>>>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
>>>>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
>>>>     EXT4-fs (sda1): I/O error while writing superblock
>>>>
>>>> This may happen either when mounting the root file system (leading to an
>>>> unable to mount root fs panic), or later (leading to a read-only
>>>> rootfs).
>>>
>>> BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
>>> caused by commit 900e080752025f00, and the following patch can fix it:
>>>
>>> - blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
>>>
>>> https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28
>>>
>>> Maybe you can try the above patch.
>>
>> Thanks! I have applied both patches, but it doesn't make a difference.
> 
> Thanks for your test!
> 
> Can you try the following patch?
> 
> 
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index f511db395c7f..a5ab2f2e9f67 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -2517,7 +2517,6 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
>  	struct blk_mq_alloc_data data = {
>  		.q		= q,
>  		.nr_tags	= 1,
> -		.cmd_flags	= bio->bi_opf,
>  	};
>  	struct request *rq;
>  
> @@ -2525,6 +2524,7 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
>  		return NULL;
>  	if (unlikely(!submit_bio_checks(bio)))
>  		goto put_exit;
> +	data.cmd_flags	= bio->bi_opf;
>  	if (blk_mq_attempt_bio_merge(q, bio, nsegs, same_queue_rq))
>  		goto put_exit;

That's definitely a real fix, akin to the other pre-enter variants, this
one just post checks. Geert, can you give this a whirl?

Ming, would you mind sending this as a real patch?

-- 
Jens Axboe

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Ming Lei]

On Thu, Nov 11, 2021 at 03:17:27PM -0700, Jens Axboe wrote:
> On 11/11/21 8:23 AM, Ming Lei wrote:
> > Hi Geert,
> > 
> > On Thu, Nov 11, 2021 at 03:51:28PM +0100, Geert Uytterhoeven wrote:
> >> Hi Ming,
> >>
> >> On Thu, Nov 11, 2021 at 2:45 PM Ming Lei <ming.lei@redhat.com> wrote:
> >>> On Thu, Nov 11, 2021 at 01:58:38PM +0100, Geert Uytterhoeven wrote:
> >>>> On Thu, 4 Nov 2021, Jens Axboe wrote:
> >>>>> On 11/4/21 1:04 PM, Christoph Hellwig wrote:
> >>>>>> On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
> >>>>>>> On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> >>>>>>>> Looks good:
> >>>>>>>>
> >>>>>>>> Reviewed-by: Christoph Hellwig <hch@lst.de>
> >>>>>>>
> >>>>>>> So these two are now:
> >>>>>>>
> >>>>>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
> >>>>>>>
> >>>>>>> which is the one I sent here, and then the next one gets cleaned up to
> >>>>>>> remove that queue enter helper:
> >>>>>>>
> >>>>>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
> >>>>>>>
> >>>>>>> Can I add your reviewed-by to this last one as well? Only change is the
> >>>>>>> removal of blk_mq_enter_queue() and the weird construct there, it's just
> >>>>>>> bio_queue_enter() now.
> >>>>>>
> >>>>>> Sure.
> >>>>>
> >>>>> Thanks, prematurely already done, as you could tell :-)
> >>>>
> >>>> The updated version is now commit 900e080752025f00 ("block: move queue
> >>>> enter logic into blk_mq_submit_bio()") in Linus' tree.
> >>>>
> >>>> I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
> >>>> root device) to this commit, e.g.:
> >>>>
> >>>>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> >>>>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> >>>>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> >>>>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> >>>>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> >>>>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> >>>>
> >>>>     EXT4-fs (sda1): I/O error while writing superblock
> >>>>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> >>>>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> >>>>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> >>>>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> >>>>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> >>>>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> >>>>     EXT4-fs (sda1): I/O error while writing superblock
> >>>>
> >>>> This may happen either when mounting the root file system (leading to an
> >>>> unable to mount root fs panic), or later (leading to a read-only
> >>>> rootfs).
> >>>
> >>> BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
> >>> caused by commit 900e080752025f00, and the following patch can fix it:
> >>>
> >>> - blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
> >>>
> >>> https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28
> >>>
> >>> Maybe you can try the above patch.
> >>
> >> Thanks! I have applied both patches, but it doesn't make a difference.
> > 
> > Thanks for your test!
> > 
> > Can you try the following patch?
> > 
> > 
> > diff --git a/block/blk-mq.c b/block/blk-mq.c
> > index f511db395c7f..a5ab2f2e9f67 100644
> > --- a/block/blk-mq.c
> > +++ b/block/blk-mq.c
> > @@ -2517,7 +2517,6 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
> >  	struct blk_mq_alloc_data data = {
> >  		.q		= q,
> >  		.nr_tags	= 1,
> > -		.cmd_flags	= bio->bi_opf,
> >  	};
> >  	struct request *rq;
> >  
> > @@ -2525,6 +2524,7 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
> >  		return NULL;
> >  	if (unlikely(!submit_bio_checks(bio)))
> >  		goto put_exit;
> > +	data.cmd_flags	= bio->bi_opf;
> >  	if (blk_mq_attempt_bio_merge(q, bio, nsegs, same_queue_rq))
> >  		goto put_exit;
> 
> That's definitely a real fix, akin to the other pre-enter variants, this
> one just post checks. Geert, can you give this a whirl?
> 
> Ming, would you mind sending this as a real patch?

Hi Jens,

The above patch may not be enough, since submit_bio_checks() is done in
case of using cached request, so how about the following patch(un-tested)?


diff --git a/block/blk-mq.c b/block/blk-mq.c
index f511db395c7f..f84044c8de3f 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2517,7 +2517,6 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 	struct blk_mq_alloc_data data = {
 		.q		= q,
 		.nr_tags	= 1,
-		.cmd_flags	= bio->bi_opf,
 	};
 	struct request *rq;
 
@@ -2525,6 +2524,7 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 		return NULL;
 	if (unlikely(!submit_bio_checks(bio)))
 		goto put_exit;
+	data.cmd_flags	= bio->bi_opf;
 	if (blk_mq_attempt_bio_merge(q, bio, nsegs, same_queue_rq))
 		goto put_exit;
 
@@ -2564,13 +2564,15 @@ static inline struct request *blk_mq_get_request(struct request_queue *q,
 			if (blk_mq_attempt_bio_merge(q, bio, nsegs,
 						same_queue_rq))
 				return NULL;
+			if (bio->bi_opf != rq->cmd_flags)
+				goto fallback;
 			plug->cached_rq = rq_list_next(rq);
 			INIT_LIST_HEAD(&rq->queuelist);
 			rq_qos_throttle(q, bio);
 			return rq;
 		}
 	}
-
+fallback:
 	return blk_mq_get_new_requests(q, plug, bio, nsegs, same_queue_rq);
 }
 

-- 
Ming

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Geert Uytterhoeven]

Hi Ming,

On Fri, Nov 12, 2021 at 1:44 AM Ming Lei <ming.lei@redhat.com> wrote:
> On Thu, Nov 11, 2021 at 03:17:27PM -0700, Jens Axboe wrote:
> > On 11/11/21 8:23 AM, Ming Lei wrote:
> > > On Thu, Nov 11, 2021 at 03:51:28PM +0100, Geert Uytterhoeven wrote:
> > >> On Thu, Nov 11, 2021 at 2:45 PM Ming Lei <ming.lei@redhat.com> wrote:
> > >>> On Thu, Nov 11, 2021 at 01:58:38PM +0100, Geert Uytterhoeven wrote:
> > >>>> On Thu, 4 Nov 2021, Jens Axboe wrote:
> > >>>>> On 11/4/21 1:04 PM, Christoph Hellwig wrote:
> > >>>>>> On Thu, Nov 04, 2021 at 01:02:54PM -0600, Jens Axboe wrote:
> > >>>>>>> On 11/4/21 12:52 PM, Christoph Hellwig wrote:
> > >>>>>>>> Looks good:
> > >>>>>>>>
> > >>>>>>>> Reviewed-by: Christoph Hellwig <hch@lst.de>
> > >>>>>>>
> > >>>>>>> So these two are now:
> > >>>>>>>
> > >>>>>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=c98cb5bbdab10d187aff9b4e386210eb2332af96
> > >>>>>>>
> > >>>>>>> which is the one I sent here, and then the next one gets cleaned up to
> > >>>>>>> remove that queue enter helper:
> > >>>>>>>
> > >>>>>>> https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.16/block&id=7f930eb31eeb07f1b606b3316d8ad3ab6a92905b
> > >>>>>>>
> > >>>>>>> Can I add your reviewed-by to this last one as well? Only change is the
> > >>>>>>> removal of blk_mq_enter_queue() and the weird construct there, it's just
> > >>>>>>> bio_queue_enter() now.
> > >>>>>>
> > >>>>>> Sure.
> > >>>>>
> > >>>>> Thanks, prematurely already done, as you could tell :-)
> > >>>>
> > >>>> The updated version is now commit 900e080752025f00 ("block: move queue
> > >>>> enter logic into blk_mq_submit_bio()") in Linus' tree.
> > >>>>
> > >>>> I have bisected failures on m68k/atari (on ARAnyM, using nfhd as the
> > >>>> root device) to this commit, e.g.:
> > >>>>
> > >>>>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> > >>>>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> > >>>>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> > >>>>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> > >>>>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> > >>>>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> > >>>>
> > >>>>     EXT4-fs (sda1): I/O error while writing superblock
> > >>>>     sd 0:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
> > >>>>     sd 0:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current]
> > >>>>     sd 0:0:0:0: [sda] tag#0 Add. Sense: Invalid field in cdb
> > >>>>     sd 0:0:0:0: [sda] tag#0 CDB: Write(10) 2a 08 00 00 00 01 00 00 08 00
> > >>>>     critical target error, dev sda, sector 1 op 0x1:(WRITE) flags 0x20800 phys_seg 1 prio class 0
> > >>>>     Buffer I/O error on dev sda1, logical block 0, lost sync page write
> > >>>>     EXT4-fs (sda1): I/O error while writing superblock
> > >>>>
> > >>>> This may happen either when mounting the root file system (leading to an
> > >>>> unable to mount root fs panic), or later (leading to a read-only
> > >>>> rootfs).
> > >>>
> > >>> BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
> > >>> caused by commit 900e080752025f00, and the following patch can fix it:
> > >>>
> > >>> - blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
> > >>>
> > >>> https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28
> > >>>
> > >>> Maybe you can try the above patch.
> > >>
> > >> Thanks! I have applied both patches, but it doesn't make a difference.
> > >
> > > Thanks for your test!
> > >
> > > Can you try the following patch?

[...]

> > That's definitely a real fix, akin to the other pre-enter variants, this
> > one just post checks. Geert, can you give this a whirl?

With both of

    blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
    blk-mq: rename blk_attempt_bio_merge

applied, and the version above, I no longer saw the error, but the
boot sometimes hangs after:

    ext3 filesystem being remounted at / supports timestamps until
2038 (0x7fffffff)

I don't know how easy that is to trigger: it hung on my first try, but
the second and third tries it booted fully into old Debian userspace.

> > Ming, would you mind sending this as a real patch?
>
> The above patch may not be enough, since submit_bio_checks() is done in
> case of using cached request, so how about the following patch(un-tested)?

Worked fine in five subsequent boots. Thanks!
Tested-by: Geert Uytterhoeven <geert@linux-m68k.org>

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Michael Schmitz]

On 12/11/21 20:51, Geert Uytterhoeven wrote:
[...]
>>>>>> BTW, today I just found that hang in blk_mq_freeze_queue_wait() is
>>>>>> caused by commit 900e080752025f00, and the following patch can fix it:
>>>>>>
>>>>>> - blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
>>>>>>
>>>>>> https://lore.kernel.org/linux-block/20211111085650.GA476@lst.de/T/#m759b88fda094a65ebf29bc81b780967cdaf9cf28
>>>>>>
>>>>>> Maybe you can try the above patch.
>>>>>
>>>>> Thanks! I have applied both patches, but it doesn't make a difference.
>>>>
>>>> Thanks for your test!
>>>>
>>>> Can you try the following patch?
>
> [...]
>
>>> That's definitely a real fix, akin to the other pre-enter variants, this
>>> one just post checks. Geert, can you give this a whirl?
>
> With both of
>
>     blk-mq: don't grab ->q_usage_counter in blk_mq_sched_bio_merge
>     blk-mq: rename blk_attempt_bio_merge
>
> applied, and the version above, I no longer saw the error, but the
> boot sometimes hangs after:
>
>     ext3 filesystem being remounted at / supports timestamps until
> 2038 (0x7fffffff)
>
> I don't know how easy that is to trigger: it hung on my first try, but
> the second and third tries it booted fully into old Debian userspace.
>
>>> Ming, would you mind sending this as a real patch?
>>
>> The above patch may not be enough, since submit_bio_checks() is done in
>> case of using cached request, so how about the following patch(un-tested)?
>
> Worked fine in five subsequent boots. Thanks!
> Tested-by: Geert Uytterhoeven <geert@linux-m68k.org>

For good measure: block-5.16-2021-11-13 tested fine running IO stress 
tests on a mix of IDE and SCSI disks, only one of those supporting DPO/FUA.

Using

WARN_ON(rq->cmd_flags & REQ_FUA && !sdkp->DPOFUA);

in sd.c:sd_setup_read_write_cmnd(), nothing seen in the logs.

Cheers,

	Michael


>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds
>

[PATCH 5/5] block: ensure cached plug request matches the current queue

[Jens Axboe]

If we're driving multiple devices, we could have pre-populated the cache
for a different device. Ensure that the empty request matches the current
queue.

Fixes: 47c122e35d7e ("block: pre-allocate requests if plug is started and is a batch")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-mq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 875bd0c04409..6c8d02bd1b06 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2529,7 +2529,7 @@ static inline struct request *blk_mq_get_request(struct request_queue *q,
 		struct request *rq;
 
 		rq = rq_list_peek(&plug->cached_rq);
-		if (rq) {
+		if (rq && rq->q == q) {
 			if (unlikely(!submit_bio_checks(bio)))
 				return NULL;
 			plug->cached_rq = rq_list_next(rq);
-- 
2.33.1

Re: [PATCH 5/5] block: ensure cached plug request matches the current queue

[Christoph Hellwig]

On Thu, Nov 04, 2021 at 12:22:01PM -0600, Jens Axboe wrote:
> If we're driving multiple devices, we could have pre-populated the cache
> for a different device. Ensure that the empty request matches the current
> queue.
> 
> Fixes: 47c122e35d7e ("block: pre-allocate requests if plug is started and is a batch")
> Signed-off-by: Jens Axboe <axboe@kernel.dk>

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCHSET v2 0/5] Alloc batch fixes

[Jens Axboe]

Hi,

A few fixes related to the batched allocations:

- Have the requests hold a queue reference, flush them on schedule
  unplug as well.

- Make sure the queue matches, could be a mismatch if we're driving
  multiple devices.

Since v1:

- Reshuffle series to do plug rq alloc helper before enter changes
- Protect submit_bio_checks() by queue enter reference as well

-- 
Jens Axboe

[PATCH 4/5] block: move queue enter logic into blk_mq_submit_bio()

[Jens Axboe]

Retain the old logic for the fops based submit, but for our internal
blk_mq_submit_bio(), move the queue entering logic into the core
function itself.

We need to be a bit careful if going into the scheduler, as a scheduler
or queue mappings can arbitrarily change before we have entered the queue.
Have the bio scheduler mapping do that separately, it's a very cheap
operation compared to actually doing merging locking and lookups.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-core.c     | 17 ++++++---------
 block/blk-mq-sched.c | 13 ++++++++---
 block/blk-mq.c       | 51 +++++++++++++++++++++++++++++---------------
 block/blk.h          |  1 +
 4 files changed, 52 insertions(+), 30 deletions(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index e00f5a2287cc..18aab7f8469a 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -746,7 +746,7 @@ static inline blk_status_t blk_check_zone_append(struct request_queue *q,
 	return BLK_STS_OK;
 }
 
-static noinline_for_stack bool submit_bio_checks(struct bio *bio)
+noinline_for_stack bool submit_bio_checks(struct bio *bio)
 {
 	struct block_device *bdev = bio->bi_bdev;
 	struct request_queue *q = bdev_get_queue(bdev);
@@ -868,18 +868,15 @@ static void __submit_bio(struct bio *bio)
 {
 	struct gendisk *disk = bio->bi_bdev->bd_disk;
 
-	if (unlikely(bio_queue_enter(bio) != 0))
-		return;
-
-	if (!submit_bio_checks(bio) || !blk_crypto_bio_prep(&bio))
-		goto queue_exit;
 	if (!disk->fops->submit_bio) {
 		blk_mq_submit_bio(bio);
-		return;
+	} else {
+		if (unlikely(bio_queue_enter(bio) != 0))
+			return;
+		if (submit_bio_checks(bio) && blk_crypto_bio_prep(&bio))
+			disk->fops->submit_bio(bio);
+		blk_queue_exit(disk->queue);
 	}
-	disk->fops->submit_bio(bio);
-queue_exit:
-	blk_queue_exit(disk->queue);
 }
 
 /*
diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c
index 4a6789e4398b..4be652fa38e7 100644
--- a/block/blk-mq-sched.c
+++ b/block/blk-mq-sched.c
@@ -370,15 +370,20 @@ bool blk_mq_sched_bio_merge(struct request_queue *q, struct bio *bio,
 	bool ret = false;
 	enum hctx_type type;
 
-	if (e && e->type->ops.bio_merge)
-		return e->type->ops.bio_merge(q, bio, nr_segs);
+	if (bio_queue_enter(bio))
+		return false;
+
+	if (e && e->type->ops.bio_merge) {
+		ret = e->type->ops.bio_merge(q, bio, nr_segs);
+		goto out_put;
+	}
 
 	ctx = blk_mq_get_ctx(q);
 	hctx = blk_mq_map_queue(q, bio->bi_opf, ctx);
 	type = hctx->type;
 	if (!(hctx->flags & BLK_MQ_F_SHOULD_MERGE) ||
 	    list_empty_careful(&ctx->rq_lists[type]))
-		return false;
+		goto out_put;
 
 	/* default per sw-queue merge */
 	spin_lock(&ctx->lock);
@@ -391,6 +396,8 @@ bool blk_mq_sched_bio_merge(struct request_queue *q, struct bio *bio,
 		ret = true;
 
 	spin_unlock(&ctx->lock);
+out_put:
+	blk_queue_exit(q);
 	return ret;
 }
 
diff --git a/block/blk-mq.c b/block/blk-mq.c
index f7f36d5ed25a..b0c0eac43eef 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2487,12 +2487,21 @@ static inline struct request *blk_get_plug_request(struct request_queue *q,
 	if (!plug)
 		return NULL;
 	rq = rq_list_peek(&plug->cached_rq);
-	if (rq) {
-		plug->cached_rq = rq_list_next(rq);
-		INIT_LIST_HEAD(&rq->queuelist);
-		return rq;
-	}
-	return NULL;
+	if (!rq)
+		return NULL;
+	if (unlikely(!submit_bio_checks(bio)))
+		return ERR_PTR(-EIO);
+	plug->cached_rq = rq_list_next(rq);
+	INIT_LIST_HEAD(&rq->queuelist);
+	rq_qos_throttle(q, bio);
+	return rq;
+}
+
+static inline bool blk_mq_queue_enter(struct request_queue *q, struct bio *bio)
+{
+	if (!blk_try_enter_queue(q, false) && bio_queue_enter(bio))
+		return false;
+	return true;
 }
 
 /**
@@ -2518,31 +2527,41 @@ void blk_mq_submit_bio(struct bio *bio)
 	unsigned int nr_segs = 1;
 	blk_status_t ret;
 
+	if (unlikely(!blk_crypto_bio_prep(&bio)))
+		return;
+
 	blk_queue_bounce(q, &bio);
 	if (blk_may_split(q, bio))
 		__blk_queue_split(q, &bio, &nr_segs);
 
 	if (!bio_integrity_prep(bio))
-		goto queue_exit;
+		return;
 
 	if (!blk_queue_nomerges(q) && bio_mergeable(bio)) {
 		if (blk_attempt_plug_merge(q, bio, nr_segs, &same_queue_rq))
-			goto queue_exit;
+			return;
 		if (blk_mq_sched_bio_merge(q, bio, nr_segs))
-			goto queue_exit;
+			return;
 	}
 
-	rq_qos_throttle(q, bio);
-
 	plug = blk_mq_plug(q, bio);
 	rq = blk_get_plug_request(q, plug, bio);
-	if (!rq) {
+	if (IS_ERR(rq)) {
+		return;
+	} else if (!rq) {
 		struct blk_mq_alloc_data data = {
 			.q		= q,
 			.nr_tags	= 1,
 			.cmd_flags	= bio->bi_opf,
 		};
 
+		if (unlikely(!blk_mq_queue_enter(q, bio)))
+			return;
+		if (unlikely(!submit_bio_checks(bio)))
+			goto put_exit;
+
+		rq_qos_throttle(q, bio);
+
 		if (plug) {
 			data.nr_tags = plug->nr_ios;
 			plug->nr_ios = 1;
@@ -2553,7 +2572,9 @@ void blk_mq_submit_bio(struct bio *bio)
 			rq_qos_cleanup(q, bio);
 			if (bio->bi_opf & REQ_NOWAIT)
 				bio_wouldblock_error(bio);
-			goto queue_exit;
+put_exit:
+			blk_queue_exit(q);
+			return;
 		}
 	}
 
@@ -2636,10 +2657,6 @@ void blk_mq_submit_bio(struct bio *bio)
 		/* Default case. */
 		blk_mq_sched_insert_request(rq, false, true, true);
 	}
-
-	return;
-queue_exit:
-	blk_queue_exit(q);
 }
 
 static size_t order_to_size(unsigned int order)
diff --git a/block/blk.h b/block/blk.h
index f7371d3b1522..79c98ced59c8 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -56,6 +56,7 @@ void blk_freeze_queue(struct request_queue *q);
 void __blk_mq_unfreeze_queue(struct request_queue *q, bool force_atomic);
 void blk_queue_start_drain(struct request_queue *q);
 int bio_queue_enter(struct bio *bio);
+bool submit_bio_checks(struct bio *bio);
 
 static inline bool blk_try_enter_queue(struct request_queue *q, bool pm)
 {
-- 
2.33.1