xref: /linux/Documentation/ABI/stable/sysfs-block (revision a1ff5a7d78a036d6c2178ee5acd6ba4946243800)
1What:		/sys/block/<disk>/alignment_offset
2Date:		April 2009
3Contact:	Martin K. Petersen <martin.petersen@oracle.com>
4Description:
5		Storage devices may report a physical block size that is
6		bigger than the logical block size (for instance a drive
7		with 4KB physical sectors exposing 512-byte logical
8		blocks to the operating system).  This parameter
9		indicates how many bytes the beginning of the device is
10		offset from the disk's natural alignment.
11
12
13What:		/sys/block/<disk>/discard_alignment
14Date:		May 2011
15Contact:	Martin K. Petersen <martin.petersen@oracle.com>
16Description:
17		Devices that support discard functionality may
18		internally allocate space in units that are bigger than
19		the exported logical block size. The discard_alignment
20		parameter indicates how many bytes the beginning of the
21		device is offset from the internal allocation unit's
22		natural alignment.
23
24What:		/sys/block/<disk>/atomic_write_max_bytes
25Date:		February 2024
26Contact:	Himanshu Madhani <himanshu.madhani@oracle.com>
27Description:
28		[RO] This parameter specifies the maximum atomic write
29		size reported by the device. This parameter is relevant
30		for merging of writes, where a merged atomic write
31		operation must not exceed this number of bytes.
32		This parameter may be greater than the value in
33		atomic_write_unit_max_bytes as
34		atomic_write_unit_max_bytes will be rounded down to a
35		power-of-two and atomic_write_unit_max_bytes may also be
36		limited by some other queue limits, such as max_segments.
37		This parameter - along with atomic_write_unit_min_bytes
38		and atomic_write_unit_max_bytes - will not be larger than
39		max_hw_sectors_kb, but may be larger than max_sectors_kb.
40
41
42What:		/sys/block/<disk>/atomic_write_unit_min_bytes
43Date:		February 2024
44Contact:	Himanshu Madhani <himanshu.madhani@oracle.com>
45Description:
46		[RO] This parameter specifies the smallest block which can
47		be written atomically with an atomic write operation. All
48		atomic write operations must begin at a
49		atomic_write_unit_min boundary and must be multiples of
50		atomic_write_unit_min. This value must be a power-of-two.
51
52
53What:		/sys/block/<disk>/atomic_write_unit_max_bytes
54Date:		February 2024
55Contact:	Himanshu Madhani <himanshu.madhani@oracle.com>
56Description:
57		[RO] This parameter defines the largest block which can be
58		written atomically with an atomic write operation. This
59		value must be a multiple of atomic_write_unit_min and must
60		be a power-of-two. This value will not be larger than
61		atomic_write_max_bytes.
62
63
64What:		/sys/block/<disk>/atomic_write_boundary_bytes
65Date:		February 2024
66Contact:	Himanshu Madhani <himanshu.madhani@oracle.com>
67Description:
68		[RO] A device may need to internally split an atomic write I/O
69		which straddles a given logical block address boundary. This
70		parameter specifies the size in bytes of the atomic boundary if
71		one is reported by the device. This value must be a
72		power-of-two and at least the size as in
73		atomic_write_unit_max_bytes.
74		Any attempt to merge atomic write I/Os must not result in a
75		merged I/O which crosses this boundary (if any).
76
77
78What:		/sys/block/<disk>/diskseq
79Date:		February 2021
80Contact:	Matteo Croce <mcroce@microsoft.com>
81Description:
82		The /sys/block/<disk>/diskseq files reports the disk
83		sequence number, which is a monotonically increasing
84		number assigned to every drive.
85		Some devices, like the loop device, refresh such number
86		every time the backing file is changed.
87		The value type is 64 bit unsigned.
88
89
90What:		/sys/block/<disk>/inflight
91Date:		October 2009
92Contact:	Jens Axboe <axboe@kernel.dk>, Nikanth Karthikesan <knikanth@suse.de>
93Description:
94		Reports the number of I/O requests currently in progress
95		(pending / in flight) in a device driver. This can be less
96		than the number of requests queued in the block device queue.
97		The report contains 2 fields: one for read requests
98		and one for write requests.
99		The value type is unsigned int.
100		Cf. Documentation/block/stat.rst which contains a single value for
101		requests in flight.
102		This is related to /sys/block/<disk>/queue/nr_requests
103		and for SCSI device also its queue_depth.
104
105
106What:		/sys/block/<disk>/integrity/device_is_integrity_capable
107Date:		July 2014
108Contact:	Martin K. Petersen <martin.petersen@oracle.com>
109Description:
110		Indicates whether a storage device is capable of storing
111		integrity metadata. Set if the device is T10 PI-capable.
112
113
114What:		/sys/block/<disk>/integrity/format
115Date:		June 2008
116Contact:	Martin K. Petersen <martin.petersen@oracle.com>
117Description:
118		Metadata format for integrity capable block device.
119		E.g. T10-DIF-TYPE1-CRC.
120
121
122What:		/sys/block/<disk>/integrity/protection_interval_bytes
123Date:		July 2015
124Contact:	Martin K. Petersen <martin.petersen@oracle.com>
125Description:
126		Describes the number of data bytes which are protected
127		by one integrity tuple. Typically the device's logical
128		block size.
129
130
131What:		/sys/block/<disk>/integrity/read_verify
132Date:		June 2008
133Contact:	Martin K. Petersen <martin.petersen@oracle.com>
134Description:
135		Indicates whether the block layer should verify the
136		integrity of read requests serviced by devices that
137		support sending integrity metadata.
138
139
140What:		/sys/block/<disk>/integrity/tag_size
141Date:		June 2008
142Contact:	Martin K. Petersen <martin.petersen@oracle.com>
143Description:
144		Number of bytes of integrity tag space available per
145		512 bytes of data.
146
147
148What:		/sys/block/<disk>/integrity/write_generate
149Date:		June 2008
150Contact:	Martin K. Petersen <martin.petersen@oracle.com>
151Description:
152		Indicates whether the block layer should automatically
153		generate checksums for write requests bound for
154		devices that support receiving integrity metadata.
155
156
157What:		/sys/block/<disk>/partscan
158Date:		May 2024
159Contact:	Christoph Hellwig <hch@lst.de>
160Description:
161		The /sys/block/<disk>/partscan files reports if partition
162		scanning is enabled for the disk.  It returns "1" if partition
163		scanning is enabled, or "0" if not.  The value type is a 32-bit
164		unsigned integer, but only "0" and "1" are valid values.
165
166
167What:		/sys/block/<disk>/<partition>/alignment_offset
168Date:		April 2009
169Contact:	Martin K. Petersen <martin.petersen@oracle.com>
170Description:
171		Storage devices may report a physical block size that is
172		bigger than the logical block size (for instance a drive
173		with 4KB physical sectors exposing 512-byte logical
174		blocks to the operating system).  This parameter
175		indicates how many bytes the beginning of the partition
176		is offset from the disk's natural alignment.
177
178
179What:		/sys/block/<disk>/<partition>/discard_alignment
180Date:		May 2011
181Contact:	Martin K. Petersen <martin.petersen@oracle.com>
182Description:
183		Devices that support discard functionality may
184		internally allocate space in units that are bigger than
185		the exported logical block size. The discard_alignment
186		parameter indicates how many bytes the beginning of the
187		partition is offset from the internal allocation unit's
188		natural alignment.
189
190
191What:		/sys/block/<disk>/<partition>/stat
192Date:		February 2008
193Contact:	Jerome Marchand <jmarchan@redhat.com>
194Description:
195		The /sys/block/<disk>/<partition>/stat files display the
196		I/O statistics of partition <partition>. The format is the
197		same as the format of /sys/block/<disk>/stat.
198
199
200What:		/sys/block/<disk>/queue/add_random
201Date:		June 2010
202Contact:	linux-block@vger.kernel.org
203Description:
204		[RW] This file allows to turn off the disk entropy contribution.
205		Default value of this file is '1'(on).
206
207
208What:		/sys/block/<disk>/queue/chunk_sectors
209Date:		September 2016
210Contact:	Hannes Reinecke <hare@suse.com>
211Description:
212		[RO] chunk_sectors has different meaning depending on the type
213		of the disk. For a RAID device (dm-raid), chunk_sectors
214		indicates the size in 512B sectors of the RAID volume stripe
215		segment. For a zoned block device, either host-aware or
216		host-managed, chunk_sectors indicates the size in 512B sectors
217		of the zones of the device, with the eventual exception of the
218		last zone of the device which may be smaller.
219
220
221What:		/sys/block/<disk>/queue/crypto/
222Date:		February 2022
223Contact:	linux-block@vger.kernel.org
224Description:
225		The presence of this subdirectory of /sys/block/<disk>/queue/
226		indicates that the device supports inline encryption.  This
227		subdirectory contains files which describe the inline encryption
228		capabilities of the device.  For more information about inline
229		encryption, refer to Documentation/block/inline-encryption.rst.
230
231
232What:		/sys/block/<disk>/queue/crypto/max_dun_bits
233Date:		February 2022
234Contact:	linux-block@vger.kernel.org
235Description:
236		[RO] This file shows the maximum length, in bits, of data unit
237		numbers accepted by the device in inline encryption requests.
238
239
240What:		/sys/block/<disk>/queue/crypto/modes/<mode>
241Date:		February 2022
242Contact:	linux-block@vger.kernel.org
243Description:
244		[RO] For each crypto mode (i.e., encryption/decryption
245		algorithm) the device supports with inline encryption, a file
246		will exist at this location.  It will contain a hexadecimal
247		number that is a bitmask of the supported data unit sizes, in
248		bytes, for that crypto mode.
249
250		Currently, the crypto modes that may be supported are:
251
252		   * AES-256-XTS
253		   * AES-128-CBC-ESSIV
254		   * Adiantum
255
256		For example, if a device supports AES-256-XTS inline encryption
257		with data unit sizes of 512 and 4096 bytes, the file
258		/sys/block/<disk>/queue/crypto/modes/AES-256-XTS will exist and
259		will contain "0x1200".
260
261
262What:		/sys/block/<disk>/queue/crypto/num_keyslots
263Date:		February 2022
264Contact:	linux-block@vger.kernel.org
265Description:
266		[RO] This file shows the number of keyslots the device has for
267		use with inline encryption.
268
269
270What:		/sys/block/<disk>/queue/dax
271Date:		June 2016
272Contact:	linux-block@vger.kernel.org
273Description:
274		[RO] This file indicates whether the device supports Direct
275		Access (DAX), used by CPU-addressable storage to bypass the
276		pagecache.  It shows '1' if true, '0' if not.
277
278
279What:		/sys/block/<disk>/queue/discard_granularity
280Date:		May 2011
281Contact:	Martin K. Petersen <martin.petersen@oracle.com>
282Description:
283		[RO] Devices that support discard functionality may internally
284		allocate space using units that are bigger than the logical
285		block size. The discard_granularity parameter indicates the size
286		of the internal allocation unit in bytes if reported by the
287		device. Otherwise the discard_granularity will be set to match
288		the device's physical block size. A discard_granularity of 0
289		means that the device does not support discard functionality.
290
291
292What:		/sys/block/<disk>/queue/discard_max_bytes
293Date:		May 2011
294Contact:	Martin K. Petersen <martin.petersen@oracle.com>
295Description:
296		[RW] While discard_max_hw_bytes is the hardware limit for the
297		device, this setting is the software limit. Some devices exhibit
298		large latencies when large discards are issued, setting this
299		value lower will make Linux issue smaller discards and
300		potentially help reduce latencies induced by large discard
301		operations.
302
303
304What:		/sys/block/<disk>/queue/discard_max_hw_bytes
305Date:		July 2015
306Contact:	linux-block@vger.kernel.org
307Description:
308		[RO] Devices that support discard functionality may have
309		internal limits on the number of bytes that can be trimmed or
310		unmapped in a single operation.  The `discard_max_hw_bytes`
311		parameter is set by the device driver to the maximum number of
312		bytes that can be discarded in a single operation.  Discard
313		requests issued to the device must not exceed this limit.  A
314		`discard_max_hw_bytes` value of 0 means that the device does not
315		support discard functionality.
316
317
318What:		/sys/block/<disk>/queue/discard_zeroes_data
319Date:		May 2011
320Contact:	Martin K. Petersen <martin.petersen@oracle.com>
321Description:
322		[RO] Will always return 0.  Don't rely on any specific behavior
323		for discards, and don't read this file.
324
325
326What:		/sys/block/<disk>/queue/dma_alignment
327Date:		May 2022
328Contact:	linux-block@vger.kernel.org
329Description:
330		Reports the alignment that user space addresses must have to be
331		used for raw block device access with O_DIRECT and other driver
332		specific passthrough mechanisms.
333
334
335What:		/sys/block/<disk>/queue/fua
336Date:		May 2018
337Contact:	linux-block@vger.kernel.org
338Description:
339		[RO] Whether or not the block driver supports the FUA flag for
340		write requests.  FUA stands for Force Unit Access. If the FUA
341		flag is set that means that write requests must bypass the
342		volatile cache of the storage device.
343
344
345What:		/sys/block/<disk>/queue/hw_sector_size
346Date:		January 2008
347Contact:	linux-block@vger.kernel.org
348Description:
349		[RO] This is the hardware sector size of the device, in bytes.
350
351
352What:		/sys/block/<disk>/queue/independent_access_ranges/
353Date:		October 2021
354Contact:	linux-block@vger.kernel.org
355Description:
356		[RO] The presence of this sub-directory of the
357		/sys/block/xxx/queue/ directory indicates that the device is
358		capable of executing requests targeting different sector ranges
359		in parallel. For instance, single LUN multi-actuator hard-disks
360		will have an independent_access_ranges directory if the device
361		correctly advertises the sector ranges of its actuators.
362
363		The independent_access_ranges directory contains one directory
364		per access range, with each range described using the sector
365		(RO) attribute file to indicate the first sector of the range
366		and the nr_sectors (RO) attribute file to indicate the total
367		number of sectors in the range starting from the first sector of
368		the range.  For example, a dual-actuator hard-disk will have the
369		following independent_access_ranges entries.::
370
371			$ tree /sys/block/<disk>/queue/independent_access_ranges/
372			/sys/block/<disk>/queue/independent_access_ranges/
373			|-- 0
374			|   |-- nr_sectors
375			|   `-- sector
376			`-- 1
377			    |-- nr_sectors
378			    `-- sector
379
380		The sector and nr_sectors attributes use 512B sector unit,
381		regardless of the actual block size of the device. Independent
382		access ranges do not overlap and include all sectors within the
383		device capacity. The access ranges are numbered in increasing
384		order of the range start sector, that is, the sector attribute
385		of range 0 always has the value 0.
386
387
388What:		/sys/block/<disk>/queue/io_poll
389Date:		November 2015
390Contact:	linux-block@vger.kernel.org
391Description:
392		[RW] When read, this file shows whether polling is enabled (1)
393		or disabled (0).  Writing '0' to this file will disable polling
394		for this device.  Writing any non-zero value will enable this
395		feature.
396
397
398What:		/sys/block/<disk>/queue/io_poll_delay
399Date:		November 2016
400Contact:	linux-block@vger.kernel.org
401Description:
402		[RW] This was used to control what kind of polling will be
403		performed.  It is now fixed to -1, which is classic polling.
404		In this mode, the CPU will repeatedly ask for completions
405		without giving up any time.
406		<deprecated>
407
408
409What:		/sys/block/<disk>/queue/io_timeout
410Date:		November 2018
411Contact:	Weiping Zhang <zhangweiping@didiglobal.com>
412Description:
413		[RW] io_timeout is the request timeout in milliseconds. If a
414		request does not complete in this time then the block driver
415		timeout handler is invoked. That timeout handler can decide to
416		retry the request, to fail it or to start a device recovery
417		strategy.
418
419
420What:		/sys/block/<disk>/queue/iostats
421Date:		January 2009
422Contact:	linux-block@vger.kernel.org
423Description:
424		[RW] This file is used to control (on/off) the iostats
425		accounting of the disk.
426
427
428What:		/sys/block/<disk>/queue/logical_block_size
429Date:		May 2009
430Contact:	Martin K. Petersen <martin.petersen@oracle.com>
431Description:
432		[RO] This is the smallest unit the storage device can address.
433		It is typically 512 bytes.
434
435
436What:		/sys/block/<disk>/queue/max_active_zones
437Date:		July 2020
438Contact:	Niklas Cassel <niklas.cassel@wdc.com>
439Description:
440		[RO] For zoned block devices (zoned attribute indicating
441		"host-managed" or "host-aware"), the sum of zones belonging to
442		any of the zone states: EXPLICIT OPEN, IMPLICIT OPEN or CLOSED,
443		is limited by this value. If this value is 0, there is no limit.
444
445		If the host attempts to exceed this limit, the driver should
446		report this error with BLK_STS_ZONE_ACTIVE_RESOURCE, which user
447		space may see as the EOVERFLOW errno.
448
449
450What:		/sys/block/<disk>/queue/max_discard_segments
451Date:		February 2017
452Contact:	linux-block@vger.kernel.org
453Description:
454		[RO] The maximum number of DMA scatter/gather entries in a
455		discard request.
456
457
458What:		/sys/block/<disk>/queue/max_hw_sectors_kb
459Date:		September 2004
460Contact:	linux-block@vger.kernel.org
461Description:
462		[RO] This is the maximum number of kilobytes supported in a
463		single data transfer.
464
465
466What:		/sys/block/<disk>/queue/max_integrity_segments
467Date:		September 2010
468Contact:	linux-block@vger.kernel.org
469Description:
470		[RO] Maximum number of elements in a DMA scatter/gather list
471		with integrity data that will be submitted by the block layer
472		core to the associated block driver.
473
474
475What:		/sys/block/<disk>/queue/max_open_zones
476Date:		July 2020
477Contact:	Niklas Cassel <niklas.cassel@wdc.com>
478Description:
479		[RO] For zoned block devices (zoned attribute indicating
480		"host-managed" or "host-aware"), the sum of zones belonging to
481		any of the zone states: EXPLICIT OPEN or IMPLICIT OPEN, is
482		limited by this value. If this value is 0, there is no limit.
483
484
485What:		/sys/block/<disk>/queue/max_sectors_kb
486Date:		September 2004
487Contact:	linux-block@vger.kernel.org
488Description:
489		[RW] This is the maximum number of kilobytes that the block
490		layer will allow for a filesystem request. Must be smaller than
491		or equal to the maximum size allowed by the hardware. Write 0
492		to use default kernel settings.
493
494
495What:		/sys/block/<disk>/queue/max_segment_size
496Date:		March 2010
497Contact:	linux-block@vger.kernel.org
498Description:
499		[RO] Maximum size in bytes of a single element in a DMA
500		scatter/gather list.
501
502
503What:		/sys/block/<disk>/queue/max_segments
504Date:		March 2010
505Contact:	linux-block@vger.kernel.org
506Description:
507		[RO] Maximum number of elements in a DMA scatter/gather list
508		that is submitted to the associated block driver.
509
510
511What:		/sys/block/<disk>/queue/minimum_io_size
512Date:		April 2009
513Contact:	Martin K. Petersen <martin.petersen@oracle.com>
514Description:
515		[RO] Storage devices may report a granularity or preferred
516		minimum I/O size which is the smallest request the device can
517		perform without incurring a performance penalty.  For disk
518		drives this is often the physical block size.  For RAID arrays
519		it is often the stripe chunk size.  A properly aligned multiple
520		of minimum_io_size is the preferred request size for workloads
521		where a high number of I/O operations is desired.
522
523
524What:		/sys/block/<disk>/queue/nomerges
525Date:		January 2010
526Contact:	linux-block@vger.kernel.org
527Description:
528		[RW] Standard I/O elevator operations include attempts to merge
529		contiguous I/Os. For known random I/O loads these attempts will
530		always fail and result in extra cycles being spent in the
531		kernel. This allows one to turn off this behavior on one of two
532		ways: When set to 1, complex merge checks are disabled, but the
533		simple one-shot merges with the previous I/O request are
534		enabled. When set to 2, all merge tries are disabled. The
535		default value is 0 - which enables all types of merge tries.
536
537
538What:		/sys/block/<disk>/queue/nr_requests
539Date:		July 2003
540Contact:	linux-block@vger.kernel.org
541Description:
542		[RW] This controls how many requests may be allocated in the
543		block layer for read or write requests. Note that the total
544		allocated number may be twice this amount, since it applies only
545		to reads or writes (not the accumulated sum).
546
547		To avoid priority inversion through request starvation, a
548		request queue maintains a separate request pool per each cgroup
549		when CONFIG_BLK_CGROUP is enabled, and this parameter applies to
550		each such per-block-cgroup request pool.  IOW, if there are N
551		block cgroups, each request queue may have up to N request
552		pools, each independently regulated by nr_requests.
553
554
555What:		/sys/block/<disk>/queue/nr_zones
556Date:		November 2018
557Contact:	Damien Le Moal <damien.lemoal@wdc.com>
558Description:
559		[RO] nr_zones indicates the total number of zones of a zoned
560		block device ("host-aware" or "host-managed" zone model). For
561		regular block devices, the value is always 0.
562
563
564What:		/sys/block/<disk>/queue/optimal_io_size
565Date:		April 2009
566Contact:	Martin K. Petersen <martin.petersen@oracle.com>
567Description:
568		[RO] Storage devices may report an optimal I/O size, which is
569		the device's preferred unit for sustained I/O.  This is rarely
570		reported for disk drives.  For RAID arrays it is usually the
571		stripe width or the internal track size.  A properly aligned
572		multiple of optimal_io_size is the preferred request size for
573		workloads where sustained throughput is desired.  If no optimal
574		I/O size is reported this file contains 0.
575
576
577What:		/sys/block/<disk>/queue/physical_block_size
578Date:		May 2009
579Contact:	Martin K. Petersen <martin.petersen@oracle.com>
580Description:
581		[RO] This is the smallest unit a physical storage device can
582		write atomically.  It is usually the same as the logical block
583		size but may be bigger.  One example is SATA drives with 4KB
584		sectors that expose a 512-byte logical block size to the
585		operating system.  For stacked block devices the
586		physical_block_size variable contains the maximum
587		physical_block_size of the component devices.
588
589
590What:		/sys/block/<disk>/queue/read_ahead_kb
591Date:		May 2004
592Contact:	linux-block@vger.kernel.org
593Description:
594		[RW] Maximum number of kilobytes to read-ahead for filesystems
595		on this block device.
596
597
598What:		/sys/block/<disk>/queue/rotational
599Date:		January 2009
600Contact:	linux-block@vger.kernel.org
601Description:
602		[RW] This file is used to stat if the device is of rotational
603		type or non-rotational type.
604
605
606What:		/sys/block/<disk>/queue/rq_affinity
607Date:		September 2008
608Contact:	linux-block@vger.kernel.org
609Description:
610		[RW] If this option is '1', the block layer will migrate request
611		completions to the cpu "group" that originally submitted the
612		request. For some workloads this provides a significant
613		reduction in CPU cycles due to caching effects.
614
615		For storage configurations that need to maximize distribution of
616		completion processing setting this option to '2' forces the
617		completion to run on the requesting cpu (bypassing the "group"
618		aggregation logic).
619
620
621What:		/sys/block/<disk>/queue/scheduler
622Date:		October 2004
623Contact:	linux-block@vger.kernel.org
624Description:
625		[RW] When read, this file will display the current and available
626		IO schedulers for this block device. The currently active IO
627		scheduler will be enclosed in [] brackets. Writing an IO
628		scheduler name to this file will switch control of this block
629		device to that new IO scheduler. Note that writing an IO
630		scheduler name to this file will attempt to load that IO
631		scheduler module, if it isn't already present in the system.
632
633
634What:		/sys/block/<disk>/queue/stable_writes
635Date:		September 2020
636Contact:	linux-block@vger.kernel.org
637Description:
638		[RW] This file will contain '1' if memory must not be modified
639		while it is being used in a write request to this device.  When
640		this is the case and the kernel is performing writeback of a
641		page, the kernel will wait for writeback to complete before
642		allowing the page to be modified again, rather than allowing
643		immediate modification as is normally the case.  This
644		restriction arises when the device accesses the memory multiple
645		times where the same data must be seen every time -- for
646		example, once to calculate a checksum and once to actually write
647		the data.  If no such restriction exists, this file will contain
648		'0'.  This file is writable for testing purposes.
649
650What:		/sys/block/<disk>/queue/virt_boundary_mask
651Date:		April 2021
652Contact:	linux-block@vger.kernel.org
653Description:
654		[RO] This file shows the I/O segment memory alignment mask for
655		the block device.  I/O requests to this device will be split
656		between segments wherever either the memory address of the end
657		of the previous segment or the memory address of the beginning
658		of the current segment is not aligned to virt_boundary_mask + 1
659		bytes.
660
661
662What:		/sys/block/<disk>/queue/wbt_lat_usec
663Date:		November 2016
664Contact:	linux-block@vger.kernel.org
665Description:
666		[RW] If the device is registered for writeback throttling, then
667		this file shows the target minimum read latency. If this latency
668		is exceeded in a given window of time (see wb_window_usec), then
669		the writeback throttling will start scaling back writes. Writing
670		a value of '0' to this file disables the feature. Writing a
671		value of '-1' to this file resets the value to the default
672		setting.
673
674
675What:		/sys/block/<disk>/queue/write_cache
676Date:		April 2016
677Contact:	linux-block@vger.kernel.org
678Description:
679		[RW] When read, this file will display whether the device has
680		write back caching enabled or not. It will return "write back"
681		for the former case, and "write through" for the latter. Writing
682		to this file can change the kernels view of the device, but it
683		doesn't alter the device state. This means that it might not be
684		safe to toggle the setting from "write back" to "write through",
685		since that will also eliminate cache flushes issued by the
686		kernel.
687
688
689What:		/sys/block/<disk>/queue/write_same_max_bytes
690Date:		January 2012
691Contact:	Martin K. Petersen <martin.petersen@oracle.com>
692Description:
693		[RO] Some devices support a write same operation in which a
694		single data block can be written to a range of several
695		contiguous blocks on storage. This can be used to wipe areas on
696		disk or to initialize drives in a RAID configuration.
697		write_same_max_bytes indicates how many bytes can be written in
698		a single write same command. If write_same_max_bytes is 0, write
699		same is not supported by the device.
700
701
702What:		/sys/block/<disk>/queue/write_zeroes_max_bytes
703Date:		November 2016
704Contact:	Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
705Description:
706		[RO] Devices that support write zeroes operation in which a
707		single request can be issued to zero out the range of contiguous
708		blocks on storage without having any payload in the request.
709		This can be used to optimize writing zeroes to the devices.
710		write_zeroes_max_bytes indicates how many bytes can be written
711		in a single write zeroes command. If write_zeroes_max_bytes is
712		0, write zeroes is not supported by the device.
713
714
715What:		/sys/block/<disk>/queue/zone_append_max_bytes
716Date:		May 2020
717Contact:	linux-block@vger.kernel.org
718Description:
719		[RO] This is the maximum number of bytes that can be written to
720		a sequential zone of a zoned block device using a zone append
721		write operation (REQ_OP_ZONE_APPEND). This value is always 0 for
722		regular block devices.
723
724
725What:		/sys/block/<disk>/queue/zone_write_granularity
726Date:		January 2021
727Contact:	linux-block@vger.kernel.org
728Description:
729		[RO] This indicates the alignment constraint, in bytes, for
730		write operations in sequential zones of zoned block devices
731		(devices with a zoned attributed that reports "host-managed" or
732		"host-aware"). This value is always 0 for regular block devices.
733
734
735What:		/sys/block/<disk>/queue/zoned
736Date:		September 2016
737Contact:	Damien Le Moal <damien.lemoal@wdc.com>
738Description:
739		[RO] zoned indicates if the device is a zoned block device and
740		the zone model of the device if it is indeed zoned.  The
741		possible values indicated by zoned are "none" for regular block
742		devices and "host-aware" or "host-managed" for zoned block
743		devices. The characteristics of host-aware and host-managed
744		zoned block devices are described in the ZBC (Zoned Block
745		Commands) and ZAC (Zoned Device ATA Command Set) standards.
746		These standards also define the "drive-managed" zone model.
747		However, since drive-managed zoned block devices do not support
748		zone commands, they will be treated as regular block devices and
749		zoned will report "none".
750
751
752What:		/sys/block/<disk>/hidden
753Date:		March 2023
754Contact:	linux-block@vger.kernel.org
755Description:
756		[RO] the block device is hidden. it doesn’t produce events, and
757		can’t be opened from userspace or using blkdev_get*.
758		Used for the underlying components of multipath devices.
759
760
761What:		/sys/block/<disk>/stat
762Date:		February 2008
763Contact:	Jerome Marchand <jmarchan@redhat.com>
764Description:
765		The /sys/block/<disk>/stat files displays the I/O
766		statistics of disk <disk>. They contain 11 fields:
767
768		==  ==============================================
769		 1  reads completed successfully
770		 2  reads merged
771		 3  sectors read
772		 4  time spent reading (ms)
773		 5  writes completed
774		 6  writes merged
775		 7  sectors written
776		 8  time spent writing (ms)
777		 9  I/Os currently in progress
778		10  time spent doing I/Os (ms)
779		11  weighted time spent doing I/Os (ms)
780		12  discards completed
781		13  discards merged
782		14  sectors discarded
783		15  time spent discarding (ms)
784		16  flush requests completed
785		17  time spent flushing (ms)
786		==  ==============================================
787
788		For more details refer Documentation/admin-guide/iostats.rst
789