2019-04-30 18:42:43 +00:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
2016-10-18 06:40:33 +00:00
|
|
|
/*
|
|
|
|
* Zoned block device handling
|
|
|
|
*
|
|
|
|
* Copyright (c) 2015, Hannes Reinecke
|
|
|
|
* Copyright (c) 2015, SUSE Linux GmbH
|
|
|
|
*
|
|
|
|
* Copyright (c) 2016, Damien Le Moal
|
|
|
|
* Copyright (c) 2016, Western Digital
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
* Copyright (c) 2024, Western Digital Corporation or its affiliates.
|
2016-10-18 06:40:33 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/blkdev.h>
|
2018-10-12 10:08:50 +00:00
|
|
|
#include <linux/blk-mq.h>
|
2019-07-01 05:09:18 +00:00
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/vmalloc.h>
|
2019-07-01 05:09:16 +00:00
|
|
|
#include <linux/sched/mm.h>
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
#include <linux/spinlock.h>
|
|
|
|
#include <linux/atomic.h>
|
|
|
|
#include <linux/mempool.h>
|
2016-10-18 06:40:33 +00:00
|
|
|
|
2018-10-12 10:08:47 +00:00
|
|
|
#include "blk.h"
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
#include "blk-mq-sched.h"
|
2024-04-08 01:41:24 +00:00
|
|
|
#include "blk-mq-debugfs.h"
|
2018-10-12 10:08:47 +00:00
|
|
|
|
2020-03-25 17:49:54 +00:00
|
|
|
#define ZONE_COND_NAME(name) [BLK_ZONE_COND_##name] = #name
|
|
|
|
static const char *const zone_cond_name[] = {
|
|
|
|
ZONE_COND_NAME(NOT_WP),
|
|
|
|
ZONE_COND_NAME(EMPTY),
|
|
|
|
ZONE_COND_NAME(IMP_OPEN),
|
|
|
|
ZONE_COND_NAME(EXP_OPEN),
|
|
|
|
ZONE_COND_NAME(CLOSED),
|
|
|
|
ZONE_COND_NAME(READONLY),
|
|
|
|
ZONE_COND_NAME(FULL),
|
|
|
|
ZONE_COND_NAME(OFFLINE),
|
|
|
|
};
|
|
|
|
#undef ZONE_COND_NAME
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
/*
|
|
|
|
* Per-zone write plug.
|
|
|
|
* @node: hlist_node structure for managing the plug using a hash table.
|
|
|
|
* @link: To list the plug in the zone write plug error list of the disk.
|
|
|
|
* @ref: Zone write plug reference counter. A zone write plug reference is
|
|
|
|
* always at least 1 when the plug is hashed in the disk plug hash table.
|
|
|
|
* The reference is incremented whenever a new BIO needing plugging is
|
|
|
|
* submitted and when a function needs to manipulate a plug. The
|
|
|
|
* reference count is decremented whenever a plugged BIO completes and
|
|
|
|
* when a function that referenced the plug returns. The initial
|
|
|
|
* reference is dropped whenever the zone of the zone write plug is reset,
|
|
|
|
* finished and when the zone becomes full (last write BIO to the zone
|
|
|
|
* completes).
|
|
|
|
* @lock: Spinlock to atomically manipulate the plug.
|
|
|
|
* @flags: Flags indicating the plug state.
|
|
|
|
* @zone_no: The number of the zone the plug is managing.
|
|
|
|
* @wp_offset: The zone write pointer location relative to the start of the zone
|
|
|
|
* as a number of 512B sectors.
|
|
|
|
* @bio_list: The list of BIOs that are currently plugged.
|
|
|
|
* @bio_work: Work struct to handle issuing of plugged BIOs
|
|
|
|
* @rcu_head: RCU head to free zone write plugs with an RCU grace period.
|
|
|
|
* @disk: The gendisk the plug belongs to.
|
|
|
|
*/
|
|
|
|
struct blk_zone_wplug {
|
|
|
|
struct hlist_node node;
|
|
|
|
struct list_head link;
|
|
|
|
atomic_t ref;
|
|
|
|
spinlock_t lock;
|
|
|
|
unsigned int flags;
|
|
|
|
unsigned int zone_no;
|
|
|
|
unsigned int wp_offset;
|
|
|
|
struct bio_list bio_list;
|
|
|
|
struct work_struct bio_work;
|
|
|
|
struct rcu_head rcu_head;
|
|
|
|
struct gendisk *disk;
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Zone write plug flags bits:
|
|
|
|
* - BLK_ZONE_WPLUG_PLUGGED: Indicates that the zone write plug is plugged,
|
|
|
|
* that is, that write BIOs are being throttled due to a write BIO already
|
|
|
|
* being executed or the zone write plug bio list is not empty.
|
|
|
|
* - BLK_ZONE_WPLUG_ERROR: Indicates that a write error happened which will be
|
|
|
|
* recovered with a report zone to update the zone write pointer offset.
|
|
|
|
* - BLK_ZONE_WPLUG_UNHASHED: Indicates that the zone write plug was removed
|
|
|
|
* from the disk hash table and that the initial reference to the zone
|
|
|
|
* write plug set when the plug was first added to the hash table has been
|
|
|
|
* dropped. This flag is set when a zone is reset, finished or become full,
|
|
|
|
* to prevent new references to the zone write plug to be taken for
|
|
|
|
* newly incoming BIOs. A zone write plug flagged with this flag will be
|
|
|
|
* freed once all remaining references from BIOs or functions are dropped.
|
|
|
|
*/
|
|
|
|
#define BLK_ZONE_WPLUG_PLUGGED (1U << 0)
|
|
|
|
#define BLK_ZONE_WPLUG_ERROR (1U << 1)
|
|
|
|
#define BLK_ZONE_WPLUG_UNHASHED (1U << 2)
|
|
|
|
|
|
|
|
#define BLK_ZONE_WPLUG_BUSY (BLK_ZONE_WPLUG_PLUGGED | BLK_ZONE_WPLUG_ERROR)
|
|
|
|
|
2020-03-25 17:49:54 +00:00
|
|
|
/**
|
|
|
|
* blk_zone_cond_str - Return string XXX in BLK_ZONE_COND_XXX.
|
|
|
|
* @zone_cond: BLK_ZONE_COND_XXX.
|
|
|
|
*
|
|
|
|
* Description: Centralize block layer function to convert BLK_ZONE_COND_XXX
|
|
|
|
* into string format. Useful in the debugging and tracing zone conditions. For
|
|
|
|
* invalid BLK_ZONE_COND_XXX it returns string "UNKNOWN".
|
|
|
|
*/
|
|
|
|
const char *blk_zone_cond_str(enum blk_zone_cond zone_cond)
|
|
|
|
{
|
|
|
|
static const char *zone_cond_str = "UNKNOWN";
|
|
|
|
|
|
|
|
if (zone_cond < ARRAY_SIZE(zone_cond_name) && zone_cond_name[zone_cond])
|
|
|
|
zone_cond_str = zone_cond_name[zone_cond];
|
|
|
|
|
|
|
|
return zone_cond_str;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blk_zone_cond_str);
|
|
|
|
|
2016-10-18 06:40:33 +00:00
|
|
|
/**
|
|
|
|
* blkdev_report_zones - Get zones information
|
|
|
|
* @bdev: Target block device
|
|
|
|
* @sector: Sector from which to report zones
|
2019-11-11 02:39:30 +00:00
|
|
|
* @nr_zones: Maximum number of zones to report
|
|
|
|
* @cb: Callback function called for each reported zone
|
|
|
|
* @data: Private data for the callback
|
2016-10-18 06:40:33 +00:00
|
|
|
*
|
|
|
|
* Description:
|
2019-11-11 02:39:30 +00:00
|
|
|
* Get zone information starting from the zone containing @sector for at most
|
|
|
|
* @nr_zones, and call @cb for each zone reported by the device.
|
|
|
|
* To report all zones in a device starting from @sector, the BLK_ALL_ZONES
|
|
|
|
* constant can be passed to @nr_zones.
|
|
|
|
* Returns the number of zones reported by the device, or a negative errno
|
|
|
|
* value in case of failure.
|
|
|
|
*
|
|
|
|
* Note: The caller must use memalloc_noXX_save/restore() calls to control
|
|
|
|
* memory allocations done within this function.
|
2016-10-18 06:40:33 +00:00
|
|
|
*/
|
2018-10-12 10:08:49 +00:00
|
|
|
int blkdev_report_zones(struct block_device *bdev, sector_t sector,
|
2019-11-11 02:39:30 +00:00
|
|
|
unsigned int nr_zones, report_zones_cb cb, void *data)
|
2016-10-18 06:40:33 +00:00
|
|
|
{
|
2019-11-11 02:39:24 +00:00
|
|
|
struct gendisk *disk = bdev->bd_disk;
|
2019-11-11 02:39:25 +00:00
|
|
|
sector_t capacity = get_capacity(disk);
|
2016-10-18 06:40:33 +00:00
|
|
|
|
2022-07-06 07:03:37 +00:00
|
|
|
if (!bdev_is_zoned(bdev) || WARN_ON_ONCE(!disk->fops->report_zones))
|
2018-10-12 10:08:49 +00:00
|
|
|
return -EOPNOTSUPP;
|
2016-10-18 06:40:33 +00:00
|
|
|
|
2019-11-11 02:39:30 +00:00
|
|
|
if (!nr_zones || sector >= capacity)
|
2016-10-18 06:40:33 +00:00
|
|
|
return 0;
|
|
|
|
|
2019-11-11 02:39:30 +00:00
|
|
|
return disk->fops->report_zones(disk, sector, nr_zones, cb, data);
|
2016-10-18 06:40:33 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blkdev_report_zones);
|
|
|
|
|
2024-01-29 07:52:20 +00:00
|
|
|
static int blkdev_zone_reset_all(struct block_device *bdev)
|
2021-05-25 21:24:51 +00:00
|
|
|
{
|
|
|
|
struct bio bio;
|
|
|
|
|
2022-01-24 09:11:06 +00:00
|
|
|
bio_init(&bio, bdev, NULL, 0, REQ_OP_ZONE_RESET_ALL | REQ_SYNC);
|
2021-05-25 21:24:51 +00:00
|
|
|
return submit_bio_wait(&bio);
|
2019-08-01 17:26:36 +00:00
|
|
|
}
|
|
|
|
|
2016-10-18 06:40:33 +00:00
|
|
|
/**
|
2019-10-27 14:05:45 +00:00
|
|
|
* blkdev_zone_mgmt - Execute a zone management operation on a range of zones
|
2016-10-18 06:40:33 +00:00
|
|
|
* @bdev: Target block device
|
2019-10-27 14:05:45 +00:00
|
|
|
* @op: Operation to be performed on the zones
|
|
|
|
* @sector: Start sector of the first zone to operate on
|
|
|
|
* @nr_sectors: Number of sectors, should be at least the length of one zone and
|
|
|
|
* must be zone size aligned.
|
2016-10-18 06:40:33 +00:00
|
|
|
*
|
|
|
|
* Description:
|
2019-10-27 14:05:45 +00:00
|
|
|
* Perform the specified operation on the range of zones specified by
|
2016-10-18 06:40:33 +00:00
|
|
|
* @sector..@sector+@nr_sectors. Specifying the entire disk sector range
|
|
|
|
* is valid, but the specified range should not contain conventional zones.
|
2019-10-27 14:05:45 +00:00
|
|
|
* The operation to execute on each zone can be a zone reset, open, close
|
|
|
|
* or finish request.
|
2016-10-18 06:40:33 +00:00
|
|
|
*/
|
2022-07-14 18:06:27 +00:00
|
|
|
int blkdev_zone_mgmt(struct block_device *bdev, enum req_op op,
|
2024-01-29 07:52:20 +00:00
|
|
|
sector_t sector, sector_t nr_sectors)
|
2016-10-18 06:40:33 +00:00
|
|
|
{
|
2022-07-06 07:03:46 +00:00
|
|
|
sector_t zone_sectors = bdev_zone_sectors(bdev);
|
|
|
|
sector_t capacity = bdev_nr_sectors(bdev);
|
2016-10-18 06:40:33 +00:00
|
|
|
sector_t end_sector = sector + nr_sectors;
|
2018-10-12 10:08:47 +00:00
|
|
|
struct bio *bio = NULL;
|
2021-05-25 21:24:51 +00:00
|
|
|
int ret = 0;
|
2016-10-18 06:40:33 +00:00
|
|
|
|
2022-07-06 07:03:37 +00:00
|
|
|
if (!bdev_is_zoned(bdev))
|
2016-10-18 06:40:33 +00:00
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
2018-10-12 10:08:47 +00:00
|
|
|
if (bdev_read_only(bdev))
|
|
|
|
return -EPERM;
|
|
|
|
|
2019-10-27 14:05:45 +00:00
|
|
|
if (!op_is_zone_mgmt(op))
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
2020-02-12 17:40:27 +00:00
|
|
|
if (end_sector <= sector || end_sector > capacity)
|
2016-10-18 06:40:33 +00:00
|
|
|
/* Out of range */
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
/* Check alignment (handle eventual smaller last zone) */
|
2023-01-10 14:36:34 +00:00
|
|
|
if (!bdev_is_zone_start(bdev, sector))
|
2016-10-18 06:40:33 +00:00
|
|
|
return -EINVAL;
|
|
|
|
|
2023-01-10 14:36:34 +00:00
|
|
|
if (!bdev_is_zone_start(bdev, nr_sectors) && end_sector != capacity)
|
2016-10-18 06:40:33 +00:00
|
|
|
return -EINVAL;
|
|
|
|
|
2021-05-25 21:24:51 +00:00
|
|
|
/*
|
2024-07-04 05:28:15 +00:00
|
|
|
* In the case of a zone reset operation over all zones, use
|
|
|
|
* REQ_OP_ZONE_RESET_ALL.
|
2021-05-25 21:24:51 +00:00
|
|
|
*/
|
2024-07-04 05:28:15 +00:00
|
|
|
if (op == REQ_OP_ZONE_RESET && sector == 0 && nr_sectors == capacity)
|
2024-01-29 07:52:20 +00:00
|
|
|
return blkdev_zone_reset_all(bdev);
|
2021-05-25 21:24:51 +00:00
|
|
|
|
2016-10-18 06:40:33 +00:00
|
|
|
while (sector < end_sector) {
|
2024-01-29 07:52:20 +00:00
|
|
|
bio = blk_next_bio(bio, bdev, 0, op | REQ_SYNC, GFP_KERNEL);
|
2019-10-27 14:05:43 +00:00
|
|
|
bio->bi_iter.bi_sector = sector;
|
2016-10-18 06:40:33 +00:00
|
|
|
sector += zone_sectors;
|
|
|
|
|
|
|
|
/* This may take a while, so be nice to others */
|
|
|
|
cond_resched();
|
|
|
|
}
|
|
|
|
|
2018-10-12 10:08:47 +00:00
|
|
|
ret = submit_bio_wait(bio);
|
|
|
|
bio_put(bio);
|
|
|
|
|
|
|
|
return ret;
|
2016-10-18 06:40:33 +00:00
|
|
|
}
|
2019-10-27 14:05:45 +00:00
|
|
|
EXPORT_SYMBOL_GPL(blkdev_zone_mgmt);
|
2016-10-18 06:40:35 +00:00
|
|
|
|
2019-11-11 02:39:30 +00:00
|
|
|
struct zone_report_args {
|
|
|
|
struct blk_zone __user *zones;
|
|
|
|
};
|
|
|
|
|
|
|
|
static int blkdev_copy_zone_to_user(struct blk_zone *zone, unsigned int idx,
|
|
|
|
void *data)
|
|
|
|
{
|
|
|
|
struct zone_report_args *args = data;
|
|
|
|
|
|
|
|
if (copy_to_user(&args->zones[idx], zone, sizeof(struct blk_zone)))
|
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2018-03-08 23:28:50 +00:00
|
|
|
/*
|
2016-10-18 06:40:35 +00:00
|
|
|
* BLKREPORTZONE ioctl processing.
|
|
|
|
* Called from blkdev_ioctl.
|
|
|
|
*/
|
2023-06-08 11:02:54 +00:00
|
|
|
int blkdev_report_zones_ioctl(struct block_device *bdev, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
2016-10-18 06:40:35 +00:00
|
|
|
{
|
|
|
|
void __user *argp = (void __user *)arg;
|
2019-11-11 02:39:30 +00:00
|
|
|
struct zone_report_args args;
|
2016-10-18 06:40:35 +00:00
|
|
|
struct blk_zone_report rep;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
if (!argp)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2022-07-06 07:03:37 +00:00
|
|
|
if (!bdev_is_zoned(bdev))
|
2016-10-18 06:40:35 +00:00
|
|
|
return -ENOTTY;
|
|
|
|
|
|
|
|
if (copy_from_user(&rep, argp, sizeof(struct blk_zone_report)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (!rep.nr_zones)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2019-11-11 02:39:30 +00:00
|
|
|
args.zones = argp + sizeof(struct blk_zone_report);
|
|
|
|
ret = blkdev_report_zones(bdev, rep.sector, rep.nr_zones,
|
|
|
|
blkdev_copy_zone_to_user, &args);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
2016-10-18 06:40:35 +00:00
|
|
|
|
2019-11-11 02:39:30 +00:00
|
|
|
rep.nr_zones = ret;
|
2020-06-29 19:06:37 +00:00
|
|
|
rep.flags = BLK_ZONE_REP_CAPACITY;
|
2019-11-11 02:39:30 +00:00
|
|
|
if (copy_to_user(argp, &rep, sizeof(struct blk_zone_report)))
|
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
2016-10-18 06:40:35 +00:00
|
|
|
}
|
|
|
|
|
2023-06-08 11:02:55 +00:00
|
|
|
static int blkdev_truncate_zone_range(struct block_device *bdev,
|
|
|
|
blk_mode_t mode, const struct blk_zone_range *zrange)
|
2021-03-11 07:25:46 +00:00
|
|
|
{
|
|
|
|
loff_t start, end;
|
|
|
|
|
|
|
|
if (zrange->sector + zrange->nr_sectors <= zrange->sector ||
|
|
|
|
zrange->sector + zrange->nr_sectors > get_capacity(bdev->bd_disk))
|
|
|
|
/* Out of range */
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
start = zrange->sector << SECTOR_SHIFT;
|
|
|
|
end = ((zrange->sector + zrange->nr_sectors) << SECTOR_SHIFT) - 1;
|
|
|
|
|
|
|
|
return truncate_bdev_range(bdev, mode, start, end);
|
|
|
|
}
|
|
|
|
|
2018-03-08 23:28:50 +00:00
|
|
|
/*
|
2019-10-27 14:05:46 +00:00
|
|
|
* BLKRESETZONE, BLKOPENZONE, BLKCLOSEZONE and BLKFINISHZONE ioctl processing.
|
2016-10-18 06:40:35 +00:00
|
|
|
* Called from blkdev_ioctl.
|
|
|
|
*/
|
2023-06-08 11:02:55 +00:00
|
|
|
int blkdev_zone_mgmt_ioctl(struct block_device *bdev, blk_mode_t mode,
|
2019-10-27 14:05:46 +00:00
|
|
|
unsigned int cmd, unsigned long arg)
|
2016-10-18 06:40:35 +00:00
|
|
|
{
|
|
|
|
void __user *argp = (void __user *)arg;
|
|
|
|
struct blk_zone_range zrange;
|
2022-07-14 18:06:27 +00:00
|
|
|
enum req_op op;
|
2021-03-11 07:25:46 +00:00
|
|
|
int ret;
|
2016-10-18 06:40:35 +00:00
|
|
|
|
|
|
|
if (!argp)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2022-07-06 07:03:37 +00:00
|
|
|
if (!bdev_is_zoned(bdev))
|
2016-10-18 06:40:35 +00:00
|
|
|
return -ENOTTY;
|
|
|
|
|
2023-06-08 11:02:55 +00:00
|
|
|
if (!(mode & BLK_OPEN_WRITE))
|
2016-10-18 06:40:35 +00:00
|
|
|
return -EBADF;
|
|
|
|
|
|
|
|
if (copy_from_user(&zrange, argp, sizeof(struct blk_zone_range)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2019-10-27 14:05:46 +00:00
|
|
|
switch (cmd) {
|
|
|
|
case BLKRESETZONE:
|
|
|
|
op = REQ_OP_ZONE_RESET;
|
2021-03-11 07:25:46 +00:00
|
|
|
|
|
|
|
/* Invalidate the page cache, including dirty pages. */
|
2024-04-11 14:53:37 +00:00
|
|
|
filemap_invalidate_lock(bdev->bd_mapping);
|
2021-03-11 07:25:46 +00:00
|
|
|
ret = blkdev_truncate_zone_range(bdev, mode, &zrange);
|
|
|
|
if (ret)
|
2021-11-11 08:52:38 +00:00
|
|
|
goto fail;
|
2019-10-27 14:05:46 +00:00
|
|
|
break;
|
|
|
|
case BLKOPENZONE:
|
|
|
|
op = REQ_OP_ZONE_OPEN;
|
|
|
|
break;
|
|
|
|
case BLKCLOSEZONE:
|
|
|
|
op = REQ_OP_ZONE_CLOSE;
|
|
|
|
break;
|
|
|
|
case BLKFINISHZONE:
|
|
|
|
op = REQ_OP_ZONE_FINISH;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
return -ENOTTY;
|
|
|
|
}
|
|
|
|
|
2024-01-29 07:52:20 +00:00
|
|
|
ret = blkdev_zone_mgmt(bdev, op, zrange.sector, zrange.nr_sectors);
|
2021-03-11 07:25:46 +00:00
|
|
|
|
2021-11-11 08:52:38 +00:00
|
|
|
fail:
|
|
|
|
if (cmd == BLKRESETZONE)
|
2024-04-11 14:53:37 +00:00
|
|
|
filemap_invalidate_unlock(bdev->bd_mapping);
|
2021-03-11 07:25:46 +00:00
|
|
|
|
|
|
|
return ret;
|
2016-10-18 06:40:35 +00:00
|
|
|
}
|
2018-10-12 10:08:50 +00:00
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
static inline bool disk_zone_is_conv(struct gendisk *disk, sector_t sector)
|
|
|
|
{
|
|
|
|
if (!disk->conv_zones_bitmap)
|
|
|
|
return false;
|
|
|
|
return test_bit(disk_zone_no(disk, sector), disk->conv_zones_bitmap);
|
|
|
|
}
|
|
|
|
|
2024-05-30 05:40:33 +00:00
|
|
|
static bool disk_zone_is_last(struct gendisk *disk, struct blk_zone *zone)
|
|
|
|
{
|
|
|
|
return zone->start + zone->len >= get_capacity(disk);
|
|
|
|
}
|
|
|
|
|
2024-05-30 05:40:34 +00:00
|
|
|
static bool disk_zone_is_full(struct gendisk *disk,
|
|
|
|
unsigned int zno, unsigned int offset_in_zone)
|
|
|
|
{
|
|
|
|
if (zno < disk->nr_zones - 1)
|
|
|
|
return offset_in_zone >= disk->zone_capacity;
|
|
|
|
return offset_in_zone >= disk->last_zone_capacity;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool disk_zone_wplug_is_full(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
return disk_zone_is_full(disk, zwplug->zone_no, zwplug->wp_offset);
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
static bool disk_insert_zone_wplug(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
struct blk_zone_wplug *zwplg;
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned int idx =
|
|
|
|
hash_32(zwplug->zone_no, disk->zone_wplugs_hash_bits);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Add the new zone write plug to the hash table, but carefully as we
|
|
|
|
* are racing with other submission context, so we may already have a
|
|
|
|
* zone write plug for the same zone.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
|
|
|
|
hlist_for_each_entry_rcu(zwplg, &disk->zone_wplugs_hash[idx], node) {
|
|
|
|
if (zwplg->zone_no == zwplug->zone_no) {
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
hlist_add_head_rcu(&zwplug->node, &disk->zone_wplugs_hash[idx]);
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct blk_zone_wplug *disk_get_zone_wplug(struct gendisk *disk,
|
|
|
|
sector_t sector)
|
|
|
|
{
|
|
|
|
unsigned int zno = disk_zone_no(disk, sector);
|
|
|
|
unsigned int idx = hash_32(zno, disk->zone_wplugs_hash_bits);
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
|
|
|
|
hlist_for_each_entry_rcu(zwplug, &disk->zone_wplugs_hash[idx], node) {
|
|
|
|
if (zwplug->zone_no == zno &&
|
|
|
|
atomic_inc_not_zero(&zwplug->ref)) {
|
|
|
|
rcu_read_unlock();
|
|
|
|
return zwplug;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void disk_free_zone_wplug_rcu(struct rcu_head *rcu_head)
|
|
|
|
{
|
|
|
|
struct blk_zone_wplug *zwplug =
|
|
|
|
container_of(rcu_head, struct blk_zone_wplug, rcu_head);
|
|
|
|
|
|
|
|
mempool_free(zwplug, zwplug->disk->zone_wplugs_pool);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void disk_put_zone_wplug(struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
if (atomic_dec_and_test(&zwplug->ref)) {
|
|
|
|
WARN_ON_ONCE(!bio_list_empty(&zwplug->bio_list));
|
|
|
|
WARN_ON_ONCE(!list_empty(&zwplug->link));
|
2024-05-01 11:08:59 +00:00
|
|
|
WARN_ON_ONCE(!(zwplug->flags & BLK_ZONE_WPLUG_UNHASHED));
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
|
|
|
call_rcu(&zwplug->rcu_head, disk_free_zone_wplug_rcu);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-05-01 11:08:59 +00:00
|
|
|
static inline bool disk_should_remove_zone_wplug(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
2024-05-01 11:09:00 +00:00
|
|
|
/* If the zone write plug was already removed, we are done. */
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_UNHASHED)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* If the zone write plug is still busy, it cannot be removed. */
|
2024-05-01 11:08:59 +00:00
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_BUSY)
|
|
|
|
return false;
|
|
|
|
|
2024-05-01 11:09:00 +00:00
|
|
|
/*
|
|
|
|
* Completions of BIOs with blk_zone_write_plug_bio_endio() may
|
|
|
|
* happen after handling a request completion with
|
2024-05-01 11:09:04 +00:00
|
|
|
* blk_zone_write_plug_finish_request() (e.g. with split BIOs
|
2024-05-01 11:09:00 +00:00
|
|
|
* that are chained). In such case, disk_zone_wplug_unplug_bio()
|
|
|
|
* should not attempt to remove the zone write plug until all BIO
|
|
|
|
* completions are seen. Check by looking at the zone write plug
|
|
|
|
* reference count, which is 2 when the plug is unused (one reference
|
|
|
|
* taken when the plug was allocated and another reference taken by the
|
|
|
|
* caller context).
|
|
|
|
*/
|
|
|
|
if (atomic_read(&zwplug->ref) > 2)
|
|
|
|
return false;
|
|
|
|
|
2024-05-01 11:08:59 +00:00
|
|
|
/* We can remove zone write plugs for zones that are empty or full. */
|
2024-05-30 05:40:34 +00:00
|
|
|
return !zwplug->wp_offset || disk_zone_wplug_is_full(disk, zwplug);
|
2024-05-01 11:08:59 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void disk_remove_zone_wplug(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
/* If the zone write plug was already removed, we have nothing to do. */
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_UNHASHED)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Mark the zone write plug as unhashed and drop the extra reference we
|
|
|
|
* took when the plug was inserted in the hash table.
|
|
|
|
*/
|
|
|
|
zwplug->flags |= BLK_ZONE_WPLUG_UNHASHED;
|
|
|
|
spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
|
|
|
|
hlist_del_init_rcu(&zwplug->node);
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
static void blk_zone_wplug_bio_work(struct work_struct *work);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Get a reference on the write plug for the zone containing @sector.
|
|
|
|
* If the plug does not exist, it is allocated and hashed.
|
|
|
|
* Return a pointer to the zone write plug with the plug spinlock held.
|
|
|
|
*/
|
|
|
|
static struct blk_zone_wplug *disk_get_and_lock_zone_wplug(struct gendisk *disk,
|
|
|
|
sector_t sector, gfp_t gfp_mask,
|
|
|
|
unsigned long *flags)
|
2018-10-12 10:08:50 +00:00
|
|
|
{
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
unsigned int zno = disk_zone_no(disk, sector);
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
|
|
|
|
again:
|
|
|
|
zwplug = disk_get_zone_wplug(disk, sector);
|
|
|
|
if (zwplug) {
|
|
|
|
/*
|
|
|
|
* Check that a BIO completion or a zone reset or finish
|
|
|
|
* operation has not already removed the zone write plug from
|
|
|
|
* the hash table and dropped its reference count. In such case,
|
|
|
|
* we need to get a new plug so start over from the beginning.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&zwplug->lock, *flags);
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_UNHASHED) {
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, *flags);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
goto again;
|
|
|
|
}
|
|
|
|
return zwplug;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate and initialize a zone write plug with an extra reference
|
|
|
|
* so that it is not freed when the zone write plug becomes idle without
|
|
|
|
* the zone being full.
|
|
|
|
*/
|
|
|
|
zwplug = mempool_alloc(disk->zone_wplugs_pool, gfp_mask);
|
|
|
|
if (!zwplug)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
INIT_HLIST_NODE(&zwplug->node);
|
|
|
|
INIT_LIST_HEAD(&zwplug->link);
|
|
|
|
atomic_set(&zwplug->ref, 2);
|
|
|
|
spin_lock_init(&zwplug->lock);
|
|
|
|
zwplug->flags = 0;
|
|
|
|
zwplug->zone_no = zno;
|
|
|
|
zwplug->wp_offset = sector & (disk->queue->limits.chunk_sectors - 1);
|
|
|
|
bio_list_init(&zwplug->bio_list);
|
|
|
|
INIT_WORK(&zwplug->bio_work, blk_zone_wplug_bio_work);
|
|
|
|
zwplug->disk = disk;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&zwplug->lock, *flags);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Insert the new zone write plug in the hash table. This can fail only
|
|
|
|
* if another context already inserted a plug. Retry from the beginning
|
|
|
|
* in such case.
|
|
|
|
*/
|
|
|
|
if (!disk_insert_zone_wplug(disk, zwplug)) {
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, *flags);
|
|
|
|
mempool_free(zwplug, disk->zone_wplugs_pool);
|
|
|
|
goto again;
|
|
|
|
}
|
|
|
|
|
|
|
|
return zwplug;
|
|
|
|
}
|
|
|
|
|
2024-05-01 11:09:06 +00:00
|
|
|
static inline void blk_zone_wplug_bio_io_error(struct blk_zone_wplug *zwplug,
|
|
|
|
struct bio *bio)
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
{
|
2024-05-01 11:09:06 +00:00
|
|
|
struct request_queue *q = zwplug->disk->queue;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
|
|
|
bio_clear_flag(bio, BIO_ZONE_WRITE_PLUGGING);
|
|
|
|
bio_io_error(bio);
|
2024-05-01 11:09:06 +00:00
|
|
|
disk_put_zone_wplug(zwplug);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
blk_queue_exit(q);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Abort (fail) all plugged BIOs of a zone write plug.
|
|
|
|
*/
|
|
|
|
static void disk_zone_wplug_abort(struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
struct bio *bio;
|
|
|
|
|
2024-05-01 11:09:06 +00:00
|
|
|
while ((bio = bio_list_pop(&zwplug->bio_list)))
|
|
|
|
blk_zone_wplug_bio_io_error(zwplug, bio);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Abort (fail) all plugged BIOs of a zone write plug that are not aligned
|
|
|
|
* with the assumed write pointer location of the zone when the BIO will
|
|
|
|
* be unplugged.
|
|
|
|
*/
|
|
|
|
static void disk_zone_wplug_abort_unaligned(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
unsigned int wp_offset = zwplug->wp_offset;
|
|
|
|
struct bio_list bl = BIO_EMPTY_LIST;
|
|
|
|
struct bio *bio;
|
|
|
|
|
|
|
|
while ((bio = bio_list_pop(&zwplug->bio_list))) {
|
2024-05-30 05:40:34 +00:00
|
|
|
if (disk_zone_is_full(disk, zwplug->zone_no, wp_offset) ||
|
2024-04-08 01:41:10 +00:00
|
|
|
(bio_op(bio) != REQ_OP_ZONE_APPEND &&
|
|
|
|
bio_offset_from_zone_start(bio) != wp_offset)) {
|
2024-05-01 11:09:06 +00:00
|
|
|
blk_zone_wplug_bio_io_error(zwplug, bio);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
wp_offset += bio_sectors(bio);
|
|
|
|
bio_list_add(&bl, bio);
|
|
|
|
}
|
|
|
|
|
|
|
|
bio_list_merge(&zwplug->bio_list, &bl);
|
|
|
|
}
|
|
|
|
|
2024-05-01 11:08:57 +00:00
|
|
|
static inline void disk_zone_wplug_set_error(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_ERROR)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* At this point, we already have a reference on the zone write plug.
|
|
|
|
* However, since we are going to add the plug to the disk zone write
|
|
|
|
* plugs work list, increase its reference count. This reference will
|
|
|
|
* be dropped in disk_zone_wplugs_work() once the error state is
|
|
|
|
* handled, or in disk_zone_wplug_clear_error() if the zone is reset or
|
|
|
|
* finished.
|
|
|
|
*/
|
|
|
|
zwplug->flags |= BLK_ZONE_WPLUG_ERROR;
|
|
|
|
atomic_inc(&zwplug->ref);
|
|
|
|
|
|
|
|
spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
|
|
|
|
list_add_tail(&zwplug->link, &disk->zone_wplugs_err_list);
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void disk_zone_wplug_clear_error(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
if (!(zwplug->flags & BLK_ZONE_WPLUG_ERROR))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We are racing with the error handling work which drops the reference
|
|
|
|
* on the zone write plug after handling the error state. So remove the
|
|
|
|
* plug from the error list and drop its reference count only if the
|
|
|
|
* error handling has not yet started, that is, if the zone write plug
|
|
|
|
* is still listed.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
|
|
|
|
if (!list_empty(&zwplug->link)) {
|
|
|
|
list_del_init(&zwplug->link);
|
|
|
|
zwplug->flags &= ~BLK_ZONE_WPLUG_ERROR;
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
/*
|
|
|
|
* Set a zone write plug write pointer offset to either 0 (zone reset case)
|
|
|
|
* or to the zone size (zone finish case). This aborts all plugged BIOs, which
|
|
|
|
* is fine to do as doing a zone reset or zone finish while writes are in-flight
|
|
|
|
* is a mistake from the user which will most likely cause all plugged BIOs to
|
|
|
|
* fail anyway.
|
|
|
|
*/
|
|
|
|
static void disk_zone_wplug_set_wp_offset(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug,
|
|
|
|
unsigned int wp_offset)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Make sure that a BIO completion or another zone reset or finish
|
|
|
|
* operation has not already removed the plug from the hash table.
|
|
|
|
*/
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_UNHASHED) {
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Update the zone write pointer and abort all plugged BIOs. */
|
|
|
|
zwplug->wp_offset = wp_offset;
|
|
|
|
disk_zone_wplug_abort(zwplug);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Updating the write pointer offset puts back the zone
|
|
|
|
* in a good state. So clear the error flag and decrement the
|
|
|
|
* error count if we were in error state.
|
|
|
|
*/
|
2024-05-01 11:08:57 +00:00
|
|
|
disk_zone_wplug_clear_error(disk, zwplug);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The zone write plug now has no BIO plugged: remove it from the
|
|
|
|
* hash table so that it cannot be seen. The plug will be freed
|
|
|
|
* when the last reference is dropped.
|
|
|
|
*/
|
|
|
|
if (disk_should_remove_zone_wplug(disk, zwplug))
|
|
|
|
disk_remove_zone_wplug(disk, zwplug);
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool blk_zone_wplug_handle_reset_or_finish(struct bio *bio,
|
|
|
|
unsigned int wp_offset)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = bio->bi_bdev->bd_disk;
|
|
|
|
sector_t sector = bio->bi_iter.bi_sector;
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
|
|
|
|
/* Conventional zones cannot be reset nor finished. */
|
|
|
|
if (disk_zone_is_conv(disk, sector)) {
|
|
|
|
bio_io_error(bio);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we have a zone write plug, set its write pointer offset to 0
|
|
|
|
* (reset case) or to the zone size (finish case). This will abort all
|
|
|
|
* BIOs plugged for the target zone. It is fine as resetting or
|
|
|
|
* finishing zones while writes are still in-flight will result in the
|
|
|
|
* writes failing anyway.
|
|
|
|
*/
|
|
|
|
zwplug = disk_get_zone_wplug(disk, sector);
|
|
|
|
if (zwplug) {
|
|
|
|
disk_zone_wplug_set_wp_offset(disk, zwplug, wp_offset);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool blk_zone_wplug_handle_reset_all(struct bio *bio)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = bio->bi_bdev->bd_disk;
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
sector_t sector;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Set the write pointer offset of all zone write plugs to 0. This will
|
|
|
|
* abort all plugged BIOs. It is fine as resetting zones while writes
|
|
|
|
* are still in-flight will result in the writes failing anyway.
|
|
|
|
*/
|
|
|
|
for (sector = 0; sector < get_capacity(disk);
|
|
|
|
sector += disk->queue->limits.chunk_sectors) {
|
|
|
|
zwplug = disk_get_zone_wplug(disk, sector);
|
|
|
|
if (zwplug) {
|
|
|
|
disk_zone_wplug_set_wp_offset(disk, zwplug, 0);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void blk_zone_wplug_add_bio(struct blk_zone_wplug *zwplug,
|
|
|
|
struct bio *bio, unsigned int nr_segs)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Grab an extra reference on the BIO request queue usage counter.
|
|
|
|
* This reference will be reused to submit a request for the BIO for
|
|
|
|
* blk-mq devices and dropped when the BIO is failed and after
|
|
|
|
* it is issued in the case of BIO-based devices.
|
|
|
|
*/
|
|
|
|
percpu_ref_get(&bio->bi_bdev->bd_disk->queue->q_usage_counter);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The BIO is being plugged and thus will have to wait for the on-going
|
|
|
|
* write and for all other writes already plugged. So polling makes
|
|
|
|
* no sense.
|
|
|
|
*/
|
|
|
|
bio_clear_polled(bio);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Reuse the poll cookie field to store the number of segments when
|
|
|
|
* split to the hardware limits.
|
|
|
|
*/
|
|
|
|
bio->__bi_nr_segments = nr_segs;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We always receive BIOs after they are split and ready to be issued.
|
|
|
|
* The block layer passes the parts of a split BIO in order, and the
|
|
|
|
* user must also issue write sequentially. So simply add the new BIO
|
|
|
|
* at the tail of the list to preserve the sequential write order.
|
|
|
|
*/
|
|
|
|
bio_list_add(&zwplug->bio_list, bio);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Called from bio_attempt_back_merge() when a BIO was merged with a request.
|
|
|
|
*/
|
|
|
|
void blk_zone_write_plug_bio_merged(struct bio *bio)
|
|
|
|
{
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the BIO was already plugged, then we were called through
|
2024-05-01 11:09:02 +00:00
|
|
|
* blk_zone_write_plug_init_request() -> blk_attempt_bio_merge().
|
|
|
|
* For this case, we already hold a reference on the zone write plug for
|
|
|
|
* the BIO and blk_zone_write_plug_init_request() will handle the
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
* zone write pointer offset update.
|
|
|
|
*/
|
|
|
|
if (bio_flagged(bio, BIO_ZONE_WRITE_PLUGGING))
|
|
|
|
return;
|
|
|
|
|
|
|
|
bio_set_flag(bio, BIO_ZONE_WRITE_PLUGGING);
|
|
|
|
|
|
|
|
/*
|
2024-05-01 11:09:03 +00:00
|
|
|
* Get a reference on the zone write plug of the target zone and advance
|
|
|
|
* the zone write pointer offset. Given that this is a merge, we already
|
|
|
|
* have at least one request and one BIO referencing the zone write
|
|
|
|
* plug. So this should not fail.
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
*/
|
|
|
|
zwplug = disk_get_zone_wplug(bio->bi_bdev->bd_disk,
|
|
|
|
bio->bi_iter.bi_sector);
|
2024-05-01 11:09:03 +00:00
|
|
|
if (WARN_ON_ONCE(!zwplug))
|
|
|
|
return;
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
zwplug->wp_offset += bio_sectors(bio);
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Attempt to merge plugged BIOs with a newly prepared request for a BIO that
|
|
|
|
* already went through zone write plugging (either a new BIO or one that was
|
|
|
|
* unplugged).
|
|
|
|
*/
|
2024-05-01 11:09:02 +00:00
|
|
|
void blk_zone_write_plug_init_request(struct request *req)
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
{
|
|
|
|
sector_t req_back_sector = blk_rq_pos(req) + blk_rq_sectors(req);
|
|
|
|
struct request_queue *q = req->q;
|
|
|
|
struct gendisk *disk = q->disk;
|
|
|
|
struct blk_zone_wplug *zwplug =
|
|
|
|
disk_get_zone_wplug(disk, blk_rq_pos(req));
|
|
|
|
unsigned long flags;
|
|
|
|
struct bio *bio;
|
|
|
|
|
2024-05-01 11:09:02 +00:00
|
|
|
if (WARN_ON_ONCE(!zwplug))
|
|
|
|
return;
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
/*
|
2024-05-01 11:09:00 +00:00
|
|
|
* Indicate that completion of this request needs to be handled with
|
2024-05-01 11:09:04 +00:00
|
|
|
* blk_zone_write_plug_finish_request(), which will drop the reference
|
2024-05-01 11:09:00 +00:00
|
|
|
* on the zone write plug we took above on entry to this function.
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
*/
|
|
|
|
req->rq_flags |= RQF_ZONE_WRITE_PLUGGING;
|
|
|
|
|
|
|
|
if (blk_queue_nomerges(q))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Walk through the list of plugged BIOs to check if they can be merged
|
|
|
|
* into the back of the request.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
2024-05-30 05:40:34 +00:00
|
|
|
while (!disk_zone_wplug_is_full(disk, zwplug)) {
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
bio = bio_list_peek(&zwplug->bio_list);
|
|
|
|
if (!bio)
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (bio->bi_iter.bi_sector != req_back_sector ||
|
|
|
|
!blk_rq_merge_ok(req, bio))
|
|
|
|
break;
|
|
|
|
|
|
|
|
WARN_ON_ONCE(bio_op(bio) != REQ_OP_WRITE_ZEROES &&
|
|
|
|
!bio->__bi_nr_segments);
|
|
|
|
|
|
|
|
bio_list_pop(&zwplug->bio_list);
|
|
|
|
if (bio_attempt_back_merge(req, bio, bio->__bi_nr_segments) !=
|
|
|
|
BIO_MERGE_OK) {
|
|
|
|
bio_list_add_head(&zwplug->bio_list, bio);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Drop the extra reference on the queue usage we got when
|
|
|
|
* plugging the BIO and advance the write pointer offset.
|
|
|
|
*/
|
|
|
|
blk_queue_exit(q);
|
|
|
|
zwplug->wp_offset += bio_sectors(bio);
|
|
|
|
|
|
|
|
req_back_sector += bio_sectors(bio);
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check and prepare a BIO for submission by incrementing the write pointer
|
2024-04-08 01:41:10 +00:00
|
|
|
* offset of its zone write plug and changing zone append operations into
|
|
|
|
* regular write when zone append emulation is needed.
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
*/
|
|
|
|
static bool blk_zone_wplug_prepare_bio(struct blk_zone_wplug *zwplug,
|
|
|
|
struct bio *bio)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = bio->bi_bdev->bd_disk;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check that the user is not attempting to write to a full zone.
|
|
|
|
* We know such BIO will fail, and that would potentially overflow our
|
|
|
|
* write pointer offset beyond the end of the zone.
|
|
|
|
*/
|
2024-05-30 05:40:34 +00:00
|
|
|
if (disk_zone_wplug_is_full(disk, zwplug))
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
goto err;
|
|
|
|
|
2024-04-08 01:41:10 +00:00
|
|
|
if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
|
|
|
|
/*
|
|
|
|
* Use a regular write starting at the current write pointer.
|
|
|
|
* Similarly to native zone append operations, do not allow
|
|
|
|
* merging.
|
|
|
|
*/
|
|
|
|
bio->bi_opf &= ~REQ_OP_MASK;
|
|
|
|
bio->bi_opf |= REQ_OP_WRITE | REQ_NOMERGE;
|
|
|
|
bio->bi_iter.bi_sector += zwplug->wp_offset;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Remember that this BIO is in fact a zone append operation
|
|
|
|
* so that we can restore its operation code on completion.
|
|
|
|
*/
|
|
|
|
bio_set_flag(bio, BIO_EMULATES_ZONE_APPEND);
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* Check for non-sequential writes early because we avoid a
|
|
|
|
* whole lot of error handling trouble if we don't send it off
|
|
|
|
* to the driver.
|
|
|
|
*/
|
|
|
|
if (bio_offset_from_zone_start(bio) != zwplug->wp_offset)
|
|
|
|
goto err;
|
|
|
|
}
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
|
|
|
/* Advance the zone write pointer offset. */
|
|
|
|
zwplug->wp_offset += bio_sectors(bio);
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
|
|
|
err:
|
|
|
|
/* We detected an invalid write BIO: schedule error recovery. */
|
|
|
|
disk_zone_wplug_set_error(disk, zwplug);
|
|
|
|
kblockd_schedule_work(&disk->zone_wplugs_work);
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool blk_zone_wplug_handle_write(struct bio *bio, unsigned int nr_segs)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = bio->bi_bdev->bd_disk;
|
|
|
|
sector_t sector = bio->bi_iter.bi_sector;
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
gfp_t gfp_mask = GFP_NOIO;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* BIOs must be fully contained within a zone so that we use the correct
|
|
|
|
* zone write plug for the entire BIO. For blk-mq devices, the block
|
|
|
|
* layer should already have done any splitting required to ensure this
|
|
|
|
* and this BIO should thus not be straddling zone boundaries. For
|
|
|
|
* BIO-based devices, it is the responsibility of the driver to split
|
|
|
|
* the bio before submitting it.
|
|
|
|
*/
|
|
|
|
if (WARN_ON_ONCE(bio_straddles_zones(bio))) {
|
|
|
|
bio_io_error(bio);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Conventional zones do not need write plugging. */
|
2024-04-08 01:41:10 +00:00
|
|
|
if (disk_zone_is_conv(disk, sector)) {
|
|
|
|
/* Zone append to conventional zones is not allowed. */
|
|
|
|
if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
|
|
|
|
bio_io_error(bio);
|
|
|
|
return true;
|
|
|
|
}
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
return false;
|
2024-04-08 01:41:10 +00:00
|
|
|
}
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
|
|
|
if (bio->bi_opf & REQ_NOWAIT)
|
|
|
|
gfp_mask = GFP_NOWAIT;
|
|
|
|
|
|
|
|
zwplug = disk_get_and_lock_zone_wplug(disk, sector, gfp_mask, &flags);
|
|
|
|
if (!zwplug) {
|
|
|
|
bio_io_error(bio);
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Indicate that this BIO is being handled using zone write plugging. */
|
|
|
|
bio_set_flag(bio, BIO_ZONE_WRITE_PLUGGING);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the zone is already plugged or has a pending error, add the BIO
|
|
|
|
* to the plug BIO list. Otherwise, plug and let the BIO execute.
|
|
|
|
*/
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_BUSY)
|
|
|
|
goto plug;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If an error is detected when preparing the BIO, add it to the BIO
|
|
|
|
* list so that error recovery can deal with it.
|
|
|
|
*/
|
|
|
|
if (!blk_zone_wplug_prepare_bio(zwplug, bio))
|
|
|
|
goto plug;
|
|
|
|
|
|
|
|
zwplug->flags |= BLK_ZONE_WPLUG_PLUGGED;
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
return false;
|
|
|
|
|
|
|
|
plug:
|
|
|
|
zwplug->flags |= BLK_ZONE_WPLUG_PLUGGED;
|
|
|
|
blk_zone_wplug_add_bio(zwplug, bio, nr_segs);
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* blk_zone_plug_bio - Handle a zone write BIO with zone write plugging
|
|
|
|
* @bio: The BIO being submitted
|
|
|
|
* @nr_segs: The number of physical segments of @bio
|
|
|
|
*
|
2024-04-08 01:41:10 +00:00
|
|
|
* Handle write, write zeroes and zone append operations requiring emulation
|
|
|
|
* using zone write plugging.
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
*
|
|
|
|
* Return true whenever @bio execution needs to be delayed through the zone
|
|
|
|
* write plug. Otherwise, return false to let the submission path process
|
|
|
|
* @bio normally.
|
|
|
|
*/
|
|
|
|
bool blk_zone_plug_bio(struct bio *bio, unsigned int nr_segs)
|
|
|
|
{
|
|
|
|
struct block_device *bdev = bio->bi_bdev;
|
|
|
|
|
|
|
|
if (!bdev->bd_disk->zone_wplugs_hash)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the BIO already has the plugging flag set, then it was already
|
|
|
|
* handled through this path and this is a submission from the zone
|
|
|
|
* plug bio submit work.
|
|
|
|
*/
|
|
|
|
if (bio_flagged(bio, BIO_ZONE_WRITE_PLUGGING))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We do not need to do anything special for empty flush BIOs, e.g
|
|
|
|
* BIOs such as issued by blkdev_issue_flush(). The is because it is
|
|
|
|
* the responsibility of the user to first wait for the completion of
|
|
|
|
* write operations for flush to have any effect on the persistence of
|
|
|
|
* the written data.
|
|
|
|
*/
|
|
|
|
if (op_is_flush(bio->bi_opf) && !bio_sectors(bio))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Regular writes and write zeroes need to be handled through the target
|
|
|
|
* zone write plug. This includes writes with REQ_FUA | REQ_PREFLUSH
|
|
|
|
* which may need to go through the flush machinery depending on the
|
|
|
|
* target device capabilities. Plugging such writes is fine as the flush
|
|
|
|
* machinery operates at the request level, below the plug, and
|
|
|
|
* completion of the flush sequence will go through the regular BIO
|
|
|
|
* completion, which will handle zone write plugging.
|
2024-04-08 01:41:10 +00:00
|
|
|
* Zone append operations for devices that requested emulation must
|
|
|
|
* also be plugged so that these BIOs can be changed into regular
|
|
|
|
* write BIOs.
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
* Zone reset, reset all and finish commands need special treatment
|
|
|
|
* to correctly track the write pointer offset of zones. These commands
|
|
|
|
* are not plugged as we do not need serialization with write
|
|
|
|
* operations. It is the responsibility of the user to not issue reset
|
|
|
|
* and finish commands when write operations are in flight.
|
|
|
|
*/
|
|
|
|
switch (bio_op(bio)) {
|
2024-04-08 01:41:10 +00:00
|
|
|
case REQ_OP_ZONE_APPEND:
|
|
|
|
if (!bdev_emulates_zone_append(bdev))
|
|
|
|
return false;
|
|
|
|
fallthrough;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
case REQ_OP_WRITE:
|
|
|
|
case REQ_OP_WRITE_ZEROES:
|
|
|
|
return blk_zone_wplug_handle_write(bio, nr_segs);
|
|
|
|
case REQ_OP_ZONE_RESET:
|
|
|
|
return blk_zone_wplug_handle_reset_or_finish(bio, 0);
|
|
|
|
case REQ_OP_ZONE_FINISH:
|
|
|
|
return blk_zone_wplug_handle_reset_or_finish(bio,
|
|
|
|
bdev_zone_sectors(bdev));
|
|
|
|
case REQ_OP_ZONE_RESET_ALL:
|
|
|
|
return blk_zone_wplug_handle_reset_all(bio);
|
|
|
|
default:
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blk_zone_plug_bio);
|
|
|
|
|
2024-05-01 11:08:58 +00:00
|
|
|
static void disk_zone_wplug_schedule_bio_work(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Take a reference on the zone write plug and schedule the submission
|
|
|
|
* of the next plugged BIO. blk_zone_wplug_bio_work() will release the
|
|
|
|
* reference we take here.
|
|
|
|
*/
|
|
|
|
WARN_ON_ONCE(!(zwplug->flags & BLK_ZONE_WPLUG_PLUGGED));
|
|
|
|
atomic_inc(&zwplug->ref);
|
|
|
|
queue_work(disk->zone_wplugs_wq, &zwplug->bio_work);
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
static void disk_zone_wplug_unplug_bio(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we had an error, schedule error recovery. The recovery work
|
|
|
|
* will restart submission of plugged BIOs.
|
|
|
|
*/
|
|
|
|
if (zwplug->flags & BLK_ZONE_WPLUG_ERROR) {
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
kblockd_schedule_work(&disk->zone_wplugs_work);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Schedule submission of the next plugged BIO if we have one. */
|
|
|
|
if (!bio_list_empty(&zwplug->bio_list)) {
|
2024-05-01 11:08:58 +00:00
|
|
|
disk_zone_wplug_schedule_bio_work(disk, zwplug);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
zwplug->flags &= ~BLK_ZONE_WPLUG_PLUGGED;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the zone is full (it was fully written or finished, or empty
|
|
|
|
* (it was reset), remove its zone write plug from the hash table.
|
|
|
|
*/
|
|
|
|
if (disk_should_remove_zone_wplug(disk, zwplug))
|
|
|
|
disk_remove_zone_wplug(disk, zwplug);
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
void blk_zone_write_plug_bio_endio(struct bio *bio)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = bio->bi_bdev->bd_disk;
|
|
|
|
struct blk_zone_wplug *zwplug =
|
2024-05-01 11:09:05 +00:00
|
|
|
disk_get_zone_wplug(disk, bio->bi_iter.bi_sector);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
if (WARN_ON_ONCE(!zwplug))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Make sure we do not see this BIO again by clearing the plug flag. */
|
|
|
|
bio_clear_flag(bio, BIO_ZONE_WRITE_PLUGGING);
|
|
|
|
|
2024-04-08 01:41:10 +00:00
|
|
|
/*
|
|
|
|
* If this is a regular write emulating a zone append operation,
|
|
|
|
* restore the original operation code.
|
|
|
|
*/
|
|
|
|
if (bio_flagged(bio, BIO_EMULATES_ZONE_APPEND)) {
|
|
|
|
bio->bi_opf &= ~REQ_OP_MASK;
|
|
|
|
bio->bi_opf |= REQ_OP_ZONE_APPEND;
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
/*
|
|
|
|
* If the BIO failed, mark the plug as having an error to trigger
|
|
|
|
* recovery.
|
|
|
|
*/
|
|
|
|
if (bio->bi_status != BLK_STS_OK) {
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
disk_zone_wplug_set_error(disk, zwplug);
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
}
|
|
|
|
|
2024-05-01 11:09:00 +00:00
|
|
|
/* Drop the reference we took when the BIO was issued. */
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
/*
|
2024-05-01 11:09:04 +00:00
|
|
|
* For BIO-based devices, blk_zone_write_plug_finish_request()
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
* is not called. So we need to schedule execution of the next
|
|
|
|
* plugged BIO here.
|
|
|
|
*/
|
2024-05-21 20:02:56 +00:00
|
|
|
if (bdev_test_flag(bio->bi_bdev, BD_HAS_SUBMIT_BIO))
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
disk_zone_wplug_unplug_bio(disk, zwplug);
|
|
|
|
|
2024-05-01 11:09:00 +00:00
|
|
|
/* Drop the reference we took when entering this function. */
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
|
2024-05-01 11:09:04 +00:00
|
|
|
void blk_zone_write_plug_finish_request(struct request *req)
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
{
|
|
|
|
struct gendisk *disk = req->q->disk;
|
2024-05-01 11:09:04 +00:00
|
|
|
struct blk_zone_wplug *zwplug;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
2024-05-01 11:09:04 +00:00
|
|
|
zwplug = disk_get_zone_wplug(disk, req->__sector);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
if (WARN_ON_ONCE(!zwplug))
|
|
|
|
return;
|
|
|
|
|
|
|
|
req->rq_flags &= ~RQF_ZONE_WRITE_PLUGGING;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Drop the reference we took when the request was initialized in
|
2024-05-01 11:09:02 +00:00
|
|
|
* blk_zone_write_plug_init_request().
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
*/
|
2024-05-01 11:09:00 +00:00
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
|
|
|
|
disk_zone_wplug_unplug_bio(disk, zwplug);
|
|
|
|
|
|
|
|
/* Drop the reference we took when entering this function. */
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void blk_zone_wplug_bio_work(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct blk_zone_wplug *zwplug =
|
|
|
|
container_of(work, struct blk_zone_wplug, bio_work);
|
|
|
|
struct block_device *bdev;
|
|
|
|
unsigned long flags;
|
|
|
|
struct bio *bio;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Submit the next plugged BIO. If we do not have any, clear
|
|
|
|
* the plugged flag.
|
|
|
|
*/
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
bio = bio_list_pop(&zwplug->bio_list);
|
|
|
|
if (!bio) {
|
|
|
|
zwplug->flags &= ~BLK_ZONE_WPLUG_PLUGGED;
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
2024-05-01 11:08:58 +00:00
|
|
|
goto put_zwplug;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!blk_zone_wplug_prepare_bio(zwplug, bio)) {
|
|
|
|
/* Error recovery will decide what to do with the BIO. */
|
|
|
|
bio_list_add_head(&zwplug->bio_list, bio);
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
2024-05-01 11:08:58 +00:00
|
|
|
goto put_zwplug;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
bdev = bio->bi_bdev;
|
|
|
|
submit_bio_noacct_nocheck(bio);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* blk-mq devices will reuse the extra reference on the request queue
|
|
|
|
* usage counter we took when the BIO was plugged, but the submission
|
|
|
|
* path for BIO-based devices will not do that. So drop this extra
|
|
|
|
* reference here.
|
|
|
|
*/
|
2024-05-21 20:02:56 +00:00
|
|
|
if (bdev_test_flag(bdev, BD_HAS_SUBMIT_BIO))
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
blk_queue_exit(bdev->bd_disk->queue);
|
2024-05-01 11:08:58 +00:00
|
|
|
|
|
|
|
put_zwplug:
|
|
|
|
/* Drop the reference we took in disk_zone_wplug_schedule_bio_work(). */
|
|
|
|
disk_put_zone_wplug(zwplug);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned int blk_zone_wp_offset(struct blk_zone *zone)
|
|
|
|
{
|
|
|
|
switch (zone->cond) {
|
|
|
|
case BLK_ZONE_COND_IMP_OPEN:
|
|
|
|
case BLK_ZONE_COND_EXP_OPEN:
|
|
|
|
case BLK_ZONE_COND_CLOSED:
|
|
|
|
return zone->wp - zone->start;
|
|
|
|
case BLK_ZONE_COND_FULL:
|
|
|
|
return zone->len;
|
|
|
|
case BLK_ZONE_COND_EMPTY:
|
|
|
|
return 0;
|
|
|
|
case BLK_ZONE_COND_NOT_WP:
|
|
|
|
case BLK_ZONE_COND_OFFLINE:
|
|
|
|
case BLK_ZONE_COND_READONLY:
|
|
|
|
default:
|
|
|
|
/*
|
|
|
|
* Conventional, offline and read-only zones do not have a valid
|
|
|
|
* write pointer.
|
|
|
|
*/
|
|
|
|
return UINT_MAX;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int blk_zone_wplug_report_zone_cb(struct blk_zone *zone,
|
|
|
|
unsigned int idx, void *data)
|
|
|
|
{
|
|
|
|
struct blk_zone *zonep = data;
|
|
|
|
|
|
|
|
*zonep = *zone;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void disk_zone_wplug_handle_error(struct gendisk *disk,
|
|
|
|
struct blk_zone_wplug *zwplug)
|
|
|
|
{
|
|
|
|
sector_t zone_start_sector =
|
|
|
|
bdev_zone_sectors(disk->part0) * zwplug->zone_no;
|
|
|
|
unsigned int noio_flag;
|
|
|
|
struct blk_zone zone;
|
|
|
|
unsigned long flags;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
/* Get the current zone information from the device. */
|
|
|
|
noio_flag = memalloc_noio_save();
|
|
|
|
ret = disk->fops->report_zones(disk, zone_start_sector, 1,
|
|
|
|
blk_zone_wplug_report_zone_cb, &zone);
|
|
|
|
memalloc_noio_restore(noio_flag);
|
|
|
|
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* A zone reset or finish may have cleared the error already. In such
|
|
|
|
* case, do nothing as the report zones may have seen the "old" write
|
|
|
|
* pointer value before the reset/finish operation completed.
|
|
|
|
*/
|
|
|
|
if (!(zwplug->flags & BLK_ZONE_WPLUG_ERROR))
|
|
|
|
goto unlock;
|
|
|
|
|
|
|
|
zwplug->flags &= ~BLK_ZONE_WPLUG_ERROR;
|
|
|
|
|
|
|
|
if (ret != 1) {
|
|
|
|
/*
|
|
|
|
* We failed to get the zone information, meaning that something
|
|
|
|
* is likely really wrong with the device. Abort all remaining
|
|
|
|
* plugged BIOs as otherwise we could endup waiting forever on
|
|
|
|
* plugged BIOs to complete if there is a queue freeze on-going.
|
|
|
|
*/
|
|
|
|
disk_zone_wplug_abort(zwplug);
|
|
|
|
goto unplug;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Update the zone write pointer offset. */
|
|
|
|
zwplug->wp_offset = blk_zone_wp_offset(&zone);
|
|
|
|
disk_zone_wplug_abort_unaligned(disk, zwplug);
|
|
|
|
|
|
|
|
/* Restart BIO submission if we still have any BIO left. */
|
|
|
|
if (!bio_list_empty(&zwplug->bio_list)) {
|
2024-05-01 11:08:58 +00:00
|
|
|
disk_zone_wplug_schedule_bio_work(disk, zwplug);
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
goto unlock;
|
|
|
|
}
|
|
|
|
|
|
|
|
unplug:
|
|
|
|
zwplug->flags &= ~BLK_ZONE_WPLUG_PLUGGED;
|
|
|
|
if (disk_should_remove_zone_wplug(disk, zwplug))
|
|
|
|
disk_remove_zone_wplug(disk, zwplug);
|
|
|
|
|
|
|
|
unlock:
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void disk_zone_wplugs_work(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct gendisk *disk =
|
|
|
|
container_of(work, struct gendisk, zone_wplugs_work);
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
|
|
|
|
|
|
|
|
while (!list_empty(&disk->zone_wplugs_err_list)) {
|
|
|
|
zwplug = list_first_entry(&disk->zone_wplugs_err_list,
|
|
|
|
struct blk_zone_wplug, link);
|
|
|
|
list_del_init(&zwplug->link);
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
|
|
|
|
disk_zone_wplug_handle_error(disk, zwplug);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
|
|
|
|
spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline unsigned int disk_zone_wplugs_hash_size(struct gendisk *disk)
|
|
|
|
{
|
|
|
|
return 1U << disk->zone_wplugs_hash_bits;
|
|
|
|
}
|
|
|
|
|
|
|
|
void disk_init_zone_resources(struct gendisk *disk)
|
|
|
|
{
|
|
|
|
spin_lock_init(&disk->zone_wplugs_lock);
|
|
|
|
INIT_LIST_HEAD(&disk->zone_wplugs_err_list);
|
|
|
|
INIT_WORK(&disk->zone_wplugs_work, disk_zone_wplugs_work);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* For the size of a disk zone write plug hash table, use the size of the
|
|
|
|
* zone write plug mempool, which is the maximum of the disk open zones and
|
|
|
|
* active zones limits. But do not exceed 4KB (512 hlist head entries), that is,
|
|
|
|
* 9 bits. For a disk that has no limits, mempool size defaults to 128.
|
|
|
|
*/
|
|
|
|
#define BLK_ZONE_WPLUG_MAX_HASH_BITS 9
|
|
|
|
#define BLK_ZONE_WPLUG_DEFAULT_POOL_SIZE 128
|
|
|
|
|
|
|
|
static int disk_alloc_zone_resources(struct gendisk *disk,
|
|
|
|
unsigned int pool_size)
|
|
|
|
{
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
disk->zone_wplugs_hash_bits =
|
|
|
|
min(ilog2(pool_size) + 1, BLK_ZONE_WPLUG_MAX_HASH_BITS);
|
|
|
|
|
|
|
|
disk->zone_wplugs_hash =
|
|
|
|
kcalloc(disk_zone_wplugs_hash_size(disk),
|
|
|
|
sizeof(struct hlist_head), GFP_KERNEL);
|
|
|
|
if (!disk->zone_wplugs_hash)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
for (i = 0; i < disk_zone_wplugs_hash_size(disk); i++)
|
|
|
|
INIT_HLIST_HEAD(&disk->zone_wplugs_hash[i]);
|
|
|
|
|
|
|
|
disk->zone_wplugs_pool = mempool_create_kmalloc_pool(pool_size,
|
|
|
|
sizeof(struct blk_zone_wplug));
|
2024-04-20 07:58:11 +00:00
|
|
|
if (!disk->zone_wplugs_pool)
|
|
|
|
goto free_hash;
|
|
|
|
|
|
|
|
disk->zone_wplugs_wq =
|
|
|
|
alloc_workqueue("%s_zwplugs", WQ_MEM_RECLAIM | WQ_HIGHPRI,
|
|
|
|
pool_size, disk->disk_name);
|
|
|
|
if (!disk->zone_wplugs_wq)
|
|
|
|
goto destroy_pool;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
|
|
|
|
return 0;
|
2024-04-20 07:58:11 +00:00
|
|
|
|
|
|
|
destroy_pool:
|
|
|
|
mempool_destroy(disk->zone_wplugs_pool);
|
|
|
|
disk->zone_wplugs_pool = NULL;
|
|
|
|
free_hash:
|
|
|
|
kfree(disk->zone_wplugs_hash);
|
|
|
|
disk->zone_wplugs_hash = NULL;
|
|
|
|
disk->zone_wplugs_hash_bits = 0;
|
|
|
|
return -ENOMEM;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void disk_destroy_zone_wplugs_hash_table(struct gendisk *disk)
|
|
|
|
{
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
unsigned int i;
|
|
|
|
|
|
|
|
if (!disk->zone_wplugs_hash)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Free all the zone write plugs we have. */
|
|
|
|
for (i = 0; i < disk_zone_wplugs_hash_size(disk); i++) {
|
|
|
|
while (!hlist_empty(&disk->zone_wplugs_hash[i])) {
|
|
|
|
zwplug = hlist_entry(disk->zone_wplugs_hash[i].first,
|
|
|
|
struct blk_zone_wplug, node);
|
|
|
|
atomic_inc(&zwplug->ref);
|
|
|
|
disk_remove_zone_wplug(disk, zwplug);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
kfree(disk->zone_wplugs_hash);
|
|
|
|
disk->zone_wplugs_hash = NULL;
|
|
|
|
disk->zone_wplugs_hash_bits = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void disk_free_zone_resources(struct gendisk *disk)
|
|
|
|
{
|
2024-06-07 00:21:26 +00:00
|
|
|
if (!disk->zone_wplugs_pool)
|
|
|
|
return;
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
cancel_work_sync(&disk->zone_wplugs_work);
|
|
|
|
|
2024-04-20 07:58:11 +00:00
|
|
|
if (disk->zone_wplugs_wq) {
|
|
|
|
destroy_workqueue(disk->zone_wplugs_wq);
|
|
|
|
disk->zone_wplugs_wq = NULL;
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
disk_destroy_zone_wplugs_hash_table(disk);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Wait for the zone write plugs to be RCU-freed before
|
|
|
|
* destorying the mempool.
|
|
|
|
*/
|
|
|
|
rcu_barrier();
|
|
|
|
|
|
|
|
mempool_destroy(disk->zone_wplugs_pool);
|
|
|
|
disk->zone_wplugs_pool = NULL;
|
|
|
|
|
2024-07-04 05:28:16 +00:00
|
|
|
bitmap_free(disk->conv_zones_bitmap);
|
2022-07-06 07:03:50 +00:00
|
|
|
disk->conv_zones_bitmap = NULL;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
disk->zone_capacity = 0;
|
2024-05-30 05:40:34 +00:00
|
|
|
disk->last_zone_capacity = 0;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
disk->nr_zones = 0;
|
|
|
|
}
|
|
|
|
|
2024-04-08 01:41:11 +00:00
|
|
|
static inline bool disk_need_zone_resources(struct gendisk *disk)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* All mq zoned devices need zone resources so that the block layer
|
|
|
|
* can automatically handle write BIO plugging. BIO-based device drivers
|
|
|
|
* (e.g. DM devices) are normally responsible for handling zone write
|
|
|
|
* ordering and do not need zone resources, unless the driver requires
|
|
|
|
* zone append emulation.
|
|
|
|
*/
|
|
|
|
return queue_is_mq(disk->queue) ||
|
|
|
|
queue_emulates_zone_append(disk->queue);
|
|
|
|
}
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
static int disk_revalidate_zone_resources(struct gendisk *disk,
|
|
|
|
unsigned int nr_zones)
|
|
|
|
{
|
|
|
|
struct queue_limits *lim = &disk->queue->limits;
|
|
|
|
unsigned int pool_size;
|
|
|
|
|
2024-04-08 01:41:11 +00:00
|
|
|
if (!disk_need_zone_resources(disk))
|
|
|
|
return 0;
|
|
|
|
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
/*
|
|
|
|
* If the device has no limit on the maximum number of open and active
|
|
|
|
* zones, use BLK_ZONE_WPLUG_DEFAULT_POOL_SIZE.
|
|
|
|
*/
|
|
|
|
pool_size = max(lim->max_open_zones, lim->max_active_zones);
|
|
|
|
if (!pool_size)
|
|
|
|
pool_size = min(BLK_ZONE_WPLUG_DEFAULT_POOL_SIZE, nr_zones);
|
|
|
|
|
|
|
|
if (!disk->zone_wplugs_hash)
|
|
|
|
return disk_alloc_zone_resources(disk, pool_size);
|
|
|
|
|
|
|
|
return 0;
|
2018-10-12 10:08:50 +00:00
|
|
|
}
|
|
|
|
|
2019-11-11 02:39:30 +00:00
|
|
|
struct blk_revalidate_zone_args {
|
|
|
|
struct gendisk *disk;
|
2019-12-03 09:39:05 +00:00
|
|
|
unsigned long *conv_zones_bitmap;
|
2019-12-03 09:39:06 +00:00
|
|
|
unsigned int nr_zones;
|
2024-04-08 01:41:06 +00:00
|
|
|
unsigned int zone_capacity;
|
2024-05-30 05:40:34 +00:00
|
|
|
unsigned int last_zone_capacity;
|
2019-11-11 02:39:30 +00:00
|
|
|
sector_t sector;
|
|
|
|
};
|
|
|
|
|
2024-04-08 01:41:08 +00:00
|
|
|
/*
|
|
|
|
* Update the disk zone resources information and device queue limits.
|
|
|
|
* The disk queue is frozen when this is executed.
|
|
|
|
*/
|
|
|
|
static int disk_update_zone_resources(struct gendisk *disk,
|
|
|
|
struct blk_revalidate_zone_args *args)
|
|
|
|
{
|
|
|
|
struct request_queue *q = disk->queue;
|
2024-05-01 11:08:55 +00:00
|
|
|
unsigned int nr_seq_zones, nr_conv_zones = 0;
|
|
|
|
unsigned int pool_size;
|
2024-04-08 01:41:08 +00:00
|
|
|
struct queue_limits lim;
|
|
|
|
|
|
|
|
disk->nr_zones = args->nr_zones;
|
|
|
|
disk->zone_capacity = args->zone_capacity;
|
2024-05-30 05:40:34 +00:00
|
|
|
disk->last_zone_capacity = args->last_zone_capacity;
|
2024-04-08 01:41:08 +00:00
|
|
|
swap(disk->conv_zones_bitmap, args->conv_zones_bitmap);
|
2024-05-01 11:08:55 +00:00
|
|
|
if (disk->conv_zones_bitmap)
|
|
|
|
nr_conv_zones = bitmap_weight(disk->conv_zones_bitmap,
|
|
|
|
disk->nr_zones);
|
|
|
|
if (nr_conv_zones >= disk->nr_zones) {
|
|
|
|
pr_warn("%s: Invalid number of conventional zones %u / %u\n",
|
|
|
|
disk->disk_name, nr_conv_zones, disk->nr_zones);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2024-06-11 02:36:36 +00:00
|
|
|
lim = queue_limits_start_update(q);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Some devices can advertize zone resource limits that are larger than
|
|
|
|
* the number of sequential zones of the zoned block device, e.g. a
|
|
|
|
* small ZNS namespace. For such case, assume that the zoned device has
|
|
|
|
* no zone resource limits.
|
|
|
|
*/
|
|
|
|
nr_seq_zones = disk->nr_zones - nr_conv_zones;
|
|
|
|
if (lim.max_open_zones >= nr_seq_zones)
|
|
|
|
lim.max_open_zones = 0;
|
|
|
|
if (lim.max_active_zones >= nr_seq_zones)
|
|
|
|
lim.max_active_zones = 0;
|
|
|
|
|
2024-05-01 11:08:55 +00:00
|
|
|
if (!disk->zone_wplugs_pool)
|
2024-06-11 02:36:36 +00:00
|
|
|
goto commit;
|
2024-04-08 01:41:08 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If the device has no limit on the maximum number of open and active
|
|
|
|
* zones, set its max open zone limit to the mempool size to indicate
|
|
|
|
* to the user that there is a potential performance impact due to
|
|
|
|
* dynamic zone write plug allocation when simultaneously writing to
|
|
|
|
* more zones than the size of the mempool.
|
|
|
|
*/
|
2024-05-01 11:08:55 +00:00
|
|
|
pool_size = max(lim.max_open_zones, lim.max_active_zones);
|
|
|
|
if (!pool_size)
|
|
|
|
pool_size = min(BLK_ZONE_WPLUG_DEFAULT_POOL_SIZE, nr_seq_zones);
|
|
|
|
|
|
|
|
mempool_resize(disk->zone_wplugs_pool, pool_size);
|
|
|
|
|
|
|
|
if (!lim.max_open_zones && !lim.max_active_zones) {
|
|
|
|
if (pool_size < nr_seq_zones)
|
|
|
|
lim.max_open_zones = pool_size;
|
|
|
|
else
|
|
|
|
lim.max_open_zones = 0;
|
2024-04-08 01:41:08 +00:00
|
|
|
}
|
|
|
|
|
2024-06-11 02:36:36 +00:00
|
|
|
commit:
|
2024-05-01 11:08:55 +00:00
|
|
|
return queue_limits_commit_update(q, &lim);
|
2024-04-08 01:41:08 +00:00
|
|
|
}
|
|
|
|
|
2024-05-01 11:09:07 +00:00
|
|
|
static int blk_revalidate_conv_zone(struct blk_zone *zone, unsigned int idx,
|
|
|
|
struct blk_revalidate_zone_args *args)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = args->disk;
|
|
|
|
|
|
|
|
if (zone->capacity != zone->len) {
|
|
|
|
pr_warn("%s: Invalid conventional zone capacity\n",
|
|
|
|
disk->disk_name);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2024-05-30 05:40:34 +00:00
|
|
|
if (disk_zone_is_last(disk, zone))
|
|
|
|
args->last_zone_capacity = zone->capacity;
|
|
|
|
|
2024-05-01 11:09:07 +00:00
|
|
|
if (!disk_need_zone_resources(disk))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (!args->conv_zones_bitmap) {
|
|
|
|
args->conv_zones_bitmap =
|
2024-07-04 05:28:16 +00:00
|
|
|
bitmap_zalloc(args->nr_zones, GFP_NOIO);
|
2024-05-01 11:09:07 +00:00
|
|
|
if (!args->conv_zones_bitmap)
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
set_bit(idx, args->conv_zones_bitmap);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int blk_revalidate_seq_zone(struct blk_zone *zone, unsigned int idx,
|
|
|
|
struct blk_revalidate_zone_args *args)
|
|
|
|
{
|
|
|
|
struct gendisk *disk = args->disk;
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
unsigned int wp_offset;
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Remember the capacity of the first sequential zone and check
|
2024-05-30 05:40:33 +00:00
|
|
|
* if it is constant for all zones, ignoring the last zone as it can be
|
|
|
|
* smaller.
|
2024-05-01 11:09:07 +00:00
|
|
|
*/
|
|
|
|
if (!args->zone_capacity)
|
|
|
|
args->zone_capacity = zone->capacity;
|
2024-05-30 05:40:34 +00:00
|
|
|
if (disk_zone_is_last(disk, zone)) {
|
|
|
|
args->last_zone_capacity = zone->capacity;
|
|
|
|
} else if (zone->capacity != args->zone_capacity) {
|
2024-05-01 11:09:07 +00:00
|
|
|
pr_warn("%s: Invalid variable zone capacity\n",
|
|
|
|
disk->disk_name);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We need to track the write pointer of all zones that are not
|
|
|
|
* empty nor full. So make sure we have a zone write plug for
|
|
|
|
* such zone if the device has a zone write plug hash table.
|
|
|
|
*/
|
|
|
|
if (!disk->zone_wplugs_hash)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
wp_offset = blk_zone_wp_offset(zone);
|
|
|
|
if (!wp_offset || wp_offset >= zone->capacity)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
zwplug = disk_get_and_lock_zone_wplug(disk, zone->wp, GFP_NOIO, &flags);
|
|
|
|
if (!zwplug)
|
|
|
|
return -ENOMEM;
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
|
|
|
disk_put_zone_wplug(zwplug);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
/*
|
|
|
|
* Helper function to check the validity of zones of a zoned block device.
|
|
|
|
*/
|
2019-11-11 02:39:30 +00:00
|
|
|
static int blk_revalidate_zone_cb(struct blk_zone *zone, unsigned int idx,
|
|
|
|
void *data)
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
{
|
2019-11-11 02:39:30 +00:00
|
|
|
struct blk_revalidate_zone_args *args = data;
|
|
|
|
struct gendisk *disk = args->disk;
|
2024-05-01 11:09:07 +00:00
|
|
|
sector_t zone_sectors = disk->queue->limits.chunk_sectors;
|
|
|
|
int ret;
|
2023-07-03 02:48:12 +00:00
|
|
|
|
|
|
|
/* Check for bad zones and holes in the zone report */
|
|
|
|
if (zone->start != args->sector) {
|
|
|
|
pr_warn("%s: Zone gap at sectors %llu..%llu\n",
|
|
|
|
disk->disk_name, args->sector, zone->start);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2024-05-30 05:40:33 +00:00
|
|
|
if (zone->start >= get_capacity(disk) || !zone->len) {
|
2023-07-03 02:48:12 +00:00
|
|
|
pr_warn("%s: Invalid zone start %llu, length %llu\n",
|
|
|
|
disk->disk_name, zone->start, zone->len);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* All zones must have the same size, with the exception on an eventual
|
|
|
|
* smaller last zone.
|
|
|
|
*/
|
2024-05-30 05:40:33 +00:00
|
|
|
if (!disk_zone_is_last(disk, zone)) {
|
2023-07-03 02:48:12 +00:00
|
|
|
if (zone->len != zone_sectors) {
|
2019-12-03 09:39:08 +00:00
|
|
|
pr_warn("%s: Invalid zoned device with non constant zone size\n",
|
|
|
|
disk->disk_name);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
2023-07-03 02:48:12 +00:00
|
|
|
} else if (zone->len > zone_sectors) {
|
|
|
|
pr_warn("%s: Invalid zoned device with larger last zone size\n",
|
|
|
|
disk->disk_name);
|
2019-11-11 02:39:30 +00:00
|
|
|
return -ENODEV;
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
}
|
|
|
|
|
2024-04-08 01:41:06 +00:00
|
|
|
if (!zone->capacity || zone->capacity > zone->len) {
|
|
|
|
pr_warn("%s: Invalid zone capacity\n",
|
|
|
|
disk->disk_name);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
/* Check zone type */
|
|
|
|
switch (zone->type) {
|
|
|
|
case BLK_ZONE_TYPE_CONVENTIONAL:
|
2024-05-01 11:09:07 +00:00
|
|
|
ret = blk_revalidate_conv_zone(zone, idx, args);
|
2019-12-03 09:39:06 +00:00
|
|
|
break;
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
case BLK_ZONE_TYPE_SEQWRITE_REQ:
|
2024-05-01 11:09:07 +00:00
|
|
|
ret = blk_revalidate_seq_zone(zone, idx, args);
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
break;
|
2024-01-07 07:22:12 +00:00
|
|
|
case BLK_ZONE_TYPE_SEQWRITE_PREF:
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
default:
|
|
|
|
pr_warn("%s: Invalid zone type 0x%x at sectors %llu\n",
|
|
|
|
disk->disk_name, (int)zone->type, zone->start);
|
2024-05-01 11:09:07 +00:00
|
|
|
ret = -ENODEV;
|
block: Enhance blk_revalidate_disk_zones()
For ZBC and ZAC zoned devices, the scsi driver revalidation processing
implemented by sd_revalidate_disk() includes a call to
sd_zbc_read_zones() which executes a full disk zone report used to
check that all zones of the disk are the same size. This processing is
followed by a call to blk_revalidate_disk_zones(), used to initialize
the device request queue zone bitmaps (zone type and zone write lock
bitmaps). To do so, blk_revalidate_disk_zones() also executes a full
device zone report to obtain zone types. As a result, the entire
zoned block device revalidation process includes two full device zone
report.
By moving the zone size checks into blk_revalidate_disk_zones(), this
process can be optimized to a single full device zone report, leading to
shorter device scan and revalidation times. This patch implements this
optimization, reducing the original full device zone report implemented
in sd_zbc_check_zones() to a single, small, report zones command
execution to obtain the size of the first zone of the device. Checks
whether all zones of the device are the same size as the first zone
size are moved to the generic blk_check_zone() function called from
blk_revalidate_disk_zones().
This optimization also has the following benefits:
1) fewer memory allocations in the scsi layer during disk revalidation
as the potentailly large buffer for zone report execution is not
needed.
2) Implement zone checks in a generic manner, reducing the burden on
device driver which only need to obtain the zone size and check that
this size is a power of 2 number of LBAs. Any new type of zoned
block device will benefit from this.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2019-11-11 02:39:22 +00:00
|
|
|
}
|
|
|
|
|
2024-05-01 11:09:07 +00:00
|
|
|
if (!ret)
|
|
|
|
args->sector += zone->len;
|
|
|
|
|
|
|
|
return ret;
|
2019-11-11 02:39:30 +00:00
|
|
|
}
|
|
|
|
|
2018-10-12 10:08:50 +00:00
|
|
|
/**
|
2024-04-08 01:41:26 +00:00
|
|
|
* blk_revalidate_disk_zones - (re)allocate and initialize zone write plugs
|
2018-10-12 10:08:50 +00:00
|
|
|
* @disk: Target disk
|
|
|
|
*
|
2024-04-08 01:41:20 +00:00
|
|
|
* Helper function for low-level device drivers to check, (re) allocate and
|
|
|
|
* initialize resources used for managing zoned disks. This function should
|
|
|
|
* normally be called by blk-mq based drivers when a zoned gendisk is probed
|
|
|
|
* and when the zone configuration of the gendisk changes (e.g. after a format).
|
2023-07-03 02:48:12 +00:00
|
|
|
* Before calling this function, the device driver must already have set the
|
|
|
|
* device zone size (chunk_sector limit) and the max zone append limit.
|
2024-04-08 01:41:11 +00:00
|
|
|
* BIO based drivers can also use this function as long as the device queue
|
|
|
|
* can be safely frozen.
|
2018-10-12 10:08:50 +00:00
|
|
|
*/
|
2024-04-08 01:41:20 +00:00
|
|
|
int blk_revalidate_disk_zones(struct gendisk *disk)
|
2018-10-12 10:08:50 +00:00
|
|
|
{
|
|
|
|
struct request_queue *q = disk->queue;
|
2023-07-03 02:48:12 +00:00
|
|
|
sector_t zone_sectors = q->limits.chunk_sectors;
|
|
|
|
sector_t capacity = get_capacity(disk);
|
|
|
|
struct blk_revalidate_zone_args args = { };
|
2019-12-03 09:39:08 +00:00
|
|
|
unsigned int noio_flag;
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
int ret = -ENOMEM;
|
2018-10-12 10:08:50 +00:00
|
|
|
|
2019-11-11 02:39:23 +00:00
|
|
|
if (WARN_ON_ONCE(!blk_queue_is_zoned(q)))
|
|
|
|
return -EIO;
|
2018-10-12 10:08:50 +00:00
|
|
|
|
2023-07-03 02:48:12 +00:00
|
|
|
if (!capacity)
|
|
|
|
return -ENODEV;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Checks that the device driver indicated a valid zone size and that
|
|
|
|
* the max zone append limit is set.
|
|
|
|
*/
|
|
|
|
if (!zone_sectors || !is_power_of_2(zone_sectors)) {
|
|
|
|
pr_warn("%s: Invalid non power of two zone size (%llu)\n",
|
|
|
|
disk->disk_name, zone_sectors);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
|
|
|
|
2024-04-08 01:41:09 +00:00
|
|
|
if (!queue_max_zone_append_sectors(q)) {
|
2023-07-03 02:48:12 +00:00
|
|
|
pr_warn("%s: Invalid 0 maximum zone append limit\n",
|
|
|
|
disk->disk_name);
|
|
|
|
return -ENODEV;
|
|
|
|
}
|
2020-07-30 11:25:17 +00:00
|
|
|
|
2019-12-03 09:39:06 +00:00
|
|
|
/*
|
2019-12-03 09:39:08 +00:00
|
|
|
* Ensure that all memory allocations in this context are done as if
|
|
|
|
* GFP_NOIO was specified.
|
2019-12-03 09:39:06 +00:00
|
|
|
*/
|
2023-07-03 02:48:12 +00:00
|
|
|
args.disk = disk;
|
|
|
|
args.nr_zones = (capacity + zone_sectors - 1) >> ilog2(zone_sectors);
|
2019-12-03 09:39:08 +00:00
|
|
|
noio_flag = memalloc_noio_save();
|
block: Introduce zone write plugging
Zone write plugging implements a per-zone "plug" for write operations
to control the submission and execution order of write operations to
sequential write required zones of a zoned block device. Per-zone
plugging guarantees that at any time there is at most only one write
request per zone being executed. This mechanism is intended to replace
zone write locking which implements a similar per-zone write throttling
at the scheduler level, but is implemented only by mq-deadline.
Unlike zone write locking which operates on requests, zone write
plugging operates on BIOs. A zone write plug is simply a BIO list that
is atomically manipulated using a spinlock and a kblockd submission
work. A write BIO to a zone is "plugged" to delay its execution if a
write BIO for the same zone was already issued, that is, if a write
request for the same zone is being executed. The next plugged BIO is
unplugged and issued once the write request completes.
This mechanism allows to:
- Untangle zone write ordering from block IO schedulers. This allows
removing the restriction on using mq-deadline for writing to zoned
block devices. Any block IO scheduler, including "none" can be used.
- Zone write plugging operates on BIOs instead of requests. Plugged
BIOs waiting for execution thus do not hold scheduling tags and thus
are not preventing other BIOs from executing (reads or writes to
other zones). Depending on the workload, this can significantly
improve the device use (higher queue depth operation) and
performance.
- Both blk-mq (request based) zoned devices and BIO-based zoned devices
(e.g. device mapper) can use zone write plugging. It is mandatory
for the former but optional for the latter. BIO-based drivers can
use zone write plugging to implement write ordering guarantees, or
the drivers can implement their own if needed.
- The code is less invasive in the block layer and is mostly limited to
blk-zoned.c with some small changes in blk-mq.c, blk-merge.c and
bio.c.
Zone write plugging is implemented using struct blk_zone_wplug. This
structure includes a spinlock, a BIO list and a work structure to
handle the submission of plugged BIOs. Zone write plugs structures are
managed using a per-disk hash table.
Plugging of zone write BIOs is done using the function
blk_zone_write_plug_bio() which returns false if a BIO execution does
not need to be delayed and true otherwise. This function is called
from blk_mq_submit_bio() after a BIO is split to avoid large BIOs
spanning multiple zones which would cause mishandling of zone write
plugs. This ichange enables by default zone write plugging for any mq
request-based block device. BIO-based device drivers can also use zone
write plugging by expliclty calling blk_zone_write_plug_bio() in their
->submit_bio method. For such devices, the driver must ensure that a
BIO passed to blk_zone_write_plug_bio() is already split and not
straddling zone boundaries.
Only write and write zeroes BIOs are plugged. Zone write plugging does
not introduce any significant overhead for other operations. A BIO that
is being handled through zone write plugging is flagged using the new
BIO flag BIO_ZONE_WRITE_PLUGGING. A request handling a BIO flagged with
this new flag is flagged with the new RQF_ZONE_WRITE_PLUGGING flag.
The completion of BIOs and requests flagged trigger respectively calls
to the functions blk_zone_write_bio_endio() and
blk_zone_write_complete_request(). The latter function is used to
trigger submission of the next plugged BIO using the zone plug work.
blk_zone_write_bio_endio() does the same for BIO-based devices.
This ensures that at any time, at most one request (blk-mq devices) or
one BIO (BIO-based devices) is being executed for any zone. The
handling of zone write plugs using a per-zone plug spinlock maximizes
parallelism and device usage by allowing multiple zones to be writen
simultaneously without lock contention.
Zone write plugging ignores flush BIOs without data. Hovever, any flush
BIO that has data is always plugged so that the write part of the flush
sequence is serialized with other regular writes.
Given that any BIO handled through zone write plugging will be the only
BIO in flight for the target zone when it is executed, the unplugging
and submission of a BIO will have no chance of successfully merging with
plugged requests or requests in the scheduler. To overcome this
potential performance degradation, blk_mq_submit_bio() calls the
function blk_zone_write_plug_attempt_merge() to try to merge other
plugged BIOs with the one just unplugged and submitted. Successful
merging is signaled using blk_zone_write_plug_bio_merged(), called from
bio_attempt_back_merge(). Furthermore, to avoid recalculating the number
of segments of plugged BIOs to attempt merging, the number of segments
of a plugged BIO is saved using the new struct bio field
__bi_nr_segments. To avoid growing the size of struct bio, this field is
added as a union with the bio_cookie field. This is safe to do as
polling is always disabled for plugged BIOs.
When BIOs are plugged in a zone write plug, the device request queue
usage counter is always incremented. This reference is kept and reused
for blk-mq devices when the plugged BIO is unplugged and submitted
again using submit_bio_noacct_nocheck(). For this case, the unplugged
BIO is already flagged with BIO_ZONE_WRITE_PLUGGING and
blk_mq_submit_bio() proceeds directly to allocating a new request for
the BIO, re-using the usage reference count taken when the BIO was
plugged. This extra reference count is dropped in
blk_zone_write_plug_attempt_merge() for any plugged BIO that is
successfully merged. Given that BIO-based devices will not take this
path, the extra reference is dropped after a plugged BIO is unplugged
and submitted.
Zone write plugs are dynamically allocated and managed using a hash
table (an array of struct hlist_head) with RCU protection.
A zone write plug is allocated when a write BIO is received for the
zone and not freed until the zone is fully written, reset or finished.
To detect when a zone write plug can be freed, the write state of each
zone is tracked using a write pointer offset which corresponds to the
offset of a zone write pointer relative to the zone start. Write
operations always increment this write pointer offset. Zone reset
operations set it to 0 and zone finish operations set it to the zone
size.
If a write error happens, the wp_offset value of a zone write plug may
become incorrect and out of sync with the device managed write pointer.
This is handled using the zone write plug flag BLK_ZONE_WPLUG_ERROR.
The function blk_zone_wplug_handle_error() is called from the new disk
zone write plug work when this flag is set. This function executes a
report zone to update the zone write pointer offset to the current
value as indicated by the device. The disk zone write plug work is
scheduled whenever a BIO flagged with BIO_ZONE_WRITE_PLUGGING completes
with an error or when bio_zone_wplug_prepare_bio() detects an unaligned
write. Once scheduled, the disk zone write plugs work keeps running
until all zone errors are handled.
To match the new data structures used for zoned disks, the function
disk_free_zone_bitmaps() is renamed to the more generic
disk_free_zone_resources(). The function disk_init_zone_resources() is
also introduced to initialize zone write plugs resources when a gendisk
is allocated.
In order to guarantee that the user can simultaneously write up to a
number of zones equal to a device max active zone limit or max open zone
limit, zone write plugs are allocated using a mempool sized to the
maximum of these 2 device limits. For a device that does not have
active and open zone limits, 128 is used as the default mempool size.
If a change to the device active and open zone limits is detected, the
disk mempool is resized when blk_revalidate_disk_zones() is executed.
This commit contains contributions from Christoph Hellwig <hch@lst.de>.
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Tested-by: Hans Holmberg <hans.holmberg@wdc.com>
Tested-by: Dennis Maisenbacher <dennis.maisenbacher@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240408014128.205141-8-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 01:41:07 +00:00
|
|
|
ret = disk_revalidate_zone_resources(disk, args.nr_zones);
|
|
|
|
if (ret) {
|
|
|
|
memalloc_noio_restore(noio_flag);
|
|
|
|
return ret;
|
|
|
|
}
|
2019-12-03 09:39:08 +00:00
|
|
|
ret = disk->fops->report_zones(disk, 0, UINT_MAX,
|
|
|
|
blk_revalidate_zone_cb, &args);
|
2020-11-11 07:36:06 +00:00
|
|
|
if (!ret) {
|
|
|
|
pr_warn("%s: No zones reported\n", disk->disk_name);
|
|
|
|
ret = -ENODEV;
|
|
|
|
}
|
2019-12-03 09:39:08 +00:00
|
|
|
memalloc_noio_restore(noio_flag);
|
2018-10-12 10:08:50 +00:00
|
|
|
|
2020-11-11 07:36:06 +00:00
|
|
|
/*
|
|
|
|
* If zones where reported, make sure that the entire disk capacity
|
|
|
|
* has been checked.
|
|
|
|
*/
|
2023-07-03 02:48:12 +00:00
|
|
|
if (ret > 0 && args.sector != capacity) {
|
2020-11-11 07:36:06 +00:00
|
|
|
pr_warn("%s: Missing zones from sector %llu\n",
|
|
|
|
disk->disk_name, args.sector);
|
|
|
|
ret = -ENODEV;
|
|
|
|
}
|
|
|
|
|
2018-10-12 10:08:50 +00:00
|
|
|
/*
|
2024-04-08 01:41:26 +00:00
|
|
|
* Set the new disk zone parameters only once the queue is frozen and
|
|
|
|
* all I/Os are completed.
|
2018-10-12 10:08:50 +00:00
|
|
|
*/
|
|
|
|
blk_mq_freeze_queue(q);
|
2024-04-08 01:41:20 +00:00
|
|
|
if (ret > 0)
|
2024-04-08 01:41:08 +00:00
|
|
|
ret = disk_update_zone_resources(disk, &args);
|
2024-04-08 01:41:20 +00:00
|
|
|
else
|
2018-10-12 10:08:50 +00:00
|
|
|
pr_warn("%s: failed to revalidate zones\n", disk->disk_name);
|
2024-04-08 01:41:08 +00:00
|
|
|
if (ret)
|
|
|
|
disk_free_zone_resources(disk);
|
2019-11-11 02:39:30 +00:00
|
|
|
blk_mq_unfreeze_queue(q);
|
2018-10-12 10:08:50 +00:00
|
|
|
|
2019-12-03 09:39:05 +00:00
|
|
|
kfree(args.conv_zones_bitmap);
|
2024-04-08 01:41:06 +00:00
|
|
|
|
2018-10-12 10:08:50 +00:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(blk_revalidate_disk_zones);
|
2024-04-08 01:41:24 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_BLK_DEBUG_FS
|
|
|
|
|
2024-04-08 01:41:25 +00:00
|
|
|
int queue_zone_wplugs_show(void *data, struct seq_file *m)
|
2024-04-08 01:41:24 +00:00
|
|
|
{
|
|
|
|
struct request_queue *q = data;
|
2024-04-08 01:41:25 +00:00
|
|
|
struct gendisk *disk = q->disk;
|
|
|
|
struct blk_zone_wplug *zwplug;
|
|
|
|
unsigned int zwp_wp_offset, zwp_flags;
|
|
|
|
unsigned int zwp_zone_no, zwp_ref;
|
|
|
|
unsigned int zwp_bio_list_size, i;
|
|
|
|
unsigned long flags;
|
2024-04-08 01:41:24 +00:00
|
|
|
|
2024-04-25 12:02:39 +00:00
|
|
|
if (!disk->zone_wplugs_hash)
|
|
|
|
return 0;
|
|
|
|
|
2024-04-08 01:41:25 +00:00
|
|
|
rcu_read_lock();
|
|
|
|
for (i = 0; i < disk_zone_wplugs_hash_size(disk); i++) {
|
|
|
|
hlist_for_each_entry_rcu(zwplug,
|
|
|
|
&disk->zone_wplugs_hash[i], node) {
|
|
|
|
spin_lock_irqsave(&zwplug->lock, flags);
|
|
|
|
zwp_zone_no = zwplug->zone_no;
|
|
|
|
zwp_flags = zwplug->flags;
|
|
|
|
zwp_ref = atomic_read(&zwplug->ref);
|
|
|
|
zwp_wp_offset = zwplug->wp_offset;
|
|
|
|
zwp_bio_list_size = bio_list_size(&zwplug->bio_list);
|
|
|
|
spin_unlock_irqrestore(&zwplug->lock, flags);
|
2024-04-08 01:41:24 +00:00
|
|
|
|
2024-04-08 01:41:25 +00:00
|
|
|
seq_printf(m, "%u 0x%x %u %u %u\n",
|
|
|
|
zwp_zone_no, zwp_flags, zwp_ref,
|
|
|
|
zwp_wp_offset, zwp_bio_list_size);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
rcu_read_unlock();
|
2024-04-08 01:41:24 +00:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|