* [PATCH v6 1/6] block: Add @flags argument to bdev_write_zeroes_sectors()
2020-02-10 9:33 [PATCH v6 0/6] block: Introduce REQ_ALLOCATE flag for REQ_OP_WRITE_ZEROES Kirill Tkhai
@ 2020-02-10 9:33 ` Kirill Tkhai
2020-02-10 9:33 ` [PATCH v6 2/6] block: Pass op_flags into blk_queue_get_max_sectors() Kirill Tkhai
` (4 subsequent siblings)
5 siblings, 0 replies; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-10 9:33 UTC (permalink / raw)
To: martin.petersen, bob.liu, axboe
Cc: agk, snitzer, dm-devel, song, tytso, adilger.kernel,
Chaitanya.Kulkarni, darrick.wong, ming.lei, osandov, jthumshirn,
minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj, ajay.joshi,
sagi, dsterba, chaitanya.kulkarni, bvanassche, dhowells,
asml.silence, linux-block, linux-kernel, ktkhai
This is a preparation for next patch, which introduces
a new flag BLKDEV_ZERO_ALLOCATE for calls, which need
only allocation of blocks and don't need actual blocks
zeroing.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
---
block/blk-lib.c | 6 +++---
drivers/md/dm-kcopyd.c | 2 +-
drivers/target/target_core_iblock.c | 4 ++--
include/linux/blkdev.h | 3 ++-
4 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/block/blk-lib.c b/block/blk-lib.c
index 5f2c429d4378..3e38c93cfc53 100644
--- a/block/blk-lib.c
+++ b/block/blk-lib.c
@@ -224,7 +224,7 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
return -EPERM;
/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
- max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev);
+ max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, 0);
if (max_write_zeroes_sectors == 0)
return -EOPNOTSUPP;
@@ -362,7 +362,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
sector_t bs_mask;
struct bio *bio;
struct blk_plug plug;
- bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev);
+ bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, 0);
bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
if ((sector | nr_sects) & bs_mask)
@@ -391,7 +391,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
try_write_zeroes = false;
goto retry;
}
- if (!bdev_write_zeroes_sectors(bdev)) {
+ if (!bdev_write_zeroes_sectors(bdev, 0)) {
/*
* Zeroing offload support was indicated, but the
* device reported ILLEGAL REQUEST (for some devices
diff --git a/drivers/md/dm-kcopyd.c b/drivers/md/dm-kcopyd.c
index 1bbe4a34ef4c..f1b8e7926dd4 100644
--- a/drivers/md/dm-kcopyd.c
+++ b/drivers/md/dm-kcopyd.c
@@ -831,7 +831,7 @@ void dm_kcopyd_copy(struct dm_kcopyd_client *kc, struct dm_io_region *from,
*/
job->rw = REQ_OP_WRITE_ZEROES;
for (i = 0; i < job->num_dests; i++)
- if (!bdev_write_zeroes_sectors(job->dests[i].bdev)) {
+ if (!bdev_write_zeroes_sectors(job->dests[i].bdev, 0)) {
job->rw = WRITE;
break;
}
diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c
index 51ffd5c002de..73a63e197bf5 100644
--- a/drivers/target/target_core_iblock.c
+++ b/drivers/target/target_core_iblock.c
@@ -117,7 +117,7 @@ static int iblock_configure_device(struct se_device *dev)
* Enable write same emulation for IBLOCK and use 0xFFFF as
* the smaller WRITE_SAME(10) only has a two-byte block count.
*/
- max_write_zeroes_sectors = bdev_write_zeroes_sectors(bd);
+ max_write_zeroes_sectors = bdev_write_zeroes_sectors(bd, 0);
if (max_write_zeroes_sectors)
dev->dev_attrib.max_write_same_len = max_write_zeroes_sectors;
else
@@ -468,7 +468,7 @@ iblock_execute_write_same(struct se_cmd *cmd)
return TCM_INVALID_CDB_FIELD;
}
- if (bdev_write_zeroes_sectors(bdev)) {
+ if (bdev_write_zeroes_sectors(bdev, 0)) {
if (!iblock_execute_zero_out(bdev, cmd))
return 0;
}
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 053ea4b51988..1e3c08854b09 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -1417,7 +1417,8 @@ static inline unsigned int bdev_write_same(struct block_device *bdev)
return 0;
}
-static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev)
+static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev,
+ unsigned int flags)
{
struct request_queue *q = bdev_get_queue(bdev);
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v6 2/6] block: Pass op_flags into blk_queue_get_max_sectors()
2020-02-10 9:33 [PATCH v6 0/6] block: Introduce REQ_ALLOCATE flag for REQ_OP_WRITE_ZEROES Kirill Tkhai
2020-02-10 9:33 ` [PATCH v6 1/6] block: Add @flags argument to bdev_write_zeroes_sectors() Kirill Tkhai
@ 2020-02-10 9:33 ` Kirill Tkhai
2020-02-10 9:33 ` [PATCH v6 3/6] block: Introduce blk_queue_get_max_write_zeroes_sectors() Kirill Tkhai
` (3 subsequent siblings)
5 siblings, 0 replies; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-10 9:33 UTC (permalink / raw)
To: martin.petersen, bob.liu, axboe
Cc: agk, snitzer, dm-devel, song, tytso, adilger.kernel,
Chaitanya.Kulkarni, darrick.wong, ming.lei, osandov, jthumshirn,
minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj, ajay.joshi,
sagi, dsterba, chaitanya.kulkarni, bvanassche, dhowells,
asml.silence, linux-block, linux-kernel, ktkhai
This preparation patch changes argument type, and now
the function takes full op_flags instead of just op code.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
---
block/blk-core.c | 4 ++--
include/linux/blkdev.h | 8 +++++---
2 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 089e890ab208..28a6d46eb982 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1221,10 +1221,10 @@ EXPORT_SYMBOL(submit_bio);
static int blk_cloned_rq_check_limits(struct request_queue *q,
struct request *rq)
{
- if (blk_rq_sectors(rq) > blk_queue_get_max_sectors(q, req_op(rq))) {
+ if (blk_rq_sectors(rq) > blk_queue_get_max_sectors(q, rq->cmd_flags)) {
printk(KERN_ERR "%s: over max size limit. (%u > %u)\n",
__func__, blk_rq_sectors(rq),
- blk_queue_get_max_sectors(q, req_op(rq)));
+ blk_queue_get_max_sectors(q, rq->cmd_flags));
return -EIO;
}
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 1e3c08854b09..f4ec7ae214ab 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -988,8 +988,10 @@ static inline struct bio_vec req_bvec(struct request *rq)
}
static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
- int op)
+ unsigned int op_flags)
{
+ int op = op_flags & REQ_OP_MASK;
+
if (unlikely(op == REQ_OP_DISCARD || op == REQ_OP_SECURE_ERASE))
return min(q->limits.max_discard_sectors,
UINT_MAX >> SECTOR_SHIFT);
@@ -1028,10 +1030,10 @@ static inline unsigned int blk_rq_get_max_sectors(struct request *rq,
if (!q->limits.chunk_sectors ||
req_op(rq) == REQ_OP_DISCARD ||
req_op(rq) == REQ_OP_SECURE_ERASE)
- return blk_queue_get_max_sectors(q, req_op(rq));
+ return blk_queue_get_max_sectors(q, rq->cmd_flags);
return min(blk_max_size_offset(q, offset),
- blk_queue_get_max_sectors(q, req_op(rq)));
+ blk_queue_get_max_sectors(q, rq->cmd_flags));
}
static inline unsigned int blk_rq_count_bios(struct request *rq)
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v6 3/6] block: Introduce blk_queue_get_max_write_zeroes_sectors()
2020-02-10 9:33 [PATCH v6 0/6] block: Introduce REQ_ALLOCATE flag for REQ_OP_WRITE_ZEROES Kirill Tkhai
2020-02-10 9:33 ` [PATCH v6 1/6] block: Add @flags argument to bdev_write_zeroes_sectors() Kirill Tkhai
2020-02-10 9:33 ` [PATCH v6 2/6] block: Pass op_flags into blk_queue_get_max_sectors() Kirill Tkhai
@ 2020-02-10 9:33 ` Kirill Tkhai
2020-02-10 9:33 ` [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag Kirill Tkhai
` (2 subsequent siblings)
5 siblings, 0 replies; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-10 9:33 UTC (permalink / raw)
To: martin.petersen, bob.liu, axboe
Cc: agk, snitzer, dm-devel, song, tytso, adilger.kernel,
Chaitanya.Kulkarni, darrick.wong, ming.lei, osandov, jthumshirn,
minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj, ajay.joshi,
sagi, dsterba, chaitanya.kulkarni, bvanassche, dhowells,
asml.silence, linux-block, linux-kernel, ktkhai
This introduces a new primitive, which returns max sectors
for REQ_OP_WRITE_ZEROES operation.
@op_flags is unused now, and it will be enabled in next patch.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
---
block/blk-core.c | 2 +-
block/blk-merge.c | 9 ++++++---
include/linux/blkdev.h | 8 +++++++-
3 files changed, 14 insertions(+), 5 deletions(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 28a6d46eb982..c7387b0d69e5 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -949,7 +949,7 @@ generic_make_request_checks(struct bio *bio)
goto not_supported;
break;
case REQ_OP_WRITE_ZEROES:
- if (!q->limits.max_write_zeroes_sectors)
+ if (!blk_queue_get_max_write_zeroes_sectors(q, bio->bi_opf))
goto not_supported;
break;
default:
diff --git a/block/blk-merge.c b/block/blk-merge.c
index 1534ed736363..467b292bc6e8 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -105,15 +105,18 @@ static struct bio *blk_bio_discard_split(struct request_queue *q,
static struct bio *blk_bio_write_zeroes_split(struct request_queue *q,
struct bio *bio, struct bio_set *bs, unsigned *nsegs)
{
+ unsigned int max_sectors;
+
+ max_sectors = blk_queue_get_max_write_zeroes_sectors(q, bio->bi_opf);
*nsegs = 0;
- if (!q->limits.max_write_zeroes_sectors)
+ if (!max_sectors)
return NULL;
- if (bio_sectors(bio) <= q->limits.max_write_zeroes_sectors)
+ if (bio_sectors(bio) <= max_sectors)
return NULL;
- return bio_split(bio, q->limits.max_write_zeroes_sectors, GFP_NOIO, bs);
+ return bio_split(bio, max_sectors, GFP_NOIO, bs);
}
static struct bio *blk_bio_write_same_split(struct request_queue *q,
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f4ec7ae214ab..55a714161684 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -987,6 +987,12 @@ static inline struct bio_vec req_bvec(struct request *rq)
return mp_bvec_iter_bvec(rq->bio->bi_io_vec, rq->bio->bi_iter);
}
+static inline unsigned int blk_queue_get_max_write_zeroes_sectors(
+ struct request_queue *q, unsigned int op_flags)
+{
+ return q->limits.max_write_zeroes_sectors;
+}
+
static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
unsigned int op_flags)
{
@@ -1000,7 +1006,7 @@ static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q,
return q->limits.max_write_same_sectors;
if (unlikely(op == REQ_OP_WRITE_ZEROES))
- return q->limits.max_write_zeroes_sectors;
+ return blk_queue_get_max_write_zeroes_sectors(q, op_flags);
return q->limits.max_sectors;
}
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag
2020-02-10 9:33 [PATCH v6 0/6] block: Introduce REQ_ALLOCATE flag for REQ_OP_WRITE_ZEROES Kirill Tkhai
` (2 preceding siblings ...)
2020-02-10 9:33 ` [PATCH v6 3/6] block: Introduce blk_queue_get_max_write_zeroes_sectors() Kirill Tkhai
@ 2020-02-10 9:33 ` Kirill Tkhai
2020-02-12 16:58 ` Darrick J. Wong
2020-02-10 9:33 ` [PATCH v6 5/6] block: Add blk_queue_max_allocate_sectors() Kirill Tkhai
2020-02-10 9:34 ` [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE Kirill Tkhai
5 siblings, 1 reply; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-10 9:33 UTC (permalink / raw)
To: martin.petersen, bob.liu, axboe
Cc: agk, snitzer, dm-devel, song, tytso, adilger.kernel,
Chaitanya.Kulkarni, darrick.wong, ming.lei, osandov, jthumshirn,
minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj, ajay.joshi,
sagi, dsterba, chaitanya.kulkarni, bvanassche, dhowells,
asml.silence, linux-block, linux-kernel, ktkhai
This adds support for REQ_ALLOCATE extension of REQ_OP_WRITE_ZEROES
operation, which encourages a block device driver to just allocate
blocks (or mark them allocated) instead of actual blocks zeroing.
REQ_ALLOCATE is aimed to be used for network filesystems providing
a block device interface. Also, block devices, which map a file
on other filesystem (like loop), may use this for less fragmentation
and batching fallocate() requests. Hypervisors like QEMU may
introduce optimizations of clusters allocations based on this.
BLKDEV_ZERO_ALLOCATE is a new corresponding flag for
blkdev_issue_zeroout().
Stacking devices start from zero max_allocate_sectors limit for now,
and the support is going to be implemented separate for each device
in the future.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
---
block/blk-lib.c | 17 ++++++++++-------
block/blk-settings.c | 4 ++++
fs/block_dev.c | 4 ++++
include/linux/blk_types.h | 5 ++++-
include/linux/blkdev.h | 13 ++++++++++---
5 files changed, 32 insertions(+), 11 deletions(-)
diff --git a/block/blk-lib.c b/block/blk-lib.c
index 3e38c93cfc53..9cd6f86523ba 100644
--- a/block/blk-lib.c
+++ b/block/blk-lib.c
@@ -214,7 +214,7 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
struct bio **biop, unsigned flags)
{
struct bio *bio = *biop;
- unsigned int max_write_zeroes_sectors;
+ unsigned int max_write_zeroes_sectors, req_flags = 0;
struct request_queue *q = bdev_get_queue(bdev);
if (!q)
@@ -224,18 +224,21 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
return -EPERM;
/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
- max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, 0);
+ max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, flags);
if (max_write_zeroes_sectors == 0)
return -EOPNOTSUPP;
+ if (flags & BLKDEV_ZERO_NOUNMAP)
+ req_flags |= REQ_NOUNMAP;
+ if (flags & BLKDEV_ZERO_ALLOCATE)
+ req_flags |= REQ_ALLOCATE|REQ_NOUNMAP;
+
while (nr_sects) {
bio = blk_next_bio(bio, 0, gfp_mask);
bio->bi_iter.bi_sector = sector;
bio_set_dev(bio, bdev);
- bio->bi_opf = REQ_OP_WRITE_ZEROES;
- if (flags & BLKDEV_ZERO_NOUNMAP)
- bio->bi_opf |= REQ_NOUNMAP;
+ bio->bi_opf = REQ_OP_WRITE_ZEROES | req_flags;
if (nr_sects > max_write_zeroes_sectors) {
bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
@@ -362,7 +365,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
sector_t bs_mask;
struct bio *bio;
struct blk_plug plug;
- bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, 0);
+ bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, flags);
bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
if ((sector | nr_sects) & bs_mask)
@@ -391,7 +394,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
try_write_zeroes = false;
goto retry;
}
- if (!bdev_write_zeroes_sectors(bdev, 0)) {
+ if (!bdev_write_zeroes_sectors(bdev, flags)) {
/*
* Zeroing offload support was indicated, but the
* device reported ILLEGAL REQUEST (for some devices
diff --git a/block/blk-settings.c b/block/blk-settings.c
index c8eda2e7b91e..8d5df9d37239 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -48,6 +48,7 @@ void blk_set_default_limits(struct queue_limits *lim)
lim->chunk_sectors = 0;
lim->max_write_same_sectors = 0;
lim->max_write_zeroes_sectors = 0;
+ lim->max_allocate_sectors = 0;
lim->max_discard_sectors = 0;
lim->max_hw_discard_sectors = 0;
lim->discard_granularity = 0;
@@ -83,6 +84,7 @@ void blk_set_stacking_limits(struct queue_limits *lim)
lim->max_dev_sectors = UINT_MAX;
lim->max_write_same_sectors = UINT_MAX;
lim->max_write_zeroes_sectors = UINT_MAX;
+ lim->max_allocate_sectors = 0;
}
EXPORT_SYMBOL(blk_set_stacking_limits);
@@ -506,6 +508,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
b->max_write_same_sectors);
t->max_write_zeroes_sectors = min(t->max_write_zeroes_sectors,
b->max_write_zeroes_sectors);
+ t->max_allocate_sectors = min(t->max_allocate_sectors,
+ b->max_allocate_sectors);
t->bounce_pfn = min_not_zero(t->bounce_pfn, b->bounce_pfn);
t->seg_boundary_mask = min_not_zero(t->seg_boundary_mask,
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 69bf2fb6f7cd..1ffef894b3bd 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -2122,6 +2122,10 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
GFP_KERNEL, BLKDEV_ZERO_NOFALLBACK);
break;
+ case FALLOC_FL_KEEP_SIZE:
+ error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
+ GFP_KERNEL, BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
+ break;
case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE | FALLOC_FL_NO_HIDE_STALE:
error = blkdev_issue_discard(bdev, start >> 9, len >> 9,
GFP_KERNEL, 0);
diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
index 70254ae11769..86accd2caa4e 100644
--- a/include/linux/blk_types.h
+++ b/include/linux/blk_types.h
@@ -335,7 +335,9 @@ enum req_flag_bits {
/* command specific flags for REQ_OP_WRITE_ZEROES: */
__REQ_NOUNMAP, /* do not free blocks when zeroing */
-
+ __REQ_ALLOCATE, /* only notify about allocated blocks,
+ * and do not actually zero them
+ */
__REQ_HIPRI,
/* for driver use */
@@ -362,6 +364,7 @@ enum req_flag_bits {
#define REQ_CGROUP_PUNT (1ULL << __REQ_CGROUP_PUNT)
#define REQ_NOUNMAP (1ULL << __REQ_NOUNMAP)
+#define REQ_ALLOCATE (1ULL << __REQ_ALLOCATE)
#define REQ_HIPRI (1ULL << __REQ_HIPRI)
#define REQ_DRV (1ULL << __REQ_DRV)
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 55a714161684..40707f980a2e 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -336,6 +336,7 @@ struct queue_limits {
unsigned int max_hw_discard_sectors;
unsigned int max_write_same_sectors;
unsigned int max_write_zeroes_sectors;
+ unsigned int max_allocate_sectors;
unsigned int discard_granularity;
unsigned int discard_alignment;
@@ -990,6 +991,8 @@ static inline struct bio_vec req_bvec(struct request *rq)
static inline unsigned int blk_queue_get_max_write_zeroes_sectors(
struct request_queue *q, unsigned int op_flags)
{
+ if (op_flags & REQ_ALLOCATE)
+ return q->limits.max_allocate_sectors;
return q->limits.max_write_zeroes_sectors;
}
@@ -1226,6 +1229,7 @@ extern int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
#define BLKDEV_ZERO_NOUNMAP (1 << 0) /* do not free blocks */
#define BLKDEV_ZERO_NOFALLBACK (1 << 1) /* don't write explicit zeroes */
+#define BLKDEV_ZERO_ALLOCATE (1 << 2) /* allocate range of blocks */
extern int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
@@ -1430,10 +1434,13 @@ static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev,
{
struct request_queue *q = bdev_get_queue(bdev);
- if (q)
- return q->limits.max_write_zeroes_sectors;
+ if (!q)
+ return 0;
- return 0;
+ if (flags & BLKDEV_ZERO_ALLOCATE)
+ return q->limits.max_allocate_sectors;
+ else
+ return q->limits.max_write_zeroes_sectors;
}
static inline enum blk_zoned_model bdev_zoned_model(struct block_device *bdev)
^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag
2020-02-10 9:33 ` [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag Kirill Tkhai
@ 2020-02-12 16:58 ` Darrick J. Wong
2020-02-12 17:33 ` Kirill Tkhai
0 siblings, 1 reply; 15+ messages in thread
From: Darrick J. Wong @ 2020-02-12 16:58 UTC (permalink / raw)
To: Kirill Tkhai
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On Mon, Feb 10, 2020 at 12:33:53PM +0300, Kirill Tkhai wrote:
> This adds support for REQ_ALLOCATE extension of REQ_OP_WRITE_ZEROES
> operation, which encourages a block device driver to just allocate
> blocks (or mark them allocated) instead of actual blocks zeroing.
> REQ_ALLOCATE is aimed to be used for network filesystems providing
> a block device interface. Also, block devices, which map a file
> on other filesystem (like loop), may use this for less fragmentation
> and batching fallocate() requests. Hypervisors like QEMU may
> introduce optimizations of clusters allocations based on this.
>
> BLKDEV_ZERO_ALLOCATE is a new corresponding flag for
> blkdev_issue_zeroout().
>
> Stacking devices start from zero max_allocate_sectors limit for now,
> and the support is going to be implemented separate for each device
> in the future.
>
> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> Reviewed-by: Bob Liu <bob.liu@oracle.com>
> ---
> block/blk-lib.c | 17 ++++++++++-------
> block/blk-settings.c | 4 ++++
> fs/block_dev.c | 4 ++++
> include/linux/blk_types.h | 5 ++++-
> include/linux/blkdev.h | 13 ++++++++++---
> 5 files changed, 32 insertions(+), 11 deletions(-)
>
> diff --git a/block/blk-lib.c b/block/blk-lib.c
> index 3e38c93cfc53..9cd6f86523ba 100644
> --- a/block/blk-lib.c
> +++ b/block/blk-lib.c
> @@ -214,7 +214,7 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
> struct bio **biop, unsigned flags)
> {
> struct bio *bio = *biop;
> - unsigned int max_write_zeroes_sectors;
> + unsigned int max_write_zeroes_sectors, req_flags = 0;
> struct request_queue *q = bdev_get_queue(bdev);
>
> if (!q)
> @@ -224,18 +224,21 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
> return -EPERM;
>
> /* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
> - max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, 0);
> + max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, flags);
>
> if (max_write_zeroes_sectors == 0)
> return -EOPNOTSUPP;
>
> + if (flags & BLKDEV_ZERO_NOUNMAP)
> + req_flags |= REQ_NOUNMAP;
> + if (flags & BLKDEV_ZERO_ALLOCATE)
> + req_flags |= REQ_ALLOCATE|REQ_NOUNMAP;
> +
> while (nr_sects) {
> bio = blk_next_bio(bio, 0, gfp_mask);
> bio->bi_iter.bi_sector = sector;
> bio_set_dev(bio, bdev);
> - bio->bi_opf = REQ_OP_WRITE_ZEROES;
> - if (flags & BLKDEV_ZERO_NOUNMAP)
> - bio->bi_opf |= REQ_NOUNMAP;
> + bio->bi_opf = REQ_OP_WRITE_ZEROES | req_flags;
>
> if (nr_sects > max_write_zeroes_sectors) {
> bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
> @@ -362,7 +365,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
> sector_t bs_mask;
> struct bio *bio;
> struct blk_plug plug;
> - bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, 0);
> + bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, flags);
>
> bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
> if ((sector | nr_sects) & bs_mask)
> @@ -391,7 +394,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
> try_write_zeroes = false;
> goto retry;
> }
> - if (!bdev_write_zeroes_sectors(bdev, 0)) {
> + if (!bdev_write_zeroes_sectors(bdev, flags)) {
> /*
> * Zeroing offload support was indicated, but the
> * device reported ILLEGAL REQUEST (for some devices
> diff --git a/block/blk-settings.c b/block/blk-settings.c
> index c8eda2e7b91e..8d5df9d37239 100644
> --- a/block/blk-settings.c
> +++ b/block/blk-settings.c
> @@ -48,6 +48,7 @@ void blk_set_default_limits(struct queue_limits *lim)
> lim->chunk_sectors = 0;
> lim->max_write_same_sectors = 0;
> lim->max_write_zeroes_sectors = 0;
> + lim->max_allocate_sectors = 0;
> lim->max_discard_sectors = 0;
> lim->max_hw_discard_sectors = 0;
> lim->discard_granularity = 0;
> @@ -83,6 +84,7 @@ void blk_set_stacking_limits(struct queue_limits *lim)
> lim->max_dev_sectors = UINT_MAX;
> lim->max_write_same_sectors = UINT_MAX;
> lim->max_write_zeroes_sectors = UINT_MAX;
> + lim->max_allocate_sectors = 0;
> }
> EXPORT_SYMBOL(blk_set_stacking_limits);
>
> @@ -506,6 +508,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
> b->max_write_same_sectors);
> t->max_write_zeroes_sectors = min(t->max_write_zeroes_sectors,
> b->max_write_zeroes_sectors);
> + t->max_allocate_sectors = min(t->max_allocate_sectors,
> + b->max_allocate_sectors);
> t->bounce_pfn = min_not_zero(t->bounce_pfn, b->bounce_pfn);
>
> t->seg_boundary_mask = min_not_zero(t->seg_boundary_mask,
> diff --git a/fs/block_dev.c b/fs/block_dev.c
> index 69bf2fb6f7cd..1ffef894b3bd 100644
> --- a/fs/block_dev.c
> +++ b/fs/block_dev.c
> @@ -2122,6 +2122,10 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
> GFP_KERNEL, BLKDEV_ZERO_NOFALLBACK);
> break;
> + case FALLOC_FL_KEEP_SIZE:
> + error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
> + GFP_KERNEL, BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
I think this should be ^^^ indented to match the other calls.
> + break;
> case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE | FALLOC_FL_NO_HIDE_STALE:
> error = blkdev_issue_discard(bdev, start >> 9, len >> 9,
> GFP_KERNEL, 0);
> diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
> index 70254ae11769..86accd2caa4e 100644
> --- a/include/linux/blk_types.h
> +++ b/include/linux/blk_types.h
> @@ -335,7 +335,9 @@ enum req_flag_bits {
>
> /* command specific flags for REQ_OP_WRITE_ZEROES: */
> __REQ_NOUNMAP, /* do not free blocks when zeroing */
> -
> + __REQ_ALLOCATE, /* only notify about allocated blocks,
> + * and do not actually zero them
"only notify"? Is someone getting a notification? Or are we simply
"notifying" the device that it must ensure allocated blocks?
If it's that last one, then perhaps this should be reworded:
/*
* Ensure the LBA range is backed by physical storage
* without writing zeroes to the blocks.
*/
--D
> + */
> __REQ_HIPRI,
>
> /* for driver use */
> @@ -362,6 +364,7 @@ enum req_flag_bits {
> #define REQ_CGROUP_PUNT (1ULL << __REQ_CGROUP_PUNT)
>
> #define REQ_NOUNMAP (1ULL << __REQ_NOUNMAP)
> +#define REQ_ALLOCATE (1ULL << __REQ_ALLOCATE)
> #define REQ_HIPRI (1ULL << __REQ_HIPRI)
>
> #define REQ_DRV (1ULL << __REQ_DRV)
> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> index 55a714161684..40707f980a2e 100644
> --- a/include/linux/blkdev.h
> +++ b/include/linux/blkdev.h
> @@ -336,6 +336,7 @@ struct queue_limits {
> unsigned int max_hw_discard_sectors;
> unsigned int max_write_same_sectors;
> unsigned int max_write_zeroes_sectors;
> + unsigned int max_allocate_sectors;
> unsigned int discard_granularity;
> unsigned int discard_alignment;
>
> @@ -990,6 +991,8 @@ static inline struct bio_vec req_bvec(struct request *rq)
> static inline unsigned int blk_queue_get_max_write_zeroes_sectors(
> struct request_queue *q, unsigned int op_flags)
> {
> + if (op_flags & REQ_ALLOCATE)
> + return q->limits.max_allocate_sectors;
> return q->limits.max_write_zeroes_sectors;
> }
>
> @@ -1226,6 +1229,7 @@ extern int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
>
> #define BLKDEV_ZERO_NOUNMAP (1 << 0) /* do not free blocks */
> #define BLKDEV_ZERO_NOFALLBACK (1 << 1) /* don't write explicit zeroes */
> +#define BLKDEV_ZERO_ALLOCATE (1 << 2) /* allocate range of blocks */
>
> extern int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
> sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
> @@ -1430,10 +1434,13 @@ static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev,
> {
> struct request_queue *q = bdev_get_queue(bdev);
>
> - if (q)
> - return q->limits.max_write_zeroes_sectors;
> + if (!q)
> + return 0;
>
> - return 0;
> + if (flags & BLKDEV_ZERO_ALLOCATE)
> + return q->limits.max_allocate_sectors;
> + else
> + return q->limits.max_write_zeroes_sectors;
> }
>
> static inline enum blk_zoned_model bdev_zoned_model(struct block_device *bdev)
>
>
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag
2020-02-12 16:58 ` Darrick J. Wong
@ 2020-02-12 17:33 ` Kirill Tkhai
2020-02-12 18:31 ` Darrick J. Wong
0 siblings, 1 reply; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-12 17:33 UTC (permalink / raw)
To: Darrick J. Wong
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
Hi, Darrick,
On 12.02.2020 19:58, Darrick J. Wong wrote:
> On Mon, Feb 10, 2020 at 12:33:53PM +0300, Kirill Tkhai wrote:
>> This adds support for REQ_ALLOCATE extension of REQ_OP_WRITE_ZEROES
>> operation, which encourages a block device driver to just allocate
>> blocks (or mark them allocated) instead of actual blocks zeroing.
>> REQ_ALLOCATE is aimed to be used for network filesystems providing
>> a block device interface. Also, block devices, which map a file
>> on other filesystem (like loop), may use this for less fragmentation
>> and batching fallocate() requests. Hypervisors like QEMU may
>> introduce optimizations of clusters allocations based on this.
>>
>> BLKDEV_ZERO_ALLOCATE is a new corresponding flag for
>> blkdev_issue_zeroout().
>>
>> Stacking devices start from zero max_allocate_sectors limit for now,
>> and the support is going to be implemented separate for each device
>> in the future.
>>
>> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
>> Reviewed-by: Bob Liu <bob.liu@oracle.com>
>> ---
>> block/blk-lib.c | 17 ++++++++++-------
>> block/blk-settings.c | 4 ++++
>> fs/block_dev.c | 4 ++++
>> include/linux/blk_types.h | 5 ++++-
>> include/linux/blkdev.h | 13 ++++++++++---
>> 5 files changed, 32 insertions(+), 11 deletions(-)
>>
>> diff --git a/block/blk-lib.c b/block/blk-lib.c
>> index 3e38c93cfc53..9cd6f86523ba 100644
>> --- a/block/blk-lib.c
>> +++ b/block/blk-lib.c
>> @@ -214,7 +214,7 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
>> struct bio **biop, unsigned flags)
>> {
>> struct bio *bio = *biop;
>> - unsigned int max_write_zeroes_sectors;
>> + unsigned int max_write_zeroes_sectors, req_flags = 0;
>> struct request_queue *q = bdev_get_queue(bdev);
>>
>> if (!q)
>> @@ -224,18 +224,21 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
>> return -EPERM;
>>
>> /* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
>> - max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, 0);
>> + max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, flags);
>>
>> if (max_write_zeroes_sectors == 0)
>> return -EOPNOTSUPP;
>>
>> + if (flags & BLKDEV_ZERO_NOUNMAP)
>> + req_flags |= REQ_NOUNMAP;
>> + if (flags & BLKDEV_ZERO_ALLOCATE)
>> + req_flags |= REQ_ALLOCATE|REQ_NOUNMAP;
>> +
>> while (nr_sects) {
>> bio = blk_next_bio(bio, 0, gfp_mask);
>> bio->bi_iter.bi_sector = sector;
>> bio_set_dev(bio, bdev);
>> - bio->bi_opf = REQ_OP_WRITE_ZEROES;
>> - if (flags & BLKDEV_ZERO_NOUNMAP)
>> - bio->bi_opf |= REQ_NOUNMAP;
>> + bio->bi_opf = REQ_OP_WRITE_ZEROES | req_flags;
>>
>> if (nr_sects > max_write_zeroes_sectors) {
>> bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
>> @@ -362,7 +365,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
>> sector_t bs_mask;
>> struct bio *bio;
>> struct blk_plug plug;
>> - bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, 0);
>> + bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, flags);
>>
>> bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
>> if ((sector | nr_sects) & bs_mask)
>> @@ -391,7 +394,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
>> try_write_zeroes = false;
>> goto retry;
>> }
>> - if (!bdev_write_zeroes_sectors(bdev, 0)) {
>> + if (!bdev_write_zeroes_sectors(bdev, flags)) {
>> /*
>> * Zeroing offload support was indicated, but the
>> * device reported ILLEGAL REQUEST (for some devices
>> diff --git a/block/blk-settings.c b/block/blk-settings.c
>> index c8eda2e7b91e..8d5df9d37239 100644
>> --- a/block/blk-settings.c
>> +++ b/block/blk-settings.c
>> @@ -48,6 +48,7 @@ void blk_set_default_limits(struct queue_limits *lim)
>> lim->chunk_sectors = 0;
>> lim->max_write_same_sectors = 0;
>> lim->max_write_zeroes_sectors = 0;
>> + lim->max_allocate_sectors = 0;
>> lim->max_discard_sectors = 0;
>> lim->max_hw_discard_sectors = 0;
>> lim->discard_granularity = 0;
>> @@ -83,6 +84,7 @@ void blk_set_stacking_limits(struct queue_limits *lim)
>> lim->max_dev_sectors = UINT_MAX;
>> lim->max_write_same_sectors = UINT_MAX;
>> lim->max_write_zeroes_sectors = UINT_MAX;
>> + lim->max_allocate_sectors = 0;
>> }
>> EXPORT_SYMBOL(blk_set_stacking_limits);
>>
>> @@ -506,6 +508,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
>> b->max_write_same_sectors);
>> t->max_write_zeroes_sectors = min(t->max_write_zeroes_sectors,
>> b->max_write_zeroes_sectors);
>> + t->max_allocate_sectors = min(t->max_allocate_sectors,
>> + b->max_allocate_sectors);
>> t->bounce_pfn = min_not_zero(t->bounce_pfn, b->bounce_pfn);
>>
>> t->seg_boundary_mask = min_not_zero(t->seg_boundary_mask,
>> diff --git a/fs/block_dev.c b/fs/block_dev.c
>> index 69bf2fb6f7cd..1ffef894b3bd 100644
>> --- a/fs/block_dev.c
>> +++ b/fs/block_dev.c
>> @@ -2122,6 +2122,10 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
>> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
>> GFP_KERNEL, BLKDEV_ZERO_NOFALLBACK);
>> break;
>> + case FALLOC_FL_KEEP_SIZE:
>> + error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
>> + GFP_KERNEL, BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
>
> I think this should be ^^^ indented to match the other calls.
The only idea I have about this is something like the below. But the below is over 90 char...
error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
GFP_KERNEL,
BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
Could you please clarify what you mean?
>> + break;
>> case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE | FALLOC_FL_NO_HIDE_STALE:
>> error = blkdev_issue_discard(bdev, start >> 9, len >> 9,
>> GFP_KERNEL, 0);
>> diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
>> index 70254ae11769..86accd2caa4e 100644
>> --- a/include/linux/blk_types.h
>> +++ b/include/linux/blk_types.h
>> @@ -335,7 +335,9 @@ enum req_flag_bits {
>>
>> /* command specific flags for REQ_OP_WRITE_ZEROES: */
>> __REQ_NOUNMAP, /* do not free blocks when zeroing */
>> -
>> + __REQ_ALLOCATE, /* only notify about allocated blocks,
>> + * and do not actually zero them
>
> "only notify"? Is someone getting a notification? Or are we simply
> "notifying" the device that it must ensure allocated blocks?
>
> If it's that last one, then perhaps this should be reworded:
>
> /*
> * Ensure the LBA range is backed by physical storage
> * without writing zeroes to the blocks.
> */
Sounds good.
>> + */
>> __REQ_HIPRI,
>>
>> /* for driver use */
>> @@ -362,6 +364,7 @@ enum req_flag_bits {
>> #define REQ_CGROUP_PUNT (1ULL << __REQ_CGROUP_PUNT)
>>
>> #define REQ_NOUNMAP (1ULL << __REQ_NOUNMAP)
>> +#define REQ_ALLOCATE (1ULL << __REQ_ALLOCATE)
>> #define REQ_HIPRI (1ULL << __REQ_HIPRI)
>>
>> #define REQ_DRV (1ULL << __REQ_DRV)
>> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
>> index 55a714161684..40707f980a2e 100644
>> --- a/include/linux/blkdev.h
>> +++ b/include/linux/blkdev.h
>> @@ -336,6 +336,7 @@ struct queue_limits {
>> unsigned int max_hw_discard_sectors;
>> unsigned int max_write_same_sectors;
>> unsigned int max_write_zeroes_sectors;
>> + unsigned int max_allocate_sectors;
>> unsigned int discard_granularity;
>> unsigned int discard_alignment;
>>
>> @@ -990,6 +991,8 @@ static inline struct bio_vec req_bvec(struct request *rq)
>> static inline unsigned int blk_queue_get_max_write_zeroes_sectors(
>> struct request_queue *q, unsigned int op_flags)
>> {
>> + if (op_flags & REQ_ALLOCATE)
>> + return q->limits.max_allocate_sectors;
>> return q->limits.max_write_zeroes_sectors;
>> }
>>
>> @@ -1226,6 +1229,7 @@ extern int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
>>
>> #define BLKDEV_ZERO_NOUNMAP (1 << 0) /* do not free blocks */
>> #define BLKDEV_ZERO_NOFALLBACK (1 << 1) /* don't write explicit zeroes */
>> +#define BLKDEV_ZERO_ALLOCATE (1 << 2) /* allocate range of blocks */
>>
>> extern int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
>> sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
>> @@ -1430,10 +1434,13 @@ static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev,
>> {
>> struct request_queue *q = bdev_get_queue(bdev);
>>
>> - if (q)
>> - return q->limits.max_write_zeroes_sectors;
>> + if (!q)
>> + return 0;
>>
>> - return 0;
>> + if (flags & BLKDEV_ZERO_ALLOCATE)
>> + return q->limits.max_allocate_sectors;
>> + else
>> + return q->limits.max_write_zeroes_sectors;
>> }
>>
>> static inline enum blk_zoned_model bdev_zoned_model(struct block_device *bdev)
>>
>>
Thanks,
Kirill
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag
2020-02-12 17:33 ` Kirill Tkhai
@ 2020-02-12 18:31 ` Darrick J. Wong
2020-02-12 20:04 ` Kirill Tkhai
0 siblings, 1 reply; 15+ messages in thread
From: Darrick J. Wong @ 2020-02-12 18:31 UTC (permalink / raw)
To: Kirill Tkhai
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On Wed, Feb 12, 2020 at 08:33:52PM +0300, Kirill Tkhai wrote:
> Hi, Darrick,
>
> On 12.02.2020 19:58, Darrick J. Wong wrote:
> > On Mon, Feb 10, 2020 at 12:33:53PM +0300, Kirill Tkhai wrote:
> >> This adds support for REQ_ALLOCATE extension of REQ_OP_WRITE_ZEROES
> >> operation, which encourages a block device driver to just allocate
> >> blocks (or mark them allocated) instead of actual blocks zeroing.
> >> REQ_ALLOCATE is aimed to be used for network filesystems providing
> >> a block device interface. Also, block devices, which map a file
> >> on other filesystem (like loop), may use this for less fragmentation
> >> and batching fallocate() requests. Hypervisors like QEMU may
> >> introduce optimizations of clusters allocations based on this.
> >>
> >> BLKDEV_ZERO_ALLOCATE is a new corresponding flag for
> >> blkdev_issue_zeroout().
> >>
> >> Stacking devices start from zero max_allocate_sectors limit for now,
> >> and the support is going to be implemented separate for each device
> >> in the future.
> >>
> >> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> >> Reviewed-by: Bob Liu <bob.liu@oracle.com>
> >> ---
> >> block/blk-lib.c | 17 ++++++++++-------
> >> block/blk-settings.c | 4 ++++
> >> fs/block_dev.c | 4 ++++
> >> include/linux/blk_types.h | 5 ++++-
> >> include/linux/blkdev.h | 13 ++++++++++---
> >> 5 files changed, 32 insertions(+), 11 deletions(-)
> >>
> >> diff --git a/block/blk-lib.c b/block/blk-lib.c
> >> index 3e38c93cfc53..9cd6f86523ba 100644
> >> --- a/block/blk-lib.c
> >> +++ b/block/blk-lib.c
> >> @@ -214,7 +214,7 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
> >> struct bio **biop, unsigned flags)
> >> {
> >> struct bio *bio = *biop;
> >> - unsigned int max_write_zeroes_sectors;
> >> + unsigned int max_write_zeroes_sectors, req_flags = 0;
> >> struct request_queue *q = bdev_get_queue(bdev);
> >>
> >> if (!q)
> >> @@ -224,18 +224,21 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
> >> return -EPERM;
> >>
> >> /* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
> >> - max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, 0);
> >> + max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, flags);
> >>
> >> if (max_write_zeroes_sectors == 0)
> >> return -EOPNOTSUPP;
> >>
> >> + if (flags & BLKDEV_ZERO_NOUNMAP)
> >> + req_flags |= REQ_NOUNMAP;
> >> + if (flags & BLKDEV_ZERO_ALLOCATE)
> >> + req_flags |= REQ_ALLOCATE|REQ_NOUNMAP;
> >> +
> >> while (nr_sects) {
> >> bio = blk_next_bio(bio, 0, gfp_mask);
> >> bio->bi_iter.bi_sector = sector;
> >> bio_set_dev(bio, bdev);
> >> - bio->bi_opf = REQ_OP_WRITE_ZEROES;
> >> - if (flags & BLKDEV_ZERO_NOUNMAP)
> >> - bio->bi_opf |= REQ_NOUNMAP;
> >> + bio->bi_opf = REQ_OP_WRITE_ZEROES | req_flags;
> >>
> >> if (nr_sects > max_write_zeroes_sectors) {
> >> bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
> >> @@ -362,7 +365,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
> >> sector_t bs_mask;
> >> struct bio *bio;
> >> struct blk_plug plug;
> >> - bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, 0);
> >> + bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, flags);
> >>
> >> bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
> >> if ((sector | nr_sects) & bs_mask)
> >> @@ -391,7 +394,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
> >> try_write_zeroes = false;
> >> goto retry;
> >> }
> >> - if (!bdev_write_zeroes_sectors(bdev, 0)) {
> >> + if (!bdev_write_zeroes_sectors(bdev, flags)) {
> >> /*
> >> * Zeroing offload support was indicated, but the
> >> * device reported ILLEGAL REQUEST (for some devices
> >> diff --git a/block/blk-settings.c b/block/blk-settings.c
> >> index c8eda2e7b91e..8d5df9d37239 100644
> >> --- a/block/blk-settings.c
> >> +++ b/block/blk-settings.c
> >> @@ -48,6 +48,7 @@ void blk_set_default_limits(struct queue_limits *lim)
> >> lim->chunk_sectors = 0;
> >> lim->max_write_same_sectors = 0;
> >> lim->max_write_zeroes_sectors = 0;
> >> + lim->max_allocate_sectors = 0;
> >> lim->max_discard_sectors = 0;
> >> lim->max_hw_discard_sectors = 0;
> >> lim->discard_granularity = 0;
> >> @@ -83,6 +84,7 @@ void blk_set_stacking_limits(struct queue_limits *lim)
> >> lim->max_dev_sectors = UINT_MAX;
> >> lim->max_write_same_sectors = UINT_MAX;
> >> lim->max_write_zeroes_sectors = UINT_MAX;
> >> + lim->max_allocate_sectors = 0;
> >> }
> >> EXPORT_SYMBOL(blk_set_stacking_limits);
> >>
> >> @@ -506,6 +508,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
> >> b->max_write_same_sectors);
> >> t->max_write_zeroes_sectors = min(t->max_write_zeroes_sectors,
> >> b->max_write_zeroes_sectors);
> >> + t->max_allocate_sectors = min(t->max_allocate_sectors,
> >> + b->max_allocate_sectors);
> >> t->bounce_pfn = min_not_zero(t->bounce_pfn, b->bounce_pfn);
> >>
> >> t->seg_boundary_mask = min_not_zero(t->seg_boundary_mask,
> >> diff --git a/fs/block_dev.c b/fs/block_dev.c
> >> index 69bf2fb6f7cd..1ffef894b3bd 100644
> >> --- a/fs/block_dev.c
> >> +++ b/fs/block_dev.c
> >> @@ -2122,6 +2122,10 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
> >> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
> >> GFP_KERNEL, BLKDEV_ZERO_NOFALLBACK);
> >> break;
> >> + case FALLOC_FL_KEEP_SIZE:
> >> + error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
> >> + GFP_KERNEL, BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
> >
> > I think this should be ^^^ indented to match the other calls.
>
> The only idea I have about this is something like the below. But the below is over 90 char...
>
> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
> GFP_KERNEL,
> BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
>
> Could you please clarify what you mean?
I mostly meant that the indent for the nth lines ought to be more than a
single indent to make it easier to scan through the code, but you're
right, the kernel indentation style is uglier. I could suggest
something like this, which actually does fit:
case FALLOC_FL_KEEP_SIZE:
error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
GFP_KERNEL,
BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
break;
But that's apparently apocryphal. :/
--D
> >> case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE | FALLOC_FL_NO_HIDE_STALE:
> >> error = blkdev_issue_discard(bdev, start >> 9, len >> 9,
> >> GFP_KERNEL, 0);
> >> diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
> >> index 70254ae11769..86accd2caa4e 100644
> >> --- a/include/linux/blk_types.h
> >> +++ b/include/linux/blk_types.h
> >> @@ -335,7 +335,9 @@ enum req_flag_bits {
> >>
> >> /* command specific flags for REQ_OP_WRITE_ZEROES: */
> >> __REQ_NOUNMAP, /* do not free blocks when zeroing */
> >> -
> >> + __REQ_ALLOCATE, /* only notify about allocated blocks,
> >> + * and do not actually zero them
> >
> > "only notify"? Is someone getting a notification? Or are we simply
> > "notifying" the device that it must ensure allocated blocks?
> >
> > If it's that last one, then perhaps this should be reworded:
> >
> > /*
> > * Ensure the LBA range is backed by physical storage
> > * without writing zeroes to the blocks.
> > */
>
> Sounds good.
>
> >> + */
> >> __REQ_HIPRI,
> >>
> >> /* for driver use */
> >> @@ -362,6 +364,7 @@ enum req_flag_bits {
> >> #define REQ_CGROUP_PUNT (1ULL << __REQ_CGROUP_PUNT)
> >>
> >> #define REQ_NOUNMAP (1ULL << __REQ_NOUNMAP)
> >> +#define REQ_ALLOCATE (1ULL << __REQ_ALLOCATE)
> >> #define REQ_HIPRI (1ULL << __REQ_HIPRI)
> >>
> >> #define REQ_DRV (1ULL << __REQ_DRV)
> >> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> >> index 55a714161684..40707f980a2e 100644
> >> --- a/include/linux/blkdev.h
> >> +++ b/include/linux/blkdev.h
> >> @@ -336,6 +336,7 @@ struct queue_limits {
> >> unsigned int max_hw_discard_sectors;
> >> unsigned int max_write_same_sectors;
> >> unsigned int max_write_zeroes_sectors;
> >> + unsigned int max_allocate_sectors;
> >> unsigned int discard_granularity;
> >> unsigned int discard_alignment;
> >>
> >> @@ -990,6 +991,8 @@ static inline struct bio_vec req_bvec(struct request *rq)
> >> static inline unsigned int blk_queue_get_max_write_zeroes_sectors(
> >> struct request_queue *q, unsigned int op_flags)
> >> {
> >> + if (op_flags & REQ_ALLOCATE)
> >> + return q->limits.max_allocate_sectors;
> >> return q->limits.max_write_zeroes_sectors;
> >> }
> >>
> >> @@ -1226,6 +1229,7 @@ extern int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
> >>
> >> #define BLKDEV_ZERO_NOUNMAP (1 << 0) /* do not free blocks */
> >> #define BLKDEV_ZERO_NOFALLBACK (1 << 1) /* don't write explicit zeroes */
> >> +#define BLKDEV_ZERO_ALLOCATE (1 << 2) /* allocate range of blocks */
> >>
> >> extern int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
> >> sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
> >> @@ -1430,10 +1434,13 @@ static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev,
> >> {
> >> struct request_queue *q = bdev_get_queue(bdev);
> >>
> >> - if (q)
> >> - return q->limits.max_write_zeroes_sectors;
> >> + if (!q)
> >> + return 0;
> >>
> >> - return 0;
> >> + if (flags & BLKDEV_ZERO_ALLOCATE)
> >> + return q->limits.max_allocate_sectors;
> >> + else
> >> + return q->limits.max_write_zeroes_sectors;
> >> }
> >>
> >> static inline enum blk_zoned_model bdev_zoned_model(struct block_device *bdev)
> >>
> >>
>
> Thanks,
> Kirill
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag
2020-02-12 18:31 ` Darrick J. Wong
@ 2020-02-12 20:04 ` Kirill Tkhai
0 siblings, 0 replies; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-12 20:04 UTC (permalink / raw)
To: Darrick J. Wong
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On 12.02.2020 21:31, Darrick J. Wong wrote:
> On Wed, Feb 12, 2020 at 08:33:52PM +0300, Kirill Tkhai wrote:
>> Hi, Darrick,
>>
>> On 12.02.2020 19:58, Darrick J. Wong wrote:
>>> On Mon, Feb 10, 2020 at 12:33:53PM +0300, Kirill Tkhai wrote:
>>>> This adds support for REQ_ALLOCATE extension of REQ_OP_WRITE_ZEROES
>>>> operation, which encourages a block device driver to just allocate
>>>> blocks (or mark them allocated) instead of actual blocks zeroing.
>>>> REQ_ALLOCATE is aimed to be used for network filesystems providing
>>>> a block device interface. Also, block devices, which map a file
>>>> on other filesystem (like loop), may use this for less fragmentation
>>>> and batching fallocate() requests. Hypervisors like QEMU may
>>>> introduce optimizations of clusters allocations based on this.
>>>>
>>>> BLKDEV_ZERO_ALLOCATE is a new corresponding flag for
>>>> blkdev_issue_zeroout().
>>>>
>>>> Stacking devices start from zero max_allocate_sectors limit for now,
>>>> and the support is going to be implemented separate for each device
>>>> in the future.
>>>>
>>>> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
>>>> Reviewed-by: Bob Liu <bob.liu@oracle.com>
>>>> ---
>>>> block/blk-lib.c | 17 ++++++++++-------
>>>> block/blk-settings.c | 4 ++++
>>>> fs/block_dev.c | 4 ++++
>>>> include/linux/blk_types.h | 5 ++++-
>>>> include/linux/blkdev.h | 13 ++++++++++---
>>>> 5 files changed, 32 insertions(+), 11 deletions(-)
>>>>
>>>> diff --git a/block/blk-lib.c b/block/blk-lib.c
>>>> index 3e38c93cfc53..9cd6f86523ba 100644
>>>> --- a/block/blk-lib.c
>>>> +++ b/block/blk-lib.c
>>>> @@ -214,7 +214,7 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
>>>> struct bio **biop, unsigned flags)
>>>> {
>>>> struct bio *bio = *biop;
>>>> - unsigned int max_write_zeroes_sectors;
>>>> + unsigned int max_write_zeroes_sectors, req_flags = 0;
>>>> struct request_queue *q = bdev_get_queue(bdev);
>>>>
>>>> if (!q)
>>>> @@ -224,18 +224,21 @@ static int __blkdev_issue_write_zeroes(struct block_device *bdev,
>>>> return -EPERM;
>>>>
>>>> /* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
>>>> - max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, 0);
>>>> + max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev, flags);
>>>>
>>>> if (max_write_zeroes_sectors == 0)
>>>> return -EOPNOTSUPP;
>>>>
>>>> + if (flags & BLKDEV_ZERO_NOUNMAP)
>>>> + req_flags |= REQ_NOUNMAP;
>>>> + if (flags & BLKDEV_ZERO_ALLOCATE)
>>>> + req_flags |= REQ_ALLOCATE|REQ_NOUNMAP;
>>>> +
>>>> while (nr_sects) {
>>>> bio = blk_next_bio(bio, 0, gfp_mask);
>>>> bio->bi_iter.bi_sector = sector;
>>>> bio_set_dev(bio, bdev);
>>>> - bio->bi_opf = REQ_OP_WRITE_ZEROES;
>>>> - if (flags & BLKDEV_ZERO_NOUNMAP)
>>>> - bio->bi_opf |= REQ_NOUNMAP;
>>>> + bio->bi_opf = REQ_OP_WRITE_ZEROES | req_flags;
>>>>
>>>> if (nr_sects > max_write_zeroes_sectors) {
>>>> bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
>>>> @@ -362,7 +365,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
>>>> sector_t bs_mask;
>>>> struct bio *bio;
>>>> struct blk_plug plug;
>>>> - bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, 0);
>>>> + bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev, flags);
>>>>
>>>> bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
>>>> if ((sector | nr_sects) & bs_mask)
>>>> @@ -391,7 +394,7 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
>>>> try_write_zeroes = false;
>>>> goto retry;
>>>> }
>>>> - if (!bdev_write_zeroes_sectors(bdev, 0)) {
>>>> + if (!bdev_write_zeroes_sectors(bdev, flags)) {
>>>> /*
>>>> * Zeroing offload support was indicated, but the
>>>> * device reported ILLEGAL REQUEST (for some devices
>>>> diff --git a/block/blk-settings.c b/block/blk-settings.c
>>>> index c8eda2e7b91e..8d5df9d37239 100644
>>>> --- a/block/blk-settings.c
>>>> +++ b/block/blk-settings.c
>>>> @@ -48,6 +48,7 @@ void blk_set_default_limits(struct queue_limits *lim)
>>>> lim->chunk_sectors = 0;
>>>> lim->max_write_same_sectors = 0;
>>>> lim->max_write_zeroes_sectors = 0;
>>>> + lim->max_allocate_sectors = 0;
>>>> lim->max_discard_sectors = 0;
>>>> lim->max_hw_discard_sectors = 0;
>>>> lim->discard_granularity = 0;
>>>> @@ -83,6 +84,7 @@ void blk_set_stacking_limits(struct queue_limits *lim)
>>>> lim->max_dev_sectors = UINT_MAX;
>>>> lim->max_write_same_sectors = UINT_MAX;
>>>> lim->max_write_zeroes_sectors = UINT_MAX;
>>>> + lim->max_allocate_sectors = 0;
>>>> }
>>>> EXPORT_SYMBOL(blk_set_stacking_limits);
>>>>
>>>> @@ -506,6 +508,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
>>>> b->max_write_same_sectors);
>>>> t->max_write_zeroes_sectors = min(t->max_write_zeroes_sectors,
>>>> b->max_write_zeroes_sectors);
>>>> + t->max_allocate_sectors = min(t->max_allocate_sectors,
>>>> + b->max_allocate_sectors);
>>>> t->bounce_pfn = min_not_zero(t->bounce_pfn, b->bounce_pfn);
>>>>
>>>> t->seg_boundary_mask = min_not_zero(t->seg_boundary_mask,
>>>> diff --git a/fs/block_dev.c b/fs/block_dev.c
>>>> index 69bf2fb6f7cd..1ffef894b3bd 100644
>>>> --- a/fs/block_dev.c
>>>> +++ b/fs/block_dev.c
>>>> @@ -2122,6 +2122,10 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
>>>> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
>>>> GFP_KERNEL, BLKDEV_ZERO_NOFALLBACK);
>>>> break;
>>>> + case FALLOC_FL_KEEP_SIZE:
>>>> + error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
>>>> + GFP_KERNEL, BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
>>>
>>> I think this should be ^^^ indented to match the other calls.
>>
>> The only idea I have about this is something like the below. But the below is over 90 char...
>>
>> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
>> GFP_KERNEL,
>> BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
>>
>> Could you please clarify what you mean?
>
> I mostly meant that the indent for the nth lines ought to be more than a
> single indent to make it easier to scan through the code, but you're
> right, the kernel indentation style is uglier. I could suggest
> something like this, which actually does fit:
>
> case FALLOC_FL_KEEP_SIZE:
> error = blkdev_issue_zeroout(bdev, start >> 9, len >> 9,
> GFP_KERNEL,
> BLKDEV_ZERO_ALLOCATE | BLKDEV_ZERO_NOFALLBACK);
> break;
>
> But that's apparently apocryphal. :/
Just to mention.
We could also introduce BLKDEV_ZERO_ALLOCATE_NOFALLBACK, which is combination
of both bits. But this does not look good for me, since this make grep over
kernel code more difficult. And it won't good to introduce a new identifier
for a single place.
> --D
>
>>>> case FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE | FALLOC_FL_NO_HIDE_STALE:
>>>> error = blkdev_issue_discard(bdev, start >> 9, len >> 9,
>>>> GFP_KERNEL, 0);
>>>> diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h
>>>> index 70254ae11769..86accd2caa4e 100644
>>>> --- a/include/linux/blk_types.h
>>>> +++ b/include/linux/blk_types.h
>>>> @@ -335,7 +335,9 @@ enum req_flag_bits {
>>>>
>>>> /* command specific flags for REQ_OP_WRITE_ZEROES: */
>>>> __REQ_NOUNMAP, /* do not free blocks when zeroing */
>>>> -
>>>> + __REQ_ALLOCATE, /* only notify about allocated blocks,
>>>> + * and do not actually zero them
>>>
>>> "only notify"? Is someone getting a notification? Or are we simply
>>> "notifying" the device that it must ensure allocated blocks?
>>>
>>> If it's that last one, then perhaps this should be reworded:
>>>
>>> /*
>>> * Ensure the LBA range is backed by physical storage
>>> * without writing zeroes to the blocks.
>>> */
>>
>> Sounds good.
>>
>>>> + */
>>>> __REQ_HIPRI,
>>>>
>>>> /* for driver use */
>>>> @@ -362,6 +364,7 @@ enum req_flag_bits {
>>>> #define REQ_CGROUP_PUNT (1ULL << __REQ_CGROUP_PUNT)
>>>>
>>>> #define REQ_NOUNMAP (1ULL << __REQ_NOUNMAP)
>>>> +#define REQ_ALLOCATE (1ULL << __REQ_ALLOCATE)
>>>> #define REQ_HIPRI (1ULL << __REQ_HIPRI)
>>>>
>>>> #define REQ_DRV (1ULL << __REQ_DRV)
>>>> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
>>>> index 55a714161684..40707f980a2e 100644
>>>> --- a/include/linux/blkdev.h
>>>> +++ b/include/linux/blkdev.h
>>>> @@ -336,6 +336,7 @@ struct queue_limits {
>>>> unsigned int max_hw_discard_sectors;
>>>> unsigned int max_write_same_sectors;
>>>> unsigned int max_write_zeroes_sectors;
>>>> + unsigned int max_allocate_sectors;
>>>> unsigned int discard_granularity;
>>>> unsigned int discard_alignment;
>>>>
>>>> @@ -990,6 +991,8 @@ static inline struct bio_vec req_bvec(struct request *rq)
>>>> static inline unsigned int blk_queue_get_max_write_zeroes_sectors(
>>>> struct request_queue *q, unsigned int op_flags)
>>>> {
>>>> + if (op_flags & REQ_ALLOCATE)
>>>> + return q->limits.max_allocate_sectors;
>>>> return q->limits.max_write_zeroes_sectors;
>>>> }
>>>>
>>>> @@ -1226,6 +1229,7 @@ extern int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
>>>>
>>>> #define BLKDEV_ZERO_NOUNMAP (1 << 0) /* do not free blocks */
>>>> #define BLKDEV_ZERO_NOFALLBACK (1 << 1) /* don't write explicit zeroes */
>>>> +#define BLKDEV_ZERO_ALLOCATE (1 << 2) /* allocate range of blocks */
>>>>
>>>> extern int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
>>>> sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
>>>> @@ -1430,10 +1434,13 @@ static inline unsigned int bdev_write_zeroes_sectors(struct block_device *bdev,
>>>> {
>>>> struct request_queue *q = bdev_get_queue(bdev);
>>>>
>>>> - if (q)
>>>> - return q->limits.max_write_zeroes_sectors;
>>>> + if (!q)
>>>> + return 0;
>>>>
>>>> - return 0;
>>>> + if (flags & BLKDEV_ZERO_ALLOCATE)
>>>> + return q->limits.max_allocate_sectors;
>>>> + else
>>>> + return q->limits.max_write_zeroes_sectors;
>>>> }
>>>>
>>>> static inline enum blk_zoned_model bdev_zoned_model(struct block_device *bdev)
>>>>
>>>>
>>
>> Thanks,
>> Kirill
^ permalink raw reply [flat|nested] 15+ messages in thread
* [PATCH v6 5/6] block: Add blk_queue_max_allocate_sectors()
2020-02-10 9:33 [PATCH v6 0/6] block: Introduce REQ_ALLOCATE flag for REQ_OP_WRITE_ZEROES Kirill Tkhai
` (3 preceding siblings ...)
2020-02-10 9:33 ` [PATCH v6 4/6] block: Add support for REQ_ALLOCATE flag Kirill Tkhai
@ 2020-02-10 9:33 ` Kirill Tkhai
2020-02-10 9:34 ` [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE Kirill Tkhai
5 siblings, 0 replies; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-10 9:33 UTC (permalink / raw)
To: martin.petersen, bob.liu, axboe
Cc: agk, snitzer, dm-devel, song, tytso, adilger.kernel,
Chaitanya.Kulkarni, darrick.wong, ming.lei, osandov, jthumshirn,
minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj, ajay.joshi,
sagi, dsterba, chaitanya.kulkarni, bvanassche, dhowells,
asml.silence, linux-block, linux-kernel, ktkhai
This is a new helper to assign max_allocate_sectors
limit of block device queue.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
---
block/blk-settings.c | 13 +++++++++++++
include/linux/blkdev.h | 2 ++
2 files changed, 15 insertions(+)
diff --git a/block/blk-settings.c b/block/blk-settings.c
index 8d5df9d37239..24cf8fbbd125 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -259,6 +259,19 @@ void blk_queue_max_write_zeroes_sectors(struct request_queue *q,
}
EXPORT_SYMBOL(blk_queue_max_write_zeroes_sectors);
+/**
+ * blk_queue_max_allocate_sectors - set max sectors for a single
+ * allocate request
+ * @q: the request queue for the device
+ * @max_allocate_sectors: maximum number of sectors to write per command
+ **/
+void blk_queue_max_allocate_sectors(struct request_queue *q,
+ unsigned int max_allocate_sectors)
+{
+ q->limits.max_allocate_sectors = max_allocate_sectors;
+}
+EXPORT_SYMBOL(blk_queue_max_allocate_sectors);
+
/**
* blk_queue_max_segments - set max hw segments for a request for this queue
* @q: the request queue for the device
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 40707f980a2e..f5edbfea7b84 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -1088,6 +1088,8 @@ extern void blk_queue_max_write_same_sectors(struct request_queue *q,
unsigned int max_write_same_sectors);
extern void blk_queue_max_write_zeroes_sectors(struct request_queue *q,
unsigned int max_write_same_sectors);
+extern void blk_queue_max_allocate_sectors(struct request_queue *q,
+ unsigned int max_allocate_sectors);
extern void blk_queue_logical_block_size(struct request_queue *, unsigned int);
extern void blk_queue_physical_block_size(struct request_queue *, unsigned int);
extern void blk_queue_alignment_offset(struct request_queue *q,
^ permalink raw reply related [flat|nested] 15+ messages in thread
* [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE
2020-02-10 9:33 [PATCH v6 0/6] block: Introduce REQ_ALLOCATE flag for REQ_OP_WRITE_ZEROES Kirill Tkhai
` (4 preceding siblings ...)
2020-02-10 9:33 ` [PATCH v6 5/6] block: Add blk_queue_max_allocate_sectors() Kirill Tkhai
@ 2020-02-10 9:34 ` Kirill Tkhai
2020-02-12 17:01 ` Darrick J. Wong
5 siblings, 1 reply; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-10 9:34 UTC (permalink / raw)
To: martin.petersen, bob.liu, axboe
Cc: agk, snitzer, dm-devel, song, tytso, adilger.kernel,
Chaitanya.Kulkarni, darrick.wong, ming.lei, osandov, jthumshirn,
minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj, ajay.joshi,
sagi, dsterba, chaitanya.kulkarni, bvanassche, dhowells,
asml.silence, linux-block, linux-kernel, ktkhai
Support for new modifier of REQ_OP_WRITE_ZEROES command.
This results in allocation extents in backing file instead
of actual blocks zeroing.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
---
drivers/block/loop.c | 15 ++++++++++++---
1 file changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index 739b372a5112..bfe76d9adf09 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -581,6 +581,15 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
return 0;
}
+static unsigned int write_zeroes_to_fallocate_mode(unsigned int flags)
+{
+ if (flags & REQ_ALLOCATE)
+ return 0;
+ if (flags & REQ_NOUNMAP)
+ return FALLOC_FL_ZERO_RANGE;
+ return FALLOC_FL_PUNCH_HOLE;
+}
+
static int do_req_filebacked(struct loop_device *lo, struct request *rq)
{
struct loop_cmd *cmd = blk_mq_rq_to_pdu(rq);
@@ -604,9 +613,7 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
* write zeroes the range. Otherwise, punch them out.
*/
return lo_fallocate(lo, rq, pos,
- (rq->cmd_flags & REQ_NOUNMAP) ?
- FALLOC_FL_ZERO_RANGE :
- FALLOC_FL_PUNCH_HOLE);
+ write_zeroes_to_fallocate_mode(rq->cmd_flags));
case REQ_OP_DISCARD:
return lo_fallocate(lo, rq, pos, FALLOC_FL_PUNCH_HOLE);
case REQ_OP_WRITE:
@@ -877,6 +884,7 @@ static void loop_config_discard(struct loop_device *lo)
q->limits.discard_alignment = 0;
blk_queue_max_discard_sectors(q, 0);
blk_queue_max_write_zeroes_sectors(q, 0);
+ blk_queue_max_allocate_sectors(q, 0);
blk_queue_flag_clear(QUEUE_FLAG_DISCARD, q);
return;
}
@@ -886,6 +894,7 @@ static void loop_config_discard(struct loop_device *lo)
blk_queue_max_discard_sectors(q, UINT_MAX >> 9);
blk_queue_max_write_zeroes_sectors(q, UINT_MAX >> 9);
+ blk_queue_max_allocate_sectors(q, UINT_MAX >> 9);
blk_queue_flag_set(QUEUE_FLAG_DISCARD, q);
}
^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE
2020-02-10 9:34 ` [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE Kirill Tkhai
@ 2020-02-12 17:01 ` Darrick J. Wong
2020-02-12 17:34 ` Kirill Tkhai
2020-02-12 20:00 ` Kirill Tkhai
0 siblings, 2 replies; 15+ messages in thread
From: Darrick J. Wong @ 2020-02-12 17:01 UTC (permalink / raw)
To: Kirill Tkhai
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On Mon, Feb 10, 2020 at 12:34:04PM +0300, Kirill Tkhai wrote:
> Support for new modifier of REQ_OP_WRITE_ZEROES command.
> This results in allocation extents in backing file instead
> of actual blocks zeroing.
>
> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> Reviewed-by: Bob Liu <bob.liu@oracle.com>
> ---
> drivers/block/loop.c | 15 ++++++++++++---
> 1 file changed, 12 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/block/loop.c b/drivers/block/loop.c
> index 739b372a5112..bfe76d9adf09 100644
> --- a/drivers/block/loop.c
> +++ b/drivers/block/loop.c
> @@ -581,6 +581,15 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
> return 0;
> }
>
> +static unsigned int write_zeroes_to_fallocate_mode(unsigned int flags)
> +{
> + if (flags & REQ_ALLOCATE)
> + return 0;
> + if (flags & REQ_NOUNMAP)
> + return FALLOC_FL_ZERO_RANGE;
> + return FALLOC_FL_PUNCH_HOLE;
> +}
> +
> static int do_req_filebacked(struct loop_device *lo, struct request *rq)
> {
> struct loop_cmd *cmd = blk_mq_rq_to_pdu(rq);
> @@ -604,9 +613,7 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
> * write zeroes the range. Otherwise, punch them out.
> */
Please update this comment to reflect the new REQ_ALLOCATE mode, and
move it to where you define write_zeroes_to_fallocate_mode().
With that fixed,
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
--D
> return lo_fallocate(lo, rq, pos,
> - (rq->cmd_flags & REQ_NOUNMAP) ?
> - FALLOC_FL_ZERO_RANGE :
> - FALLOC_FL_PUNCH_HOLE);
> + write_zeroes_to_fallocate_mode(rq->cmd_flags));
> case REQ_OP_DISCARD:
> return lo_fallocate(lo, rq, pos, FALLOC_FL_PUNCH_HOLE);
> case REQ_OP_WRITE:
> @@ -877,6 +884,7 @@ static void loop_config_discard(struct loop_device *lo)
> q->limits.discard_alignment = 0;
> blk_queue_max_discard_sectors(q, 0);
> blk_queue_max_write_zeroes_sectors(q, 0);
> + blk_queue_max_allocate_sectors(q, 0);
> blk_queue_flag_clear(QUEUE_FLAG_DISCARD, q);
> return;
> }
> @@ -886,6 +894,7 @@ static void loop_config_discard(struct loop_device *lo)
>
> blk_queue_max_discard_sectors(q, UINT_MAX >> 9);
> blk_queue_max_write_zeroes_sectors(q, UINT_MAX >> 9);
> + blk_queue_max_allocate_sectors(q, UINT_MAX >> 9);
> blk_queue_flag_set(QUEUE_FLAG_DISCARD, q);
> }
>
>
>
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE
2020-02-12 17:01 ` Darrick J. Wong
@ 2020-02-12 17:34 ` Kirill Tkhai
2020-02-12 20:00 ` Kirill Tkhai
1 sibling, 0 replies; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-12 17:34 UTC (permalink / raw)
To: Darrick J. Wong
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On 12.02.2020 20:01, Darrick J. Wong wrote:
> On Mon, Feb 10, 2020 at 12:34:04PM +0300, Kirill Tkhai wrote:
>> Support for new modifier of REQ_OP_WRITE_ZEROES command.
>> This results in allocation extents in backing file instead
>> of actual blocks zeroing.
>>
>> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
>> Reviewed-by: Bob Liu <bob.liu@oracle.com>
>> ---
>> drivers/block/loop.c | 15 ++++++++++++---
>> 1 file changed, 12 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/block/loop.c b/drivers/block/loop.c
>> index 739b372a5112..bfe76d9adf09 100644
>> --- a/drivers/block/loop.c
>> +++ b/drivers/block/loop.c
>> @@ -581,6 +581,15 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
>> return 0;
>> }
>>
>> +static unsigned int write_zeroes_to_fallocate_mode(unsigned int flags)
>> +{
>> + if (flags & REQ_ALLOCATE)
>> + return 0;
>> + if (flags & REQ_NOUNMAP)
>> + return FALLOC_FL_ZERO_RANGE;
>> + return FALLOC_FL_PUNCH_HOLE;
>> +}
>> +
>> static int do_req_filebacked(struct loop_device *lo, struct request *rq)
>> {
>> struct loop_cmd *cmd = blk_mq_rq_to_pdu(rq);
>> @@ -604,9 +613,7 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
>> * write zeroes the range. Otherwise, punch them out.
>> */
>
> Please update this comment to reflect the new REQ_ALLOCATE mode, and
> move it to where you define write_zeroes_to_fallocate_mode().
Ok, I'll update it in v7
> With that fixed,
>
> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Thanks
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE
2020-02-12 17:01 ` Darrick J. Wong
2020-02-12 17:34 ` Kirill Tkhai
@ 2020-02-12 20:00 ` Kirill Tkhai
2020-02-12 20:44 ` Darrick J. Wong
1 sibling, 1 reply; 15+ messages in thread
From: Kirill Tkhai @ 2020-02-12 20:00 UTC (permalink / raw)
To: Darrick J. Wong
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On 12.02.2020 20:01, Darrick J. Wong wrote:
> On Mon, Feb 10, 2020 at 12:34:04PM +0300, Kirill Tkhai wrote:
>> Support for new modifier of REQ_OP_WRITE_ZEROES command.
>> This results in allocation extents in backing file instead
>> of actual blocks zeroing.
>>
>> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
>> Reviewed-by: Bob Liu <bob.liu@oracle.com>
>> ---
>> drivers/block/loop.c | 15 ++++++++++++---
>> 1 file changed, 12 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/block/loop.c b/drivers/block/loop.c
>> index 739b372a5112..bfe76d9adf09 100644
>> --- a/drivers/block/loop.c
>> +++ b/drivers/block/loop.c
>> @@ -581,6 +581,15 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
>> return 0;
>> }
>>
>> +static unsigned int write_zeroes_to_fallocate_mode(unsigned int flags)
>> +{
>> + if (flags & REQ_ALLOCATE)
>> + return 0;
>> + if (flags & REQ_NOUNMAP)
>> + return FALLOC_FL_ZERO_RANGE;
>> + return FALLOC_FL_PUNCH_HOLE;
>> +}
>> +
>> static int do_req_filebacked(struct loop_device *lo, struct request *rq)
>> {
>> struct loop_cmd *cmd = blk_mq_rq_to_pdu(rq);
>> @@ -604,9 +613,7 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
>> * write zeroes the range. Otherwise, punch them out.
>> */
>
> Please update this comment to reflect the new REQ_ALLOCATE mode, and
> move it to where you define write_zeroes_to_fallocate_mode().
>
> With that fixed,
>
> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Just to clarify. Is this "Reviewed-by:" tag for this patch or for the whole series?
Kirill
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [PATCH v6 6/6] loop: Add support for REQ_ALLOCATE
2020-02-12 20:00 ` Kirill Tkhai
@ 2020-02-12 20:44 ` Darrick J. Wong
0 siblings, 0 replies; 15+ messages in thread
From: Darrick J. Wong @ 2020-02-12 20:44 UTC (permalink / raw)
To: Kirill Tkhai
Cc: martin.petersen, bob.liu, axboe, agk, snitzer, dm-devel, song,
tytso, adilger.kernel, Chaitanya.Kulkarni, ming.lei, osandov,
jthumshirn, minwoo.im.dev, damien.lemoal, andrea.parri, hare, tj,
ajay.joshi, sagi, dsterba, bvanassche, dhowells, asml.silence,
linux-block, linux-kernel
On Wed, Feb 12, 2020 at 11:00:15PM +0300, Kirill Tkhai wrote:
> On 12.02.2020 20:01, Darrick J. Wong wrote:
> > On Mon, Feb 10, 2020 at 12:34:04PM +0300, Kirill Tkhai wrote:
> >> Support for new modifier of REQ_OP_WRITE_ZEROES command.
> >> This results in allocation extents in backing file instead
> >> of actual blocks zeroing.
> >>
> >> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> >> Reviewed-by: Bob Liu <bob.liu@oracle.com>
> >> ---
> >> drivers/block/loop.c | 15 ++++++++++++---
> >> 1 file changed, 12 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/drivers/block/loop.c b/drivers/block/loop.c
> >> index 739b372a5112..bfe76d9adf09 100644
> >> --- a/drivers/block/loop.c
> >> +++ b/drivers/block/loop.c
> >> @@ -581,6 +581,15 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
> >> return 0;
> >> }
> >>
> >> +static unsigned int write_zeroes_to_fallocate_mode(unsigned int flags)
> >> +{
> >> + if (flags & REQ_ALLOCATE)
> >> + return 0;
> >> + if (flags & REQ_NOUNMAP)
> >> + return FALLOC_FL_ZERO_RANGE;
> >> + return FALLOC_FL_PUNCH_HOLE;
> >> +}
> >> +
> >> static int do_req_filebacked(struct loop_device *lo, struct request *rq)
> >> {
> >> struct loop_cmd *cmd = blk_mq_rq_to_pdu(rq);
> >> @@ -604,9 +613,7 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
> >> * write zeroes the range. Otherwise, punch them out.
> >> */
> >
> > Please update this comment to reflect the new REQ_ALLOCATE mode, and
> > move it to where you define write_zeroes_to_fallocate_mode().
> >
> > With that fixed,
> >
> > Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
>
> Just to clarify. Is this "Reviewed-by:" tag for this patch or for the whole series?
Only that patch.
--D
> Kirill
^ permalink raw reply [flat|nested] 15+ messages in thread