[PATCH v5 1/2] bio: limit bio max size

[PATCH v5 1/2] bio: limit bio max size

[Changheun Lee]

bio size can grow up to 4GB when muli-page bvec is enabled.
but sometimes it would lead to inefficient behaviors.
in case of large chunk direct I/O, - 32MB chunk read in user space -
all pages for 32MB would be merged to a bio structure if the pages
physical addresses are contiguous. it makes some delay to submit
until merge complete. bio max size should be limited to a proper size.

When 32MB chunk read with direct I/O option is coming from userspace,
kernel behavior is below now in do_direct_IO() loop. it's timeline.

 | bio merge for 32MB. total 8,192 pages are merged.
 | total elapsed time is over 2ms.
 |------------------ ... ----------------------->|
                                                 | 8,192 pages merged a bio.
                                                 | at this time, first bio submit is done.
                                                 | 1 bio is split to 32 read request and issue.
                                                 |--------------->
                                                  |--------------->
                                                   |--------------->
                                                              ......
                                                                   |--------------->
                                                                    |--------------->|
                          total 19ms elapsed to complete 32MB read done from device. |

If bio max size is limited with 1MB, behavior is changed below.

 | bio merge for 1MB. 256 pages are merged for each bio.
 | total 32 bio will be made.
 | total elapsed time is over 2ms. it's same.
 | but, first bio submit timing is fast. about 100us.
 |--->|--->|--->|---> ... -->|--->|--->|--->|--->|
      | 256 pages merged a bio.
      | at this time, first bio submit is done.
      | and 1 read request is issued for 1 bio.
      |--------------->
           |--------------->
                |--------------->
                                      ......
                                                 |--------------->
                                                  |--------------->|
        total 17ms elapsed to complete 32MB read done from device. |

As a result, read request issue timing is faster if bio max size is limited.
Current kernel behavior with multipage bvec, super large bio can be created.
And it lead to delay first I/O request issue.

Signed-off-by: Changheun Lee <nanich.lee@samsung.com>
---
 block/bio.c            | 13 ++++++++++++-
 include/linux/bio.h    |  2 +-
 include/linux/blkdev.h |  3 +++
 3 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/block/bio.c b/block/bio.c
index 1f2cc1fbe283..c528e1f944c7 100644
--- a/block/bio.c
+++ b/block/bio.c
@@ -287,6 +287,17 @@ void bio_init(struct bio *bio, struct bio_vec *table,
 }
 EXPORT_SYMBOL(bio_init);
 
+unsigned int bio_max_size(struct bio *bio)
+{
+	struct request_queue *q = bio->bi_disk->queue;
+
+	if (blk_queue_limit_bio_size(q))
+		return blk_queue_get_max_sectors(q, bio_op(bio))
+			<< SECTOR_SHIFT;
+
+	return UINT_MAX;
+}
+
 /**
  * bio_reset - reinitialize a bio
  * @bio:	bio to reset
@@ -877,7 +888,7 @@ bool __bio_try_merge_page(struct bio *bio, struct page *page,
 		struct bio_vec *bv = &bio->bi_io_vec[bio->bi_vcnt - 1];
 
 		if (page_is_mergeable(bv, page, len, off, same_page)) {
-			if (bio->bi_iter.bi_size > UINT_MAX - len) {
+			if (bio->bi_iter.bi_size > bio_max_size(bio) - len) {
 				*same_page = false;
 				return false;
 			}
diff --git a/include/linux/bio.h b/include/linux/bio.h
index 1edda614f7ce..13b6f6562a5b 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -113,7 +113,7 @@ static inline bool bio_full(struct bio *bio, unsigned len)
 	if (bio->bi_vcnt >= bio->bi_max_vecs)
 		return true;
 
-	if (bio->bi_iter.bi_size > UINT_MAX - len)
+	if (bio->bi_iter.bi_size > bio_max_size(bio) - len)
 		return true;
 
 	return false;
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f94ee3089e01..3aeab9e7e97b 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -621,6 +621,7 @@ struct request_queue {
 #define QUEUE_FLAG_RQ_ALLOC_TIME 27	/* record rq->alloc_time_ns */
 #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_LIMIT_BIO_SIZE 30	/* limit bio size */
 
 #define QUEUE_FLAG_MQ_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
 				 (1 << QUEUE_FLAG_SAME_COMP) |		\
@@ -667,6 +668,8 @@ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q);
 #define blk_queue_fua(q)	test_bit(QUEUE_FLAG_FUA, &(q)->queue_flags)
 #define blk_queue_registered(q)	test_bit(QUEUE_FLAG_REGISTERED, &(q)->queue_flags)
 #define blk_queue_nowait(q)	test_bit(QUEUE_FLAG_NOWAIT, &(q)->queue_flags)
+#define blk_queue_limit_bio_size(q)	\
+	test_bit(QUEUE_FLAG_LIMIT_BIO_SIZE, &(q)->queue_flags)
 
 extern void blk_set_pm_only(struct request_queue *q);
 extern void blk_clear_pm_only(struct request_queue *q);
-- 
2.28.0

[PATCH v5 2/2] bio: add limit_bio_size sysfs

[Changheun Lee]

Add limit_bio_size block sysfs node to limit bio size.
Queue flag QUEUE_FLAG_LIMIT_BIO_SIZE will be set if limit_bio_size is set.
And bio max size will be limited by queue max sectors via
QUEUE_FLAG_LIMIT_BIO_SIZE set.

Signed-off-by: Changheun Lee <nanich.lee@samsung.com>
---
 Documentation/ABI/testing/sysfs-block | 10 ++++++++++
 Documentation/block/queue-sysfs.rst   |  7 +++++++
 block/blk-sysfs.c                     |  3 +++
 3 files changed, 20 insertions(+)

diff --git a/Documentation/ABI/testing/sysfs-block b/Documentation/ABI/testing/sysfs-block
index e34cdeeeb9d4..86a7b15410cf 100644
--- a/Documentation/ABI/testing/sysfs-block
+++ b/Documentation/ABI/testing/sysfs-block
@@ -316,3 +316,13 @@ Description:
 		does not complete in this time then the block driver timeout
 		handler is invoked. That timeout handler can decide to retry
 		the request, to fail it or to start a device recovery strategy.
+
+What:		/sys/block/<disk>/queue/limit_bio_size
+Date:		Feb, 2021
+Contact:	Changheun Lee <nanich.lee@samsung.com>
+Description:
+		(RW) Toggle for set/clear QUEUE_FLAG_LIMIT_BIO_SIZE queue flag.
+		Queue flag QUEUE_FLAG_LIMIT_BIO_SIZE will be set if limit_bio_size
+		is set. And bio max size will be limited by queue max sectors.
+		QUEUE_FLAG_LIMIT_BIO_SIZE will be cleared if limit_bio_size is
+		cleard. And limit of bio max size will be cleard.
diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index 2638d3446b79..cd371a821855 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -273,4 +273,11 @@ devices are described in the ZBC (Zoned Block Commands) and ZAC
 do not support zone commands, they will be treated as regular block devices
 and zoned will report "none".
 
+limit_bio_size (RW)
+-------------------
+This indicates QUEUE_FLAG_LIMIT_BIO_SIZE queue flag value. And
+QUEUE_FLAG_LIMIT_BIO_SIZE can be changed via set(1)/clear(0) this node.
+bio max size will be limited by queue max sectors via set this node. And
+limit of bio max size will be cleard via clear this node.
+
 Jens Axboe <jens.axboe@oracle.com>, February 2009
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index b513f1683af0..840d97f427e6 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -288,6 +288,7 @@ QUEUE_SYSFS_BIT_FNS(nonrot, NONROT, 1);
 QUEUE_SYSFS_BIT_FNS(random, ADD_RANDOM, 0);
 QUEUE_SYSFS_BIT_FNS(iostats, IO_STAT, 0);
 QUEUE_SYSFS_BIT_FNS(stable_writes, STABLE_WRITES, 0);
+QUEUE_SYSFS_BIT_FNS(limit_bio_size, LIMIT_BIO_SIZE, 0);
 #undef QUEUE_SYSFS_BIT_FNS
 
 static ssize_t queue_zoned_show(struct request_queue *q, char *page)
@@ -615,6 +616,7 @@ QUEUE_RW_ENTRY(queue_nonrot, "rotational");
 QUEUE_RW_ENTRY(queue_iostats, "iostats");
 QUEUE_RW_ENTRY(queue_random, "add_random");
 QUEUE_RW_ENTRY(queue_stable_writes, "stable_writes");
+QUEUE_RW_ENTRY(queue_limit_bio_size, "limit_bio_size");
 
 static struct attribute *queue_attrs[] = {
 	&queue_requests_entry.attr,
@@ -648,6 +650,7 @@ static struct attribute *queue_attrs[] = {
 	&queue_rq_affinity_entry.attr,
 	&queue_iostats_entry.attr,
 	&queue_stable_writes_entry.attr,
+	&queue_limit_bio_size_entry.attr,
 	&queue_random_entry.attr,
 	&queue_poll_entry.attr,
 	&queue_wc_entry.attr,
-- 
2.28.0

[PATCH v5 1/2] bio: limit bio max size

[Changheun Lee]

Please feedback to me if more modification is needed to apply. :)

---
Changheun Lee

Re: [PATCH v5 1/2] bio: limit bio max size

[Greg KH]

On Tue, Feb 16, 2021 at 11:00:32AM +0900, Changheun Lee wrote:
> Please feedback to me if more modification is needed to apply. :)

No context here :(

Re: [PATCH v5 1/2] bio: limit bio max size

[Changheun Lee]

> 
> On Tue, Feb 16, 2021 at 11:00:32AM +0900, Changheun Lee wrote:
> > Please feedback to me if more modification is needed to apply. :)
> 
> No context here :(
> 

I'm so sorry. I missed it.

> bio size can grow up to 4GB when muli-page bvec is enabled.
> but sometimes it would lead to inefficient behaviors.
> in case of large chunk direct I/O, - 32MB chunk read in user space -
> all pages for 32MB would be merged to a bio structure if the pages
> physical addresses are contiguous. it makes some delay to submit
> until merge complete. bio max size should be limited to a proper size.
> 
> When 32MB chunk read with direct I/O option is coming from userspace,
> kernel behavior is below now in do_direct_IO() loop. it's timeline.
> 
>  | bio merge for 32MB. total 8,192 pages are merged.
>  | total elapsed time is over 2ms.
>  |------------------ ... ----------------------->|
>                                                  | 8,192 pages merged a bio.
>                                                  | at this time, first bio submit is done.
>                                                  | 1 bio is split to 32 read request and issue.
>                                                  |--------------->
>                                                   |--------------->
>                                                    |--------------->
>                                                               ......
>                                                                    |--------------->
>                                                                     |--------------->|
>                           total 19ms elapsed to complete 32MB read done from device. |
> 
> If bio max size is limited with 1MB, behavior is changed below.
> 
>  | bio merge for 1MB. 256 pages are merged for each bio.
>  | total 32 bio will be made.
>  | total elapsed time is over 2ms. it's same.
>  | but, first bio submit timing is fast. about 100us.
>  |--->|--->|--->|---> ... -->|--->|--->|--->|--->|
>       | 256 pages merged a bio.
>       | at this time, first bio submit is done.
>       | and 1 read request is issued for 1 bio.
>       |--------------->
>            |--------------->
>                 |--------------->
>                                       ......
>                                                  |--------------->
>                                                   |--------------->|
>         total 17ms elapsed to complete 32MB read done from device. |
> 
> As a result, read request issue timing is faster if bio max size is limited.
> Current kernel behavior with multipage bvec, super large bio can be created.
> And it lead to delay first I/O request issue.
> 
> Signed-off-by: Changheun Lee <nanich.lee@samsung.com>
> ---
>  block/bio.c            | 13 ++++++++++++-
>  include/linux/bio.h    |  2 +-
>  include/linux/blkdev.h |  3 +++
>  3 files changed, 16 insertions(+), 2 deletions(-)
> 
> diff --git a/block/bio.c b/block/bio.c
> index 1f2cc1fbe283..c528e1f944c7 100644
> --- a/block/bio.c
> +++ b/block/bio.c
> @@ -287,6 +287,17 @@ void bio_init(struct bio *bio, struct bio_vec *table,
>  }
>  EXPORT_SYMBOL(bio_init);
>  
> +unsigned int bio_max_size(struct bio *bio)
> +{
> +	struct request_queue *q = bio->bi_disk->queue;
> +
> +	if (blk_queue_limit_bio_size(q))
> +		return blk_queue_get_max_sectors(q, bio_op(bio))
> +			<< SECTOR_SHIFT;
> +
> +	return UINT_MAX;
> +}
> +
>  /**
>   * bio_reset - reinitialize a bio
>   * @bio:	bio to reset
> @@ -877,7 +888,7 @@ bool __bio_try_merge_page(struct bio *bio, struct page *page,
>  		struct bio_vec *bv = &bio->bi_io_vec[bio->bi_vcnt - 1];
>  
>  		if (page_is_mergeable(bv, page, len, off, same_page)) {
> -			if (bio->bi_iter.bi_size > UINT_MAX - len) {
> +			if (bio->bi_iter.bi_size > bio_max_size(bio) - len) {
>  				*same_page = false;
>  				return false;
>  			}
> diff --git a/include/linux/bio.h b/include/linux/bio.h
> index 1edda614f7ce..13b6f6562a5b 100644
> --- a/include/linux/bio.h
> +++ b/include/linux/bio.h
> @@ -113,7 +113,7 @@ static inline bool bio_full(struct bio *bio, unsigned len)
>  	if (bio->bi_vcnt >= bio->bi_max_vecs)
>  		return true;
>  
> -	if (bio->bi_iter.bi_size > UINT_MAX - len)
> +	if (bio->bi_iter.bi_size > bio_max_size(bio) - len)
>  		return true;
>  
>  	return false;
> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> index f94ee3089e01..3aeab9e7e97b 100644
> --- a/include/linux/blkdev.h
> +++ b/include/linux/blkdev.h
> @@ -621,6 +621,7 @@ struct request_queue {
>  #define QUEUE_FLAG_RQ_ALLOC_TIME 27	/* record rq->alloc_time_ns */
>  #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_LIMIT_BIO_SIZE 30	/* limit bio size */
>  
>  #define QUEUE_FLAG_MQ_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
>  				 (1 << QUEUE_FLAG_SAME_COMP) |		\
> @@ -667,6 +668,8 @@ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q);
>  #define blk_queue_fua(q)	test_bit(QUEUE_FLAG_FUA, &(q)->queue_flags)
>  #define blk_queue_registered(q)	test_bit(QUEUE_FLAG_REGISTERED, &(q)->queue_flags)
>  #define blk_queue_nowait(q)	test_bit(QUEUE_FLAG_NOWAIT, &(q)->queue_flags)
> +#define blk_queue_limit_bio_size(q)	\
> +	test_bit(QUEUE_FLAG_LIMIT_BIO_SIZE, &(q)->queue_flags)
>  
>  extern void blk_set_pm_only(struct request_queue *q);
>  extern void blk_clear_pm_only(struct request_queue *q);
> -- 
> 2.28.0
> 

Please feedback to me if more modification is needed to apply. :)

---
Changheun Lee

Re: [PATCH v5 1/2] bio: limit bio max size

[Greg KH]

On Tue, Feb 16, 2021 at 03:42:42PM +0900, Changheun Lee wrote:
> > 
> > On Tue, Feb 16, 2021 at 11:00:32AM +0900, Changheun Lee wrote:
> > > Please feedback to me if more modification is needed to apply. :)
> > 
> > No context here :(
> > 
> 
> I'm so sorry. I missed it.
> 
> > bio size can grow up to 4GB when muli-page bvec is enabled.
> > but sometimes it would lead to inefficient behaviors.
> > in case of large chunk direct I/O, - 32MB chunk read in user space -
> > all pages for 32MB would be merged to a bio structure if the pages
> > physical addresses are contiguous. it makes some delay to submit
> > until merge complete. bio max size should be limited to a proper size.
> > 
> > When 32MB chunk read with direct I/O option is coming from userspace,
> > kernel behavior is below now in do_direct_IO() loop. it's timeline.
> > 
> >  | bio merge for 32MB. total 8,192 pages are merged.
> >  | total elapsed time is over 2ms.
> >  |------------------ ... ----------------------->|
> >                                                  | 8,192 pages merged a bio.
> >                                                  | at this time, first bio submit is done.
> >                                                  | 1 bio is split to 32 read request and issue.
> >                                                  |--------------->
> >                                                   |--------------->
> >                                                    |--------------->
> >                                                               ......
> >                                                                    |--------------->
> >                                                                     |--------------->|
> >                           total 19ms elapsed to complete 32MB read done from device. |
> > 
> > If bio max size is limited with 1MB, behavior is changed below.
> > 
> >  | bio merge for 1MB. 256 pages are merged for each bio.
> >  | total 32 bio will be made.
> >  | total elapsed time is over 2ms. it's same.
> >  | but, first bio submit timing is fast. about 100us.
> >  |--->|--->|--->|---> ... -->|--->|--->|--->|--->|
> >       | 256 pages merged a bio.
> >       | at this time, first bio submit is done.
> >       | and 1 read request is issued for 1 bio.
> >       |--------------->
> >            |--------------->
> >                 |--------------->
> >                                       ......
> >                                                  |--------------->
> >                                                   |--------------->|
> >         total 17ms elapsed to complete 32MB read done from device. |
> > 
> > As a result, read request issue timing is faster if bio max size is limited.
> > Current kernel behavior with multipage bvec, super large bio can be created.
> > And it lead to delay first I/O request issue.
> > 
> > Signed-off-by: Changheun Lee <nanich.lee@samsung.com>
> > ---
> >  block/bio.c            | 13 ++++++++++++-
> >  include/linux/bio.h    |  2 +-
> >  include/linux/blkdev.h |  3 +++
> >  3 files changed, 16 insertions(+), 2 deletions(-)
> > 
> > diff --git a/block/bio.c b/block/bio.c
> > index 1f2cc1fbe283..c528e1f944c7 100644
> > --- a/block/bio.c
> > +++ b/block/bio.c
> > @@ -287,6 +287,17 @@ void bio_init(struct bio *bio, struct bio_vec *table,
> >  }
> >  EXPORT_SYMBOL(bio_init);
> >  
> > +unsigned int bio_max_size(struct bio *bio)
> > +{
> > +	struct request_queue *q = bio->bi_disk->queue;
> > +
> > +	if (blk_queue_limit_bio_size(q))
> > +		return blk_queue_get_max_sectors(q, bio_op(bio))
> > +			<< SECTOR_SHIFT;
> > +
> > +	return UINT_MAX;
> > +}
> > +
> >  /**
> >   * bio_reset - reinitialize a bio
> >   * @bio:	bio to reset
> > @@ -877,7 +888,7 @@ bool __bio_try_merge_page(struct bio *bio, struct page *page,
> >  		struct bio_vec *bv = &bio->bi_io_vec[bio->bi_vcnt - 1];
> >  
> >  		if (page_is_mergeable(bv, page, len, off, same_page)) {
> > -			if (bio->bi_iter.bi_size > UINT_MAX - len) {
> > +			if (bio->bi_iter.bi_size > bio_max_size(bio) - len) {
> >  				*same_page = false;
> >  				return false;
> >  			}
> > diff --git a/include/linux/bio.h b/include/linux/bio.h
> > index 1edda614f7ce..13b6f6562a5b 100644
> > --- a/include/linux/bio.h
> > +++ b/include/linux/bio.h
> > @@ -113,7 +113,7 @@ static inline bool bio_full(struct bio *bio, unsigned len)
> >  	if (bio->bi_vcnt >= bio->bi_max_vecs)
> >  		return true;
> >  
> > -	if (bio->bi_iter.bi_size > UINT_MAX - len)
> > +	if (bio->bi_iter.bi_size > bio_max_size(bio) - len)
> >  		return true;
> >  
> >  	return false;
> > diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> > index f94ee3089e01..3aeab9e7e97b 100644
> > --- a/include/linux/blkdev.h
> > +++ b/include/linux/blkdev.h
> > @@ -621,6 +621,7 @@ struct request_queue {
> >  #define QUEUE_FLAG_RQ_ALLOC_TIME 27	/* record rq->alloc_time_ns */
> >  #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_LIMIT_BIO_SIZE 30	/* limit bio size */
> >  
> >  #define QUEUE_FLAG_MQ_DEFAULT	((1 << QUEUE_FLAG_IO_STAT) |		\
> >  				 (1 << QUEUE_FLAG_SAME_COMP) |		\
> > @@ -667,6 +668,8 @@ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q);
> >  #define blk_queue_fua(q)	test_bit(QUEUE_FLAG_FUA, &(q)->queue_flags)
> >  #define blk_queue_registered(q)	test_bit(QUEUE_FLAG_REGISTERED, &(q)->queue_flags)
> >  #define blk_queue_nowait(q)	test_bit(QUEUE_FLAG_NOWAIT, &(q)->queue_flags)
> > +#define blk_queue_limit_bio_size(q)	\
> > +	test_bit(QUEUE_FLAG_LIMIT_BIO_SIZE, &(q)->queue_flags)
> >  
> >  extern void blk_set_pm_only(struct request_queue *q);
> >  extern void blk_clear_pm_only(struct request_queue *q);
> > -- 
> > 2.28.0
> > 
> 
> Please feedback to me if more modification is needed to apply. :)

It's the middle of the merge window, no maintainers can add anything new
to their trees, so please wait until after 5.12-rc1 is out before we can
do anything...

thanks,

greg k-h