1.\" 2.\" Copyright (c) 2003 Robert N. M. Watson 3.\" All rights reserved. 4.\" 5.\" Redistribution and use in source and binary forms, with or without 6.\" modification, are permitted provided that the following conditions 7.\" are met: 8.\" 1. Redistributions of source code must retain the above copyright 9.\" notice(s), this list of conditions and the following disclaimer as 10.\" the first lines of this file unmodified other than the possible 11.\" addition of one or more copyright notices. 12.\" 2. Redistributions in binary form must reproduce the above copyright 13.\" notice(s), this list of conditions and the following disclaimer in the 14.\" documentation and/or other materials provided with the distribution. 15.\" 16.\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) ``AS IS'' AND ANY 17.\" EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 18.\" WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 19.\" DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY 20.\" DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 21.\" (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR 22.\" SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 23.\" CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH 26.\" DAMAGE. 27.\" 28.\" $FreeBSD$ 29.\" 30.Dd April 30, 2020 31.Dt DISK 9 32.Os 33.Sh NAME 34.Nm disk 35.Nd kernel disk storage API 36.Sh SYNOPSIS 37.In geom/geom_disk.h 38.Ft struct disk * 39.Fn disk_alloc void 40.Ft void 41.Fn disk_create "struct disk *disk" "int version" 42.Ft void 43.Fn disk_gone "struct disk *disk" 44.Ft void 45.Fn disk_destroy "struct disk *disk" 46.Ft int 47.Fn disk_resize "struct disk *disk" "int flags" 48.Ft void 49.Fn disk_add_alias "struct disk *disk" "const char *alias" 50.Sh DESCRIPTION 51The disk storage API permits kernel device drivers providing access to 52disk-like storage devices to advertise the device to other kernel 53components, including 54.Xr GEOM 4 55and 56.Xr devfs 5 . 57.Pp 58Each disk device is described by a 59.Vt "struct disk" 60structure, which contains a variety of parameters for the disk device, 61function pointers for various methods that may be performed on the device, 62as well as private data storage for the device driver. 63In addition, some fields are reserved for use by GEOM in managing access 64to the device and its statistics. 65.Pp 66GEOM has the ownership of 67.Vt "struct disk" , 68and drivers must allocate storage for it with the 69.Fn disk_alloc 70function, 71fill in the fields and call 72.Fn disk_create 73when the device is ready to service requests. 74.Fn disk_add_alias 75adds an alias for the disk and must be called before 76.Fn disk_create , 77but may be called multiple times. 78For each alias added, a device node will be created with 79.Xr make_dev_alias 9 80in the same way primary device nodes are created with 81.Xr make_dev 9 82for 83.Va d_name 84and 85.Va d_unit . 86Care should be taken to ensure that only one driver creates aliases 87for any given name. 88.Fn disk_resize 89can be called by the driver after modifying 90.Va d_mediasize 91to notify GEOM about the disk capacity change. 92The 93.Fa flags 94field should be set to either M_WAITOK, or M_NOWAIT. 95.Fn disk_gone 96orphans all of the providers associated with the drive, setting an error 97condition of ENXIO in each one. 98In addition, it prevents a re-taste on last close for writing if an error 99condition has been set in the provider. 100After calling 101.Fn disk_destroy , 102the device driver is not allowed to access the contents of 103.Vt "struct disk" 104anymore. 105.Pp 106The 107.Fn disk_create 108function 109takes a second parameter, 110.Fa version , 111which must always be passed 112.Dv DISK_VERSION . 113If GEOM detects that the driver is compiled against an unsupported version, 114it will ignore the device and print a warning on the console. 115.Ss Descriptive Fields 116The following fields identify the disk device described by the structure 117instance, and must be filled in prior to submitting the structure to 118.Fn disk_create 119and may not be subsequently changed: 120.Bl -tag -width indent 121.It Vt u_int Va d_flags 122Optional flags indicating to the storage framework what optional features 123or descriptions the storage device driver supports. 124Currently supported flags are 125.Dv DISKFLAG_OPEN 126(maintained by storage framework), 127.Dv DISKFLAG_CANDELETE 128(maintained by device driver), 129and 130.Dv DISKFLAG_CANFLUSHCACHE 131(maintained by device driver). 132.It Vt "const char *" Va d_name 133Holds the name of the storage device class, e.g., 134.Dq Li ahd . 135This value typically uniquely identifies a particular driver device, 136and must not conflict with devices serviced by other device drivers. 137.It Vt u_int Va d_unit 138Holds the instance of the storage device class, e.g., 139.Dq Li 4 . 140This namespace is managed by the device driver, and assignment of unit 141numbers might be a property of probe order, or in some cases topology. 142Together, the 143.Va d_name 144and 145.Va d_unit 146values will uniquely identify a disk storage device. 147.El 148.Ss Disk Device Methods 149The following fields identify various disk device methods, if implemented: 150.Bl -tag -width indent 151.It Vt "disk_open_t *" Va d_open 152Optional: invoked when the disk device is opened. 153If no method is provided, open will always succeed. 154.It Vt "disk_close_t *" Va d_close 155Optional: invoked when the disk device is closed. 156Although an error code may be returned, the call should always terminate 157any state setup by the corresponding open method call. 158.It Vt "disk_strategy_t *" Va d_strategy 159Mandatory: invoked when a new 160.Vt "struct bio" 161is to be initiated on the disk device. 162.It Vt "disk_ioctl_t *" Va d_ioctl 163Optional: invoked when an I/O control operation is initiated on the disk device. 164Please note that for security reasons these operations should not 165be able to affect other devices than the one on which they are performed. 166.It Vt "dumper_t *" Va d_dump 167Optional: if configured with 168.Xr dumpon 8 , 169this function is invoked from a very restricted system state after a 170kernel panic to record a copy of the system RAM to the disk. 171.It Vt "disk_getattr_t *" Va d_getattr 172Optional: if this method is provided, it gives the disk driver the 173opportunity to override the default GEOM response to BIO_GETATTR requests. 174This function should return -1 if the attribute is not handled, 0 if the 175attribute is handled, or an errno to be passed to 176.Fn g_io_deliver . 177.It Vt "disk_gone_t *" Va d_gone 178Optional: if this method is provided, it will be called after 179.Fn disk_gone 180is called, once GEOM has finished its cleanup process. 181Once this callback is called, it is safe for the disk driver to free all of 182its resources, as it will not be receiving further calls from GEOM. 183.El 184.Ss Mandatory Media Properties 185The following fields identify the size and granularity of the disk device. 186These fields must stay stable from return of the drivers open method until 187the close method is called, but it is perfectly legal to modify them in 188the open method before returning. 189.Bl -tag -width indent 190.It Vt u_int Va d_sectorsize 191The sector size of the disk device in bytes. 192.It Vt off_t Va d_mediasize 193The size of the disk device in bytes. 194.It Vt u_int Va d_maxsize 195The maximum supported size in bytes of an I/O request. 196Requests larger than this size will be chopped up by GEOM. 197.El 198.Ss Optional Media Properties 199These optional fields can provide extra information about the disk 200device. 201Do not initialize these fields if the field/concept does not apply. 202These fields must stay stable from return of the drivers open method until 203the close method is called, but it is perfectly legal to modify them in 204the open method before returning. 205.Bl -tag -width indent 206.It Vt u_int Va d_fwsectors , Vt u_int Va d_fwheads 207The number of sectors and heads advertised on the disk device by the 208firmware or BIOS. 209These values are almost universally bogus, but on some architectures 210necessary for the correct calculation of disk partitioning. 211.It Vt u_int Va d_stripeoffset , Vt u_int Va d_stripesize 212These two fields can be used to describe the width and location of 213natural performance boundaries for most disk technologies. 214Please see 215.Pa src/sys/geom/notes 216for details. 217.It Vt char Va d_ident[DISK_IDENT_SIZE] 218This field can and should be used to store disk's serial number if the 219d_getattr method described above isn't implemented, or if it does not 220support the GEOM::ident attribute. 221.It Vt char Va d_descr[DISK_IDENT_SIZE] 222This field can be used to store the disk vendor and product description. 223.It Vt uint16_t Va d_hba_vendor 224This field can be used to store the PCI vendor ID for the HBA connected to 225the disk. 226.It Vt uint16_t Va d_hba_device 227This field can be used to store the PCI device ID for the HBA connected to 228the disk. 229.It Vt uint16_t Va d_hba_subvendor 230This field can be used to store the PCI subvendor ID for the HBA connected to 231the disk. 232.It Vt uint16_t Va d_hba_subdevice 233This field can be used to store the PCI subdevice ID for the HBA connected to 234the disk. 235.El 236.Ss Driver Private Data 237This field may be used by the device driver to store a pointer to 238private data to implement the disk service. 239.Bl -tag -width indent 240.It Vt "void *" Va d_drv1 241Private data pointer. 242Typically used to store a pointer to the drivers 243.Vt softc 244structure for this disk device. 245.El 246.Sh SEE ALSO 247.Xr GEOM 4 , 248.Xr devfs 5 249.Sh HISTORY 250The 251.Nm kernel disk storage API 252first appeared in 253.Fx 4.9 . 254.Sh AUTHORS 255This manual page was written by 256.An Robert Watson . 257.Sh BUGS 258Disk aliases are not a general purpose aliasing mechanism, but are 259intended only to ease the transition from one name to another. 260They can be used to ensure that nvd0 and nda0 are the same thing. 261They cannot be used to implement the diskX concept from macOS. 262