Improve zoned storage write performance
Subsystems (2)
v2 cover Improve zoned storage write performance
Hi Jens,
Measurements have shown that limiting the queue depth to one per sequential zone has a significant negative performance impact on zoned UFS devices. Hence this patch series that increases the queue depth for write commands for sequential zones when using the mq-deadline scheduler.
SCSI and UFS patches will be submitted at a later time (after multiqueue support is available for the UFS driver).
Please consider this patch series for kernel v5.20.
Thanks,
Bart.
Changes compared to v1:
- Left out the SCSI and NVMe patches.
- Added a null_blk patch.
- Included measurement results.
Bart Van Assche (6):
block: Document blk_queue_zone_is_seq() and blk_rq_zone_is_seq()
block: Introduce the blk_rq_is_zoned_seq_write() function
block: Introduce a request queue flag for pipelining zoned writes block/mq-deadline: Only use zone locking if necessary block/null_blk: Refactor null_queue_rq() block/null_blk: Add support for pipelining zoned writes
block/blk-zoned.c | 17 +++-------------- block/mq-deadline.c | 15 +++++++++------ drivers/block/null_blk/main.c | 30 ++++++++++++++++++++---------- drivers/block/null_blk/null_blk.h | 3 +++ drivers/block/null_blk/zoned.c | 4 +++- include/linux/blk-mq.h | 30 ++++++++++++++++++++++++++++++ include/linux/blkdev.h | 16 ++++++++++++++++ 7 files changed, 84 insertions(+), 31 deletions(-)
PATCH v2 1/6 block: Document blk_queue_zone_is_seq() and blk_rq_zone_is_seq()
Since it is nontrivial to figure out how blk_queue_zone_is_seq() and blk_rq_zone_is_seq() handle sequential write preferred zones, document this.
--- include/linux/blk-mq.h | 7 +++++++ include/linux/blkdev.h | 9 +++++++++ 2 files changed, 16 insertions(+)
include/linux/blk-mq.h +7 -0
@@ -1124,6 +1124,13 @@ static inline unsigned int blk_rq_zone_no(struct request *rq)return blk_queue_zone_no (rq->q,blk_rq_pos (rq)); } +/** +* blk_rq_zone_is_seq() - Whether a request is for a sequential zone. +* @rq: Request pointer. +* +* Return: true if and only if blk_rq_pos(@rq) refers either to a sequential +* write required or a sequential write preferred zone. +*/ static inline unsigned int blk_rq_zone_is_seq (struct request *rq) {return blk_queue_zone_is_seq (rq->q,blk_rq_pos (rq));
include/linux/blkdev.h +9 -0
@@ -688,6 +688,15 @@ static inline unsigned int blk_queue_zone_no(struct request_queue *q,return sector >>ilog2 (q->limits.chunk_sectors); } +/** +* blk_queue_zone_is_seq() - Whether a logical block is in a sequential zone. +* @q: Request queue pointer. +* @sector: Offset from start of block device in 512 byte units. +* +* Return: true if and only if @q is associated with a zoned block device and +* @sector refers either to a sequential write required or a sequential write +* preferred zone. +*/ static inline bool blk_queue_zone_is_seq (struct request_queue *q,sector_t sector) {
Re: [PATCH v2 1/6] block: Document blk_queue_zone_is_seq() and blk_rq_zone_is_seq()
Quoted 6 lines · Bart Van Assche
Since it is nontrivial to figure out how
blk_queue_zone_is_seq()andblk_rq_zone_is_seq()handle sequential write preferred zones, document this.Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>
Looks good.
Quoted 43 lines
--- include/linux/blk-mq.h | 7 +++++++ include/linux/blkdev.h | 9 +++++++++ 2 files changed, 16 insertions(+)include/linux/blk-mq.h +7
-0@@ -1124,6 +1124,13 @@ static inline unsigned int blk_rq_zone_no(struct request *rq)return blk_queue_zone_no (rq->q,blk_rq_pos (rq)); } +/** +* blk_rq_zone_is_seq() - Whether a request is for a sequential zone. +* @rq: Request pointer. +* +* Return: true if and only if blk_rq_pos(@rq) refers either to a sequential +* write required or a sequential write preferred zone. +*/ static inline unsigned int blk_rq_zone_is_seq (struct request *rq) {return blk_queue_zone_is_seq (rq->q,blk_rq_pos (rq));include/linux/blkdev.h +9
-0@@ -688,6 +688,15 @@ static inline unsigned int blk_queue_zone_no(struct request_queue *q,return sector >>ilog2 (q->limits.chunk_sectors); } +/** +* blk_queue_zone_is_seq() - Whether a logical block is in a sequential zone. +* @q: Request queue pointer. +* @sector: Offset from start of block device in 512 byte units. +* +* Return: true if and only if @q is associated with a zoned block device and +* @sector refers either to a sequential write required or a sequential write +* preferred zone. +*/ static inline bool blk_queue_zone_is_seq (struct request_queue *q,sector_t sector) {
Signature · 2 lines
Damien Le Moal Western Digital Research
PATCH v2 2/6 block: Introduce the blk_rq_is_zoned_seq_write() function
In reply to Bart Van AsscheIntroduce a function that makes it easy to verify whether a write request is for a sequential write required or sequential write preferred zone. Simplify blk_req_needs_zone_write_lock() by using the new function.
--- block/blk-zoned.c | 14 +------------- include/linux/blk-mq.h | 23 +++++++++++++++++++++++ 2 files changed, 24 insertions(+), 13 deletions(-)
block/blk-zoned.c +1 -13
@@ -57,19 +57,7 @@ EXPORT_SYMBOL_GPL(blk_zone_cond_str);*/ bool blk_req_needs_zone_write_lock (struct request *rq) { -if (!rq->q->seq_zones_wlock) -return false; - -if (blk_rq_is_passthrough (rq)) -return false; - -switch (req_op (rq)) { -case REQ_OP_WRITE_ZEROES : -case REQ_OP_WRITE : -return blk_rq_zone_is_seq (rq); -default : -return false; - } +return rq->q->seq_zones_wlock &&blk_rq_is_zoned_seq_write (rq); }EXPORT_SYMBOL_GPL (blk_req_needs_zone_write_lock);
include/linux/blk-mq.h +23 -0
@@ -1136,6 +1136,24 @@ static inline unsigned int blk_rq_zone_is_seq(struct request *rq)return blk_queue_zone_is_seq (rq->q,blk_rq_pos (rq)); } +/** +* blk_rq_is_zoned_seq_write() - Whether @rq is a write request for a sequential zone. +* @rq: Request to examine. +* +* In this context sequential zone means either a sequential write required or +* a sequential write preferred zone. +*/ +static inline bool blk_rq_is_zoned_seq_write (struct request *rq) +{ +switch (req_op (rq)) { +case REQ_OP_WRITE : +case REQ_OP_WRITE_ZEROES : +return blk_rq_zone_is_seq (rq); +default : +return false; + } +} +bool blk_req_needs_zone_write_lock (struct request *rq);bool blk_req_zone_write_trylock (struct request *rq);void __blk_req_zone_write_lock (struct request *rq);
@@ -1166,6 +1184,11 @@ static inline bool blk_req_can_dispatch_to_zone(struct request *rq) return !blk_req_zone_is_write_locked(rq); }#else /* CONFIG_BLK_DEV_ZONED */ +static inline bool blk_rq_is_zoned_seq_write (struct request *rq) +{ +return false; +} +static inline bool blk_req_needs_zone_write_lock (struct request *rq) {return false;
Re: [PATCH v2 2/6] block: Introduce the blk_rq_is_zoned_seq_write() function
Quoted 6 lines · Bart Van Assche
Introduce a function that makes it easy to verify whether a write request is for a sequential write required or sequential write preferred zone. Simplify
blk_req_needs_zone_write_lock()by using the new function.Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>
Looks good.
Quoted 71 lines
--- block/blk-zoned.c | 14 +------------- include/linux/blk-mq.h | 23 +++++++++++++++++++++++ 2 files changed, 24 insertions(+), 13 deletions(-)block/blk-zoned.c +1
-13@@ -57,19 +57,7 @@ EXPORT_SYMBOL_GPL(blk_zone_cond_str);*/ bool blk_req_needs_zone_write_lock (struct request *rq) { -if (!rq->q->seq_zones_wlock) -return false; - -if (blk_rq_is_passthrough (rq)) -return false; - -switch (req_op (rq)) { -case REQ_OP_WRITE_ZEROES : -case REQ_OP_WRITE : -return blk_rq_zone_is_seq (rq); -default : -return false; - } +return rq->q->seq_zones_wlock &&blk_rq_is_zoned_seq_write (rq); }EXPORT_SYMBOL_GPL (blk_req_needs_zone_write_lock);include/linux/blk-mq.h +23
-0@@ -1136,6 +1136,24 @@ static inline unsigned int blk_rq_zone_is_seq(struct request *rq)return blk_queue_zone_is_seq (rq->q,blk_rq_pos (rq)); } +/** +* blk_rq_is_zoned_seq_write() - Whether @rq is a write request for a sequential zone. +* @rq: Request to examine. +* +* In this context sequential zone means either a sequential write required or +* a sequential write preferred zone. +*/ +static inline bool blk_rq_is_zoned_seq_write (struct request *rq) +{ +switch (req_op (rq)) { +case REQ_OP_WRITE : +case REQ_OP_WRITE_ZEROES : +return blk_rq_zone_is_seq (rq); +default : +return false; + } +} +bool blk_req_needs_zone_write_lock (struct request *rq);bool blk_req_zone_write_trylock (struct request *rq);void __blk_req_zone_write_lock (struct request *rq);@@ -1166,6 +1184,11 @@ static inline bool blk_req_can_dispatch_to_zone(struct request *rq) return !blk_req_zone_is_write_locked(rq); }#else /* CONFIG_BLK_DEV_ZONED */ +static inline bool blk_rq_is_zoned_seq_write (struct request *rq) +{ +return false; +} +static inline bool blk_req_needs_zone_write_lock (struct request *rq) {return false;
Signature · 2 lines
Damien Le Moal Western Digital Research
Re: [PATCH v2 2/6] block: Introduce the blk_rq_is_zoned_seq_write() function
In reply to Bart Van AsscheQuoted 41 lines · Bart Van Assche
Introduce a function that makes it easy to verify whether a write request is for a sequential write required or sequential write preferred zone. Simplify
blk_req_needs_zone_write_lock()by using the new function.Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>--- block/blk-zoned.c | 14 +------------- include/linux/blk-mq.h | 23 +++++++++++++++++++++++ 2 files changed, 24 insertions(+), 13 deletions(-)block/blk-zoned.c +1
-13@@ -57,19 +57,7 @@ EXPORT_SYMBOL_GPL(blk_zone_cond_str);*/ bool blk_req_needs_zone_write_lock (struct request *rq) { -if (!rq->q->seq_zones_wlock) -return false; - -if (blk_rq_is_passthrough (rq)) -return false; - -switch (req_op (rq)) { -case REQ_OP_WRITE_ZEROES : -case REQ_OP_WRITE : -return blk_rq_zone_is_seq (rq); -default : -return false; - } +return rq->q->seq_zones_wlock &&blk_rq_is_zoned_seq_write (rq); }EXPORT_SYMBOL_GPL (blk_req_needs_zone_write_lock);include/linux/blk-mq.h +0
-0
include/linux/blk-mq.h static inline unsigned int blk_rq_zone_is_seq(struct request *rq) · line 1140
Quoted 5 lines
include/linux/blk-mq.h +2
-0@@ -1136,6 +1136,24 @@ static inline unsigned int blk_rq_zone_is_seq(struct request *rq)return blk_queue_zone_is_seq (rq->q,blk_rq_pos (rq)); } +/** +* blk_rq_is_zoned_seq_write() - Whether @rq is a write request for a sequential zone.
This doesn't parse and is overly long at the same time.
Re: [PATCH v2 2/6] block: Introduce the blk_rq_is_zoned_seq_write() function
Quoted 5 lines · Christoph Hellwig
Quoted 2 lines · Bart Van Assche
+/**
+ *blk_rq_is_zoned_seq_write()- Whether @rq is a write request for a sequential zone.This doesn't parse and is overly long at the same time.
Hi Christoph,
It is not clear to me why you wrote that "this doesn't parse"? Are you referring to the function name or to the function description? I think the text at the right side of the hyphen is grammatically correct.
Although I'm in favor of respecting the 80 column limit, I haven't found any explanation in Documentation/doc-guide/kernel-doc.rst about whether or not splitting the brief description across multiple lines is supported.
Thanks,
Bart.
PATCH v2 3/6 block: Introduce a request queue flag for pipelining zoned writes
In reply to Bart Van AsscheWrites in sequential write required zones must happen at the write pointer. Even if the submitter of the write commands (e.g. a filesystem) submits writes for sequential write required zones in order, the block layer or the storage controller may reorder these write commands.
The zone locking mechanism in the mq-deadline I/O scheduler serializes write commands for sequential zones. Some but not all storage controllers require this serialization. Introduce a new flag such that block drivers can request pipelining of writes for sequential write required zones.
An example of a storage controller standard that requires write serialization is AHCI (Advanced Host Controller Interface). Submitting commands to AHCI controllers happens by writing a bit pattern into a register. Each set bit corresponds to an active command. This mechanism does not preserve command ordering information.
--- include/linux/blkdev.h | 7 +++++++ 1 file changed, 7 insertions(+)
include/linux/blkdev.h +7 -0
@@ -581,6 +581,8 @@ struct request_queue {#define QUEUE_FLAG_HCTX_ACTIVE 28/* at least one blk-mq hctx is active */ #define QUEUE_FLAG_NOWAIT 29/* device supports NOWAIT */ #define QUEUE_FLAG_SQ_SCHED 30/* single queue style io dispatch */ +/* Writes for sequential write required zones may be pipelined. */ +#define QUEUE_FLAG_PIPELINE_ZONED_WRITES 31#define QUEUE_FLAG_MQ_DEFAULT ((1 << QUEUE_FLAG_IO_STAT) | \ (1 << QUEUE_FLAG_SAME_COMP) | \
@@ -624,6 +626,11 @@ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q);#define blk_queue_nowait (q) test_bit(QUEUE_FLAG_NOWAIT, &(q)->queue_flags)#define blk_queue_sq_sched (q) test_bit(QUEUE_FLAG_SQ_SCHED, &(q)->queue_flags) +static inline bool blk_queue_pipeline_zoned_writes (struct request_queue *q) +{ +return test_bit (QUEUE_FLAG_PIPELINE_ZONED_WRITES , &(q)->queue_flags); +} +extern void blk_set_pm_only (struct request_queue *q);extern void blk_clear_pm_only (struct request_queue *q);
Re: [PATCH v2 3/6] block: Introduce a request queue flag for pipelining zoned writes
Quoted 36 lines · Bart Van Assche
Writes in sequential write required zones must happen at the write pointer. Even if the submitter of the write commands (e.g. a filesystem) submits writes for sequential write required zones in order, the block layer or the storage controller may reorder these write commands.
The zone locking mechanism in the mq-deadline I/O scheduler serializes write commands for sequential zones. Some but not all storage controllers require this serialization. Introduce a new flag such that block drivers can request pipelining of writes for sequential write required zones.
An example of a storage controller standard that requires write serialization is AHCI (Advanced Host Controller Interface). Submitting commands to AHCI controllers happens by writing a bit pattern into a register. Each set bit corresponds to an active command. This mechanism does not preserve command ordering information.
Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>--- include/linux/blkdev.h | 7 +++++++ 1 file changed, 7 insertions(+)include/linux/blkdev.h +2
-0@@ -581,6 +581,8 @@ struct request_queue {#define QUEUE_FLAG_HCTX_ACTIVE 28/* at least one blk-mq hctx is active */ #define QUEUE_FLAG_NOWAIT 29/* device supports NOWAIT */ #define QUEUE_FLAG_SQ_SCHED 30/* single queue style io dispatch */ +/* Writes for sequential write required zones may be pipelined. */ +#define QUEUE_FLAG_PIPELINE_ZONED_WRITES 31#define QUEUE_FLAG_MQ_DEFAULT ((1 << QUEUE_FLAG_IO_STAT) | \ (1 << QUEUE_FLAG_SAME_COMP) | \
include/linux/blkdev.h bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q); · line 633
Quoted 8 lines
include/linux/blkdev.h +5
-0@@ -624,6 +626,11 @@ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q);#define blk_queue_nowait (q) test_bit(QUEUE_FLAG_NOWAIT, &(q)->queue_flags)#define blk_queue_sq_sched (q) test_bit(QUEUE_FLAG_SQ_SCHED, &(q)->queue_flags) +static inline bool blk_queue_pipeline_zoned_writes (struct request_queue *q) +{ +return test_bit (QUEUE_FLAG_PIPELINE_ZONED_WRITES , &(q)->queue_flags); +} +
Since this flag will be set by an LLD to indicate that the LLD can handle zoned write commands in order, I would suggest a different name. Something
like: QUEUE_FLAG_ORDERED_ZONED_WRITES ? And well, if the LLD says it can do that for zoned writes, it likely means that it would be the same for any command, so the flag could be generalized and named QUEUE_FLAG_ORDERED_CMD or something like that.
Quoted 3 lines
include/linux/blkdev.h +0
-0extern void blk_set_pm_only (struct request_queue *q);extern void blk_clear_pm_only (struct request_queue *q);
Signature · 2 lines
Damien Le Moal Western Digital Research
Re: [PATCH v2 3/6] block: Introduce a request queue flag for pipelining zoned writes
Quoted 13 lines · Damien Le Moal
include/linux/blkdev.h bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q); · line 633
Quoted 5 lines · Bart Van Assche
include/linux/blkdev.h +5
-0+static inline bool blk_queue_pipeline_zoned_writes (struct request_queue *q) +{ +return test_bit (QUEUE_FLAG_PIPELINE_ZONED_WRITES , &(q)->queue_flags); +} +Since this flag will be set by an LLD to indicate that the LLD can handle zoned write commands in order, I would suggest a different name. Something
like:QUEUE_FLAG_ORDERED_ZONED_WRITES? And well, if the LLD says it can do that for zoned writes, it likely means that it would be the same for any command, so the flag could be generalized and namedQUEUE_FLAG_ORDERED_CMDor something like that.
Zoned writes should always be submitted in order by the software that generates the zoned writes so I think the name QUEUE_FLAG_ORDERED_ZONED_WRITES would cause confusion. I'm concerned it would make other kernel developers wonder whether it is OK for e.g. filesystems to submit zoned writes out of order, which is not the case. I think the word "pipelining" has a well-established meaning in computer science and also that in this context it reflects the intent correctly. See also https://en.wikipedia.org/wiki/Pipeline_(computing).
Thanks,
Bart.
PATCH v2 4/6 block/mq-deadline: Only use zone locking if necessary
In reply to Bart Van AsscheMeasurements have shown that limiting the queue depth to one for zoned writes has a significant negative performance impact on zoned UFS devices. Hence this patch that disables zone locking from the mq-deadline scheduler for storage controllers that support pipelining zoned writes. This patch is based on the following assumptions:
- Applications submit write requests to sequential write required zones in order.
- The I/O priority of all pipelined write requests is the same per zone.
- If such write requests get reordered by the software or hardware queue mechanism, nr_hw_queues * nr_requests - 1 retries are sufficient to reorder the write requests.
- It happens infrequently that zoned write requests are reordered by the block layer.
- Either no I/O scheduler is used or an I/O scheduler is used that submits write requests per zone in LBA order.
See also commit 5700f69178e9 ("mq-deadline: Introduce zone locking support").
--- block/blk-zoned.c | 3 ++- block/mq-deadline.c | 15 +++++++++------ 2 files changed, 11 insertions(+), 7 deletions(-)
block/blk-zoned.c +2 -1
@@ -513,7 +513,8 @@ static int blk_revalidate_zone_cb(struct blk_zone *zone, unsigned int idx,break ;case BLK_ZONE_TYPE_SEQWRITE_REQ :case BLK_ZONE_TYPE_SEQWRITE_PREF : -if (!args->seq_zones_wlock) { +if (!blk_queue_pipeline_zoned_writes (q) && + !args->seq_zones_wlock) { args->seq_zones_wlock =blk_alloc_zone_bitmap (q->node, args->nr_zones);if (!args->seq_zones_wlock)
block/mq-deadline.c +9 -6
@@ -292,7 +292,7 @@ deadline_fifo_request(struct deadline_data *dd, struct dd_per_prio *per_prio,return NULL ; rq =rq_entry_fifo (per_prio->fifo_list[data_dir].next); -if (data_dir ==DD_READ || !blk_queue_is_zoned (rq->q)) +if (data_dir ==DD_READ ||blk_queue_pipeline_zoned_writes (rq->q))return rq;/*
@@ -326,7 +326,7 @@ deadline_next_request(struct deadline_data *dd, struct dd_per_prio *per_prio,if (!rq)return NULL ; -if (data_dir ==DD_READ || !blk_queue_is_zoned (rq->q)) +if (data_dir ==DD_READ ||blk_queue_pipeline_zoned_writes (rq->q))return rq;/*
@@ -445,8 +445,9 @@ static struct request *__dd_dispatch_request(struct deadline_data *dd, }/* -* For a zoned block device, if we only have writes queued and none of -* them can be dispatched, rq will be NULL. +* For a zoned block device that requires write serialization, if we +* only have writes queued and none of them can be dispatched, rq will +* be NULL. */ if (!rq)return NULL ;
@@ -719,6 +720,8 @@ static void dd_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,u8 ioprio_class =IOPRIO_PRIO_CLASS (ioprio);struct dd_per_prio *per_prio;enum dd_prio prio; +bool pipelined_seq_write =blk_queue_pipeline_zoned_writes (q) && +blk_rq_is_zoned_seq_write (rq);LIST_HEAD (free);lockdep_assert_held (&dd->lock);
@@ -743,7 +746,7 @@ static void dd_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,trace_block_rq_insert (rq); -if (at_head) { +if (at_head && !pipelined_seq_write) {list_add (&rq->queuelist, &per_prio->dispatch); rq->fifo_time = jiffies; }else {
@@ -823,7 +826,7 @@ static void dd_finish_request(struct request *rq)atomic_inc (&per_prio->stats.completed); -if (blk_queue_is_zoned (q)) { +if (!blk_queue_pipeline_zoned_writes (q)) {unsigned long flags;spin_lock_irqsave (&dd->zone_lock, flags);
Re: [PATCH v2 4/6] block/mq-deadline: Only use zone locking if necessary
Quoted 16 lines · Bart Van Assche
Measurements have shown that limiting the queue depth to one for zoned writes has a significant negative performance impact on zoned UFS devices. Hence this patch that disables zone locking from the mq-deadline scheduler for storage controllers that support pipelining zoned writes. This patch is based on the following assumptions:
- Applications submit write requests to sequential write required zones in order.
- The I/O priority of all pipelined write requests is the same per zone.
- If such write requests get reordered by the software or hardware queue mechanism,nr_hw_queues*nr_requests- 1 retries are sufficient to reorder the write requests.
- It happens infrequently that zoned write requests are reordered by the block layer.
- Either no I/O scheduler is used or an I/O scheduler is used that submits write requests per zone in LBA order.
Quoted 2 lines
See also commit
5700f69178e9("mq-deadline: Introduce zone locking support").
I think this patch should be squashed together with the previous patch. It would then be easier to see what effect the pipeline queue flag has.
Quoted 27 lines
Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>--- block/blk-zoned.c | 3 ++- block/mq-deadline.c | 15 +++++++++------ 2 files changed, 11 insertions(+), 7 deletions(-)block/blk-zoned.c +2
-1@@ -513,7 +513,8 @@ static int blk_revalidate_zone_cb(struct blk_zone *zone, unsigned int idx,break ;case BLK_ZONE_TYPE_SEQWRITE_REQ :case BLK_ZONE_TYPE_SEQWRITE_PREF : -if (!args->seq_zones_wlock) { +if (!blk_queue_pipeline_zoned_writes (q) && + !args->seq_zones_wlock) { args->seq_zones_wlock =blk_alloc_zone_bitmap (q->node, args->nr_zones);if (!args->seq_zones_wlock)block/mq-deadline.c +0
-0
block/mq-deadline.c deadline_fifo_request(struct deadline_data *dd, struct dd_per_prio *per_prio, · line 295
Quoted 5 lines
block/mq-deadline.c +1
-1@@ -292,7 +292,7 @@ deadline_fifo_request(struct deadline_data *dd, struct dd_per_prio *per_prio,return NULL ; rq =rq_entry_fifo (per_prio->fifo_list[data_dir].next); -if (data_dir ==DD_READ || !blk_queue_is_zoned (rq->q)) +if (data_dir ==DD_READ ||blk_queue_pipeline_zoned_writes (rq->q))
This change seems wrong. Before: both read and writes can proceed for regular disks. After, only read can proceed, assuming that the regular device does not have pipeline zoned writes enabled.
block/mq-deadline.c deadline_next_request(struct deadline_data *dd, struct dd_per_prio *per_prio, · line 329
Quoted 9 lines
block/mq-deadline.c +1
-1return rq;/* @@ -326,7 +326,7 @@ deadline_next_request(struct deadline_data *dd, struct dd_per_prio *per_prio,if (!rq)return NULL ; -if (data_dir ==DD_READ || !blk_queue_is_zoned (rq->q)) +if (data_dir ==DD_READ ||blk_queue_pipeline_zoned_writes (rq->q))
same here.
Quoted 42 lines
block/mq-deadline.c +7
-4return rq;/* @@ -445,8 +445,9 @@ static struct request *__dd_dispatch_request(struct deadline_data *dd, }/* -* For a zoned block device, if we only have writes queued and none of -* them can be dispatched, rq will be NULL. +* For a zoned block device that requires write serialization, if we +* only have writes queued and none of them can be dispatched, rq will +* be NULL. */ if (!rq)return NULL ;@@ -719,6 +720,8 @@ static void dd_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,u8 ioprio_class =IOPRIO_PRIO_CLASS (ioprio);struct dd_per_prio *per_prio;enum dd_prio prio; +bool pipelined_seq_write =blk_queue_pipeline_zoned_writes (q) && +blk_rq_is_zoned_seq_write (rq);LIST_HEAD (free);lockdep_assert_held (&dd->lock);@@ -743,7 +746,7 @@ static void dd_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,trace_block_rq_insert (rq); -if (at_head) { +if (at_head && !pipelined_seq_write) {list_add (&rq->queuelist, &per_prio->dispatch); rq->fifo_time = jiffies; }else {@@ -823,7 +826,7 @@ static void dd_finish_request(struct request *rq)atomic_inc (&per_prio->stats.completed); -if (blk_queue_is_zoned (q)) { +if (!blk_queue_pipeline_zoned_writes (q)) {unsigned long flags;spin_lock_irqsave (&dd->zone_lock, flags);
Signature · 2 lines
Damien Le Moal Western Digital Research
PATCH v2 5/6 block/null_blk: Refactor null_queue_rq()
In reply to Bart Van AsscheIntroduce a local variable for the expression bd->rq since that expression occurs multiple times. This patch does not change any functionality.
--- drivers/block/null_blk/main.c | 21 +++++++++++---------- 1 file changed, 11 insertions(+), 10 deletions(-)
drivers/block/null_blk/main.c +11 -10
@@ -1609,10 +1609,11 @@ static enum blk_eh_timer_return null_timeout_rq(struct request *rq, bool res)static blk_status_t null_queue_rq (struct blk_mq_hw_ctx *hctx,const struct blk_mq_queue_data *bd) { -struct nullb_cmd *cmd =blk_mq_rq_to_pdu (bd->rq); +struct request *rq = bd->rq; +struct nullb_cmd *cmd =blk_mq_rq_to_pdu (rq);struct nullb_queue *nq = hctx->driver_data; -sector_t nr_sectors =blk_rq_sectors (bd->rq); -sector_t sector =blk_rq_pos (bd->rq); +sector_t nr_sectors =blk_rq_sectors (rq); +sector_t sector =blk_rq_pos (rq);const bool is_poll = hctx->type ==HCTX_TYPE_POLL ;might_sleep_if (hctx->flags &BLK_MQ_F_BLOCKING );
@@ -1621,14 +1622,14 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx,hrtimer_init (&cmd->timer,CLOCK_MONOTONIC ,HRTIMER_MODE_REL ); cmd->timer.function = null_cmd_timer_expired; } - cmd->rq = bd->rq; + cmd->rq = rq; cmd->error =BLK_STS_OK ; cmd->nq = nq; - cmd->fake_timeout =should_timeout_request (bd->rq); + cmd->fake_timeout =should_timeout_request (rq); -blk_mq_start_request (bd->rq); +blk_mq_start_request (rq); -if (should_requeue_request (bd->rq)) { +if (should_requeue_request (rq)) {/* * Alternate between hitting the core BUSY path, and the * driver driven requeue path
@@ -1637,21 +1638,21 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx,if (nq->requeue_selection &1 )return BLK_STS_RESOURCE ;else { -blk_mq_requeue_request (bd->rq, true); +blk_mq_requeue_request (rq, true);return BLK_STS_OK ; } }if (is_poll) {spin_lock (&nq->poll_lock); -list_add_tail (&bd->rq->queuelist, &nq->poll_list); +list_add_tail (&rq->queuelist, &nq->poll_list);spin_unlock (&nq->poll_lock);return BLK_STS_OK ; }if (cmd->fake_timeout)return BLK_STS_OK ; -return null_handle_cmd (cmd, sector, nr_sectors,req_op (bd->rq)); +return null_handle_cmd (cmd, sector, nr_sectors,req_op (rq)); }static void cleanup_queue (struct nullb_queue *nq)
Re: [PATCH v2 5/6] block/null_blk: Refactor null_queue_rq()
Quoted 5 lines · Bart Van Assche
Introduce a local variable for the expression
bd->rqsince that expression occurs multiple times. This patch does not change any functionality.Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>
Looks good. Completely independent from the series goal though, so this can be applied independently I think.
Quoted 67 lines
--- drivers/block/null_blk/main.c | 21 +++++++++++---------- 1 file changed, 11 insertions(+), 10 deletions(-)drivers/block/null_blk/main.c +11
-10@@ -1609,10 +1609,11 @@ static enum blk_eh_timer_return null_timeout_rq(struct request *rq, bool res)static blk_status_t null_queue_rq (struct blk_mq_hw_ctx *hctx,const struct blk_mq_queue_data *bd) { -struct nullb_cmd *cmd =blk_mq_rq_to_pdu (bd->rq); +struct request *rq = bd->rq; +struct nullb_cmd *cmd =blk_mq_rq_to_pdu (rq);struct nullb_queue *nq = hctx->driver_data; -sector_t nr_sectors =blk_rq_sectors (bd->rq); -sector_t sector =blk_rq_pos (bd->rq); +sector_t nr_sectors =blk_rq_sectors (rq); +sector_t sector =blk_rq_pos (rq);const bool is_poll = hctx->type ==HCTX_TYPE_POLL ;might_sleep_if (hctx->flags &BLK_MQ_F_BLOCKING );@@ -1621,14 +1622,14 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx,hrtimer_init (&cmd->timer,CLOCK_MONOTONIC ,HRTIMER_MODE_REL ); cmd->timer.function = null_cmd_timer_expired; } - cmd->rq = bd->rq; + cmd->rq = rq; cmd->error =BLK_STS_OK ; cmd->nq = nq; - cmd->fake_timeout =should_timeout_request (bd->rq); + cmd->fake_timeout =should_timeout_request (rq); -blk_mq_start_request (bd->rq); +blk_mq_start_request (rq); -if (should_requeue_request (bd->rq)) { +if (should_requeue_request (rq)) {/* * Alternate between hitting the core BUSY path, and the * driver driven requeue path @@ -1637,21 +1638,21 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx,if (nq->requeue_selection &1 )return BLK_STS_RESOURCE ;else { -blk_mq_requeue_request (bd->rq, true); +blk_mq_requeue_request (rq, true);return BLK_STS_OK ; } }if (is_poll) {spin_lock (&nq->poll_lock); -list_add_tail (&bd->rq->queuelist, &nq->poll_list); +list_add_tail (&rq->queuelist, &nq->poll_list);spin_unlock (&nq->poll_lock);return BLK_STS_OK ; }if (cmd->fake_timeout)return BLK_STS_OK ; -return null_handle_cmd (cmd, sector, nr_sectors,req_op (bd->rq)); +return null_handle_cmd (cmd, sector, nr_sectors,req_op (rq)); }static void cleanup_queue (struct nullb_queue *nq)
Signature · 2 lines
Damien Le Moal Western Digital Research
PATCH v2 6/6 block/null_blk: Add support for pipelining zoned writes
In reply to Bart Van AsscheAdd a new configfs attribute for enabling pipelining of zoned writes. If that attribute has been set, retry zoned writes that are not aligned with the write pointer. The test script below reports 236 K IOPS with no I/O scheduler, 81 K IOPS with mq-deadline and pipelining disabled and 121 K IOPS with mq-deadline and pipelining enabled (+49%).
#!/bin/bash
for d in /sys/kernel/config/nullb/*; do
[ -d "$d" ] && rmdir "$d"
done
modprobe -r null_blk
set -e
modprobe null_blk nr_devices=0
udevadm settle
(
cd /sys/kernel/config/nullb
mkdir nullb0
cd nullb0
params=(
completion_nsec=100000
hw_queue_depth=64
irqmode=2
memory_backed=1
pipeline_zoned_writes=1
size=1
submit_queues=1
zone_size=1
zoned=1
power=1
)
for p in "${params[@]}"; do
echo "${p//*=}" > "${p//=*}"
done
)
params=(
--direct=1
--filename=/dev/nullb0
--iodepth=64
--iodepth_batch=16
--ioengine=io_uring
--ioscheduler=mq-deadline
--hipri=0
--name=nullb0
--runtime=30
--rw=write
--time_based=1
--zonemode=zbd
)
fio "${params[@]}"--- drivers/block/null_blk/main.c | 9 +++++++++ drivers/block/null_blk/null_blk.h | 3 +++ drivers/block/null_blk/zoned.c | 4 +++- 3 files changed, 15 insertions(+), 1 deletion(-)
drivers/block/null_blk/main.c +9 -0
@@ -408,6 +408,7 @@ NULLB_DEVICE_ATTR(zone_capacity, ulong, NULL);NULLB_DEVICE_ATTR (zone_nr_conv, uint,NULL );NULLB_DEVICE_ATTR (zone_max_open, uint,NULL );NULLB_DEVICE_ATTR (zone_max_active, uint,NULL ); +NULLB_DEVICE_ATTR (pipeline_zoned_writes, bool,NULL );NULLB_DEVICE_ATTR (virt_boundary, bool,NULL );static ssize_t nullb_device_power_show (struct config_item *item,char *page)
@@ -531,6 +532,7 @@ static struct configfs_attribute *nullb_device_attrs[] = { &nullb_device_attr_zone_nr_conv, &nullb_device_attr_zone_max_open, &nullb_device_attr_zone_max_active, + &nullb_device_attr_pipeline_zoned_writes, &nullb_device_attr_virt_boundary, NULL, };
@@ -1626,6 +1628,11 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx, cmd->error =BLK_STS_OK ; cmd->nq = nq; cmd->fake_timeout =should_timeout_request (rq); +if (!(rq->rq_flags &RQF_DONTPREP )) { + rq->rq_flags |=RQF_DONTPREP ; + cmd->retries =0 ; + cmd->max_attempts = rq->q->nr_hw_queues * rq->q->nr_requests; + }blk_mq_start_request (rq);
@@ -2042,6 +2049,8 @@ static int null_add_dev(struct nullb_device *dev) nullb->q->queuedata = nullb;blk_queue_flag_set (QUEUE_FLAG_NONROT , nullb->q);blk_queue_flag_clear (QUEUE_FLAG_ADD_RANDOM , nullb->q); +if (dev->pipeline_zoned_writes) +blk_queue_flag_set (QUEUE_FLAG_PIPELINE_ZONED_WRITES , nullb->q);mutex_lock (&lock); nullb->index =ida_simple_get (&nullb_indexes,0 ,0 ,GFP_KERNEL );
drivers/block/null_blk/null_blk.h +3 -0
@@ -23,6 +23,8 @@ struct nullb_cmd {unsigned int tag;blk_status_t error;bool fake_timeout; +u16 retries; +u16 max_attempts;struct nullb_queue *nq;struct hrtimer timer; };
@@ -112,6 +114,7 @@ struct nullb_device {bool memory_backed;/* if data is stored in memory */ bool discard;/* if support discard */ bool zoned;/* if device is zoned */ +bool pipeline_zoned_writes;bool virt_boundary;/* virtual boundary on/off for the device */ };
drivers/block/null_blk/zoned.c +3 -1
@@ -403,7 +403,9 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,else cmd->bio->bi_iter.bi_sector = sector; }else if (sector != zone->wp) { - ret =BLK_STS_IOERR ; + ret = dev->pipeline_zoned_writes && + ++cmd->retries < cmd->max_attempts ? +BLK_STS_DEV_RESOURCE :BLK_STS_IOERR ; goto unlock; }
Re: [PATCH v2 6/6] block/null_blk: Add support for pipelining zoned writes
Quoted 126 lines · Bart Van Assche
Add a new configfs attribute for enabling pipelining of zoned writes. If that attribute has been set, retry zoned writes that are not aligned with the write pointer. The test script below reports 236 K IOPS with no I/O scheduler, 81 K IOPS with mq-deadline and pipelining disabled and 121 K IOPS with mq-deadline and pipelining enabled (+49%).
#!/bin/bash for d in /sys/kernel/config/nullb/*; do [ -d "$d" ] && rmdir "$d" done modprobe -r null_blk set -e modprobe null_blk nr_devices=0 udevadm settle ( cd /sys/kernel/config/nullb mkdir nullb0 cd nullb0 params=( completion_nsec=100000 hw_queue_depth=64 irqmode=2 memory_backed=1 pipeline_zoned_writes=1 size=1 submit_queues=1 zone_size=1 zoned=1 power=1 ) for p in "${params[@]}"; do echo "${p//*=}" > "${p//=*}" done ) params=( --direct=1 --filename=/dev/nullb0 --iodepth=64 --iodepth_batch=16 --ioengine=io_uring --ioscheduler=mq-deadline --hipri=0 --name=nullb0 --runtime=30 --rw=write --time_based=1 --zonemode=zbd ) fio "${params[@]}"Cc Damien Le Moal <[email protected]>Signed-off-by Bart Van Assche <[email protected]>--- drivers/block/null_blk/main.c | 9 +++++++++ drivers/block/null_blk/null_blk.h | 3 +++ drivers/block/null_blk/zoned.c | 4 +++- 3 files changed, 15 insertions(+), 1 deletion(-)drivers/block/null_blk/main.c +9
-0@@ -408,6 +408,7 @@ NULLB_DEVICE_ATTR(zone_capacity, ulong, NULL);NULLB_DEVICE_ATTR (zone_nr_conv, uint,NULL );NULLB_DEVICE_ATTR (zone_max_open, uint,NULL );NULLB_DEVICE_ATTR (zone_max_active, uint,NULL ); +NULLB_DEVICE_ATTR (pipeline_zoned_writes, bool,NULL );NULLB_DEVICE_ATTR (virt_boundary, bool,NULL );static ssize_t nullb_device_power_show (struct config_item *item,char *page)@@ -531,6 +532,7 @@ static struct configfs_attribute *nullb_device_attrs[] = { &nullb_device_attr_zone_nr_conv, &nullb_device_attr_zone_max_open, &nullb_device_attr_zone_max_active, + &nullb_device_attr_pipeline_zoned_writes, &nullb_device_attr_virt_boundary, NULL, };@@ -1626,6 +1628,11 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx, cmd->error =BLK_STS_OK ; cmd->nq = nq; cmd->fake_timeout =should_timeout_request (rq); +if (!(rq->rq_flags &RQF_DONTPREP )) { + rq->rq_flags |=RQF_DONTPREP ; + cmd->retries =0 ; + cmd->max_attempts = rq->q->nr_hw_queues * rq->q->nr_requests; + }blk_mq_start_request (rq);@@ -2042,6 +2049,8 @@ static int null_add_dev(struct nullb_device *dev) nullb->q->queuedata = nullb;blk_queue_flag_set (QUEUE_FLAG_NONROT , nullb->q);blk_queue_flag_clear (QUEUE_FLAG_ADD_RANDOM , nullb->q); +if (dev->pipeline_zoned_writes) +blk_queue_flag_set (QUEUE_FLAG_PIPELINE_ZONED_WRITES , nullb->q);mutex_lock (&lock); nullb->index =ida_simple_get (&nullb_indexes,0 ,0 ,GFP_KERNEL );drivers/block/null_blk/null_blk.h +3
-0@@ -23,6 +23,8 @@ struct nullb_cmd {unsigned int tag;blk_status_t error;bool fake_timeout; +u16 retries; +u16 max_attempts;struct nullb_queue *nq;struct hrtimer timer; };@@ -112,6 +114,7 @@ struct nullb_device {bool memory_backed;/* if data is stored in memory */ bool discard;/* if support discard */ bool zoned;/* if device is zoned */ +bool pipeline_zoned_writes;bool virt_boundary;/* virtual boundary on/off for the device */ };drivers/block/null_blk/zoned.c +0
-0
drivers/block/null_blk/zoned.c static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector, · line 408
Quoted 7 lines
drivers/block/null_blk/zoned.c +3
-1@@ -403,7 +403,9 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,else cmd->bio->bi_iter.bi_sector = sector; }else if (sector != zone->wp) { - ret =BLK_STS_IOERR ; + ret = dev->pipeline_zoned_writes && + ++cmd->retries < cmd->max_attempts ? +BLK_STS_DEV_RESOURCE :BLK_STS_IOERR ;
Hard to read. Could you change this to a plain "if()" ?
if (dev->pipeline_zoned_writes && ++cmd->retries < cmd->max_attempts) ret = BLK_STS_DEV_RESOURCE; else ret = BLK_STS_IOERR;
Adding a comment on top of this hunk explaining the difference between the 2 cases would be nice too.
Quoted 3 lines
drivers/block/null_blk/zoned.c +0
-0goto unlock; }
Signature · 2 lines
Damien Le Moal Western Digital Research
Re: [PATCH v2 6/6] block/null_blk: Add support for pipelining zoned writes
In reply to Bart Van AsscheI don't think adding all the previous code just for null_blk is a good idea. We need a real user of it first.
Re: [PATCH v2 6/6] block/null_blk: Add support for pipelining zoned writes
Quoted 2 lines · Christoph Hellwig
I don't think adding all the previous code just for
null_blkis a good idea. We need a real user of it first.
Hi Christoph,
Although these patches work fine in our tests with zoned UFS prototypes, these patches have been tested with a UFSHCI 3.0 controller. While the UFSHCI 4.0 standard supports multiple queues and while UFSHCI 4.0 controllers process each queue in order, with UFSHCI 3.0 controllers submitting requests happens by writing into a 32-bit doorbell register, just like for AHCI controllers. So although our UFS tests pass it cannot be guaranteed for UFSHCI 3.0 controllers that requests are propagated to the UFS device in the same order these have been submitted to the controller.
I have not yet seen any UFSHCI 4.0 implementation. The upstream UFSHCI driver does not yet support it either.
Hence the choice to add support to the null_blk driver for setting the QUEUE_FLAG_PIPELINE_ZONED_WRITES flag.
Thanks,
Bart.
Re: [PATCH v2 6/6] block/null_blk: Add support for pipelining zoned writes
Quoted 5 lines · Bart Van Assche
I have not yet seen any UFSHCI 4.0 implementation. The upstream UFSHCI driver does not yet support it either.
Hence the choice to add support to the
null_blkdriver for setting theQUEUE_FLAG_PIPELINE_ZONED_WRITESflag.
Also I don't think UFS even supports zoned devices as is? Either way without an in-tree driver for actually useful and existing hardware we should not add extra code to the block layer.
Re: [PATCH v2 6/6] block/null_blk: Add support for pipelining zoned writes
Quoted 8 lines · Christoph Hellwig
Quoted 5 lines · Bart Van Assche
I have not yet seen any UFSHCI 4.0 implementation. The upstream UFSHCI driver does not yet support it either.
Hence the choice to add support to the
null_blkdriver for setting theQUEUE_FLAG_PIPELINE_ZONED_WRITESflag.Also I don't think UFS even supports zoned devices as is?
The following ballot has been approved about two weeks ago and is on its way to become a standard: https://www.jedec.org/members/documents/84467/download (file name: Zoned_Storage_for_UFS_Ballot_v3; only JEDEC members have permission to download this document).
Quoted 2 lines
Either way> without an in-tree driver for actually useful and existing hardware we should not add extra code to the block layer.
There is a zoned NVMe SSD in my workstation. How about adding support for zoned NVMe SSDs in this patch series?
Thanks,
Bart.