linux/block
Luis Chamberlain 425fbcd62d
bdev: use bdev_io_min() for statx block size
You can use lsblk to query for a block device block device block size:

lsblk -o MIN-IO /dev/nvme0n1
MIN-IO
 4096

The min-io is the minimum IO the block device prefers for optimal
performance. In turn we map this to the block device block size.
The current block size exposed even for block devices with an
LBA format of 16k is 4k. Likewise devices which support 4k LBA format
but have a larger Indirection Unit of 16k have an exposed block size
of 4k.

This incurs read-modify-writes on direct IO against devices with a
min-io larger than the page size. To fix this, use the block device
min io, which is the minimal optimal IO the device prefers.

With this we now get:

lsblk -o MIN-IO /dev/nvme0n1
MIN-IO
 16384

And so userspace gets the appropriate information it needs for optimal
performance. This is verified with blkalgn against mkfs against a
device with LBA format of 4k but an NPWG of 16k (min io size)

mkfs.xfs -f -b size=16k  /dev/nvme3n1
blkalgn -d nvme3n1 --ops Write

     Block size          : count     distribution
         0 -> 1          : 0        |                                        |
         2 -> 3          : 0        |                                        |
         4 -> 7          : 0        |                                        |
         8 -> 15         : 0        |                                        |
        16 -> 31         : 0        |                                        |
        32 -> 63         : 0        |                                        |
        64 -> 127        : 0        |                                        |
       128 -> 255        : 0        |                                        |
       256 -> 511        : 0        |                                        |
       512 -> 1023       : 0        |                                        |
      1024 -> 2047       : 0        |                                        |
      2048 -> 4095       : 0        |                                        |
      4096 -> 8191       : 0        |                                        |
      8192 -> 16383      : 0        |                                        |
     16384 -> 32767      : 66       |****************************************|
     32768 -> 65535      : 0        |                                        |
     65536 -> 131071     : 0        |                                        |
    131072 -> 262143     : 2        |*                                       |
Block size: 14 - 66
Block size: 17 - 2

     Algn size           : count     distribution
         0 -> 1          : 0        |                                        |
         2 -> 3          : 0        |                                        |
         4 -> 7          : 0        |                                        |
         8 -> 15         : 0        |                                        |
        16 -> 31         : 0        |                                        |
        32 -> 63         : 0        |                                        |
        64 -> 127        : 0        |                                        |
       128 -> 255        : 0        |                                        |
       256 -> 511        : 0        |                                        |
       512 -> 1023       : 0        |                                        |
      1024 -> 2047       : 0        |                                        |
      2048 -> 4095       : 0        |                                        |
      4096 -> 8191       : 0        |                                        |
      8192 -> 16383      : 0        |                                        |
     16384 -> 32767      : 66       |****************************************|
     32768 -> 65535      : 0        |                                        |
     65536 -> 131071     : 0        |                                        |
    131072 -> 262143     : 2        |*                                       |
Algn size: 14 - 66
Algn size: 17 - 2

Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Link: https://lore.kernel.org/r/20250221223823.1680616-9-mcgrof@kernel.org
Reviewed-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-02-24 11:44:44 +01:00
..
partitions partitions: ldm: remove the initial kernel-doc notation 2025-01-13 07:47:19 -07:00
Kconfig
Kconfig.iosched
Makefile blk-mq: remove unused queue mapping helpers 2024-12-23 08:17:23 -07:00
badblocks.c
bdev.c bdev: use bdev_io_min() for statx block size 2025-02-24 11:44:44 +01:00
bfq-cgroup.c Revert "block, bfq: merge bfq_release_process_ref() into bfq_put_cooperator()" 2024-11-19 19:05:32 -07:00
bfq-iosched.c for-6.14/block-20250118 2025-01-20 19:38:46 -08:00
bfq-iosched.h
bfq-wf2q.c
bio-integrity.c block: add support to pass user meta buffer 2024-12-23 08:17:17 -07:00
bio.c block: Don't trim an atomic write 2025-01-17 13:13:55 -07:00
blk-cgroup-fc-appid.c
blk-cgroup-rwstat.c
blk-cgroup-rwstat.h blk-cgroup: rwstat: fix kernel-doc warnings in header file 2025-01-13 07:47:09 -07:00
blk-cgroup.c block-6.14-20250131 2025-01-31 11:49:30 -08:00
blk-cgroup.h blk-cgroup: fix kernel-doc warnings in header file 2025-01-13 07:46:55 -07:00
blk-core.c block: get rid of request queue ->sysfs_dir_lock 2025-01-29 07:16:47 -07:00
blk-crypto-fallback.c block: Rework bio_split() return value 2024-11-11 08:35:46 -07:00
blk-crypto-internal.h
blk-crypto-profile.c
blk-crypto-sysfs.c
blk-crypto.c
blk-flush.c
blk-ia-ranges.c block: get rid of request queue ->sysfs_dir_lock 2025-01-29 07:16:47 -07:00
blk-integrity.c for-6.14/io_uring-20250119 2025-01-20 20:27:33 -08:00
blk-ioc.c block: replace call_rcu by kfree_rcu for simple kmem_cache_free callback 2024-10-22 08:16:40 -06:00
blk-iocost.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-iolatency.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-ioprio.c
blk-ioprio.h
blk-lib.c
blk-map.c block: remove blk_rq_bio_prep 2025-01-04 15:27:35 -07:00
blk-merge.c block: add a dma mapping iterator 2025-01-06 07:37:11 -07:00
blk-mq-cpumap.c blk-mq: create correct map for fallback case 2025-01-23 06:34:32 -07:00
blk-mq-debugfs.c block: simplify tag allocation policy selection 2025-01-06 07:37:41 -07:00
blk-mq-debugfs.h
blk-mq-sched.c block: remove BLK_MQ_F_SHOULD_MERGE 2024-12-23 08:17:23 -07:00
blk-mq-sched.h
blk-mq-sysfs.c block: fix nr_hw_queue update racing with disk addition/removal 2025-01-29 07:16:47 -07:00
blk-mq-tag.c block: simplify tag allocation policy selection 2025-01-06 07:37:41 -07:00
blk-mq.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-mq.h block: don't update BLK_FEAT_POLL in __blk_mq_update_nr_hw_queues 2025-01-10 07:29:23 -07:00
blk-pm.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-pm.h
blk-rq-qos.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-rq-qos.h
blk-settings.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-stat.c
blk-stat.h
blk-sysfs.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-throttle.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk-throttle.h
blk-timeout.c
blk-wbt.c
blk-wbt.h
blk-zoned.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
blk.h block: limit disk max sectors to (LLONG_MAX >> 9) 2025-01-15 15:46:56 -07:00
bounce.c
bsg-lib.c block: remove BLK_MQ_F_NO_SCHED 2025-01-06 07:37:41 -07:00
bsg.c
disk-events.c
early-lookup.c
elevator.c block: force noio scope in blk_mq_freeze_queue 2025-01-31 07:20:08 -07:00
elevator.h elevator: Enable const sysfs attributes 2025-01-02 13:20:29 -07:00
fops.c block: don't revert iter for -EIOCBQUEUED 2025-01-23 06:18:41 -07:00
genhd.c block: limit disk max sectors to (LLONG_MAX >> 9) 2025-01-15 15:46:56 -07:00
holder.c
ioctl.c
ioprio.c
kyber-iosched.c kyber: constify sysfs attributes 2025-01-02 13:20:29 -07:00
mq-deadline.c block: mq-deadline: Constify sysfs attributes 2025-01-02 13:20:29 -07:00
opal_proto.h
sed-opal.c block: sed-opal: add ioctl IOC_OPAL_SET_SID_PW 2024-10-22 08:16:40 -06:00
t10-pi.c move asm/unaligned.h to linux/unaligned.h 2024-10-02 17:23:23 -04:00