1 Linux Kernel 2.6 series
2 SCSI mid_level - lower_level driver interface
3 =============================================
7 This document outlines the interface between the Linux SCSI mid level and
8 SCSI lower level drivers. Lower level drivers (LLDs) are variously called
9 host bus adapter (HBA) drivers and host drivers (HD). A "host" in this
10 context is a bridge between a computer IO bus (e.g. PCI or ISA) and a
11 single SCSI initiator port on a SCSI transport. An "initiator" port
12 (SCSI terminology, see SAM-3 at http://www.t10.org) sends SCSI commands
13 to "target" SCSI ports (e.g. disks). There can be many LLDs in a running
14 system, but only one per hardware type. Most LLDs can control one or more
15 SCSI HBAs. Some HBAs contain multiple hosts.
17 In some cases the SCSI transport is an external bus that already has
18 its own subsystem in Linux (e.g. USB and ieee1394). In such cases the
19 SCSI subsystem LLD is a software bridge to the other driver subsystem.
20 Examples are the usb-storage driver (found in the drivers/usb/storage
21 directory) and the ieee1394/sbp2 driver (found in the drivers/ieee1394
24 For example, the aic7xxx LLD controls Adaptec SCSI parallel interface
25 (SPI) controllers based on that company's 7xxx chip series. The aic7xxx
26 LLD can be built into the kernel or loaded as a module. There can only be
27 one aic7xxx LLD running in a Linux system but it may be controlling many
28 HBAs. These HBAs might be either on PCI daughter-boards or built into
29 the motherboard (or both). Some aic7xxx based HBAs are dual controllers
30 and thus represent two hosts. Like most modern HBAs, each aic7xxx host
31 has its own PCI device address. [The one-to-one correspondence between
32 a SCSI host and a PCI device is common but not required (e.g. with
33 ISA or MCA adapters).]
35 The SCSI mid level isolates an LLD from other layers such as the SCSI
36 upper layer drivers and the block layer.
38 This version of the document roughly matches linux kernel version 2.6.8 .
42 There is a SCSI documentation directory within the kernel source tree,
43 typically Documentation/scsi . Most documents are in plain
44 (i.e. ASCII) text. This file is named scsi_mid_low_api.txt and can be
45 found in that directory. A more recent copy of this document may be found
46 at http://www.torque.net/scsi/scsi_mid_low_api.txt.gz .
47 Many LLDs are documented there (e.g. aic7xxx.txt). The SCSI mid-level is
48 briefly described in scsi.txt which contains a url to a document
49 describing the SCSI subsystem in the lk 2.4 series. Two upper level
50 drivers have documents in that directory: st.txt (SCSI tape driver) and
51 scsi-generic.txt (for the sg driver).
53 Some documentation (or urls) for LLDs may be found in the C source code
54 or in the same directory as the C source code. For example to find a url
55 about the USB mass storage driver see the
56 /usr/src/linux/drivers/usb/storage directory.
58 The Linux kernel source Documentation/DocBook/scsidrivers.tmpl file
59 refers to this file. With the appropriate DocBook tool-set, this permits
60 users to generate html, ps and pdf renderings of information within this
61 file (e.g. the interface functions).
65 Traditionally an LLD for the SCSI subsystem has been at least two files in
66 the drivers/scsi directory. For example, a driver called "xyz" has a header
67 file "xyz.h" and a source file "xyz.c". [Actually there is no good reason
68 why this couldn't all be in one file; the header file is superfluous.] Some
69 drivers that have been ported to several operating systems have more than
70 two files. For example the aic7xxx driver has separate files for generic
71 and OS-specific code (e.g. FreeBSD and Linux). Such drivers tend to have
72 their own directory under the drivers/scsi directory.
74 When a new LLD is being added to Linux, the following files (found in the
75 drivers/scsi directory) will need some attention: Makefile and Kconfig .
76 It is probably best to study how existing LLDs are organized.
78 As the 2.5 series development kernels evolve into the 2.6 series
79 production series, changes are being introduced into this interface. An
80 example of this is driver initialization code where there are now 2 models
81 available. The older one, similar to what was found in the lk 2.4 series,
82 is based on hosts that are detected at HBA driver load time. This will be
83 referred to the "passive" initialization model. The newer model allows HBAs
84 to be hot plugged (and unplugged) during the lifetime of the LLD and will
85 be referred to as the "hotplug" initialization model. The newer model is
86 preferred as it can handle both traditional SCSI equipment that is
87 permanently connected as well as modern "SCSI" devices (e.g. USB or
88 IEEE 1394 connected digital cameras) that are hotplugged. Both
89 initialization models are discussed in the following sections.
91 An LLD interfaces to the SCSI subsystem several ways:
92 a) directly invoking functions supplied by the mid level
93 b) passing a set of function pointers to a registration function
94 supplied by the mid level. The mid level will then invoke these
95 functions at some point in the future. The LLD will supply
96 implementations of these functions.
97 c) direct access to instances of well known data structures maintained
100 Those functions in group a) are listed in a section entitled "Mid level
101 supplied functions" below.
103 Those functions in group b) are listed in a section entitled "Interface
104 functions" below. Their function pointers are placed in the members of
105 "struct scsi_host_template", an instance of which is passed to
106 scsi_host_alloc() ** . Those interface functions that the LLD does not
107 wish to supply should have NULL placed in the corresponding member of
108 struct scsi_host_template. Defining an instance of struct
109 scsi_host_template at file scope will cause NULL to be placed in function
110 pointer members not explicitly initialized.
112 Those usages in group c) should be handled with care, especially in a
113 "hotplug" environment. LLDs should be aware of the lifetime of instances
114 that are shared with the mid level and other layers.
116 All functions defined within an LLD and all data defined at file scope
117 should be static. For example the slave_alloc() function in an LLD
118 called "xxx" could be defined as
119 "static int xxx_slave_alloc(struct scsi_device * sdev) { /* code */ }"
121 ** the scsi_host_alloc() function is a replacement for the rather vaguely
122 named scsi_register() function in most situations. The scsi_register()
123 and scsi_unregister() functions remain to support legacy LLDs that use
124 the passive initialization model.
127 Hotplug initialization model
128 ============================
129 In this model an LLD controls when SCSI hosts are introduced and removed
130 from the SCSI subsystem. Hosts can be introduced as early as driver
131 initialization and removed as late as driver shutdown. Typically a driver
132 will respond to a sysfs probe() callback that indicates an HBA has been
133 detected. After confirming that the new device is one that the LLD wants
134 to control, the LLD will initialize the HBA and then register a new host
135 with the SCSI mid level.
137 During LLD initialization the driver should register itself with the
138 appropriate IO bus on which it expects to find HBA(s) (e.g. the PCI bus).
139 This can probably be done via sysfs. Any driver parameters (especially
140 those that are writable after the driver is loaded) could also be
141 registered with sysfs at this point. The SCSI mid level first becomes
142 aware of an LLD when that LLD registers its first HBA.
144 At some later time, the LLD becomes aware of an HBA and what follows
145 is a typical sequence of calls between the LLD and the mid level.
146 This example shows the mid level scanning the newly introduced HBA for 3
147 scsi devices of which only the first 2 respond:
149 HBA PROBE: assume 2 SCSI devices found in scan
151 ===-------------------=========--------------------===------
152 scsi_host_alloc() -->
153 scsi_add_host() ---->
154 scsi_scan_host() -------+
157 slave_configure() --> scsi_adjust_queue_depth()
164 ------------------------------------------------------------
166 If the LLD wants to adjust the default queue settings, it can invoke
167 scsi_adjust_queue_depth() in its slave_configure() routine.
169 *** For scsi devices that the mid level tries to scan but do not
170 respond, a slave_alloc(), slave_destroy() pair is called.
172 When an HBA is being removed it could be as part of an orderly shutdown
173 associated with the LLD module being unloaded (e.g. with the "rmmod"
174 command) or in response to a "hot unplug" indicated by sysfs()'s
175 remove() callback being invoked. In either case, the sequence is the
178 HBA REMOVE: assume 2 SCSI devices attached
180 ===----------------------=========-----------------===------
181 scsi_remove_host() ---------+
186 ------------------------------------------------------------
188 It may be useful for a LLD to keep track of struct Scsi_Host instances
189 (a pointer is returned by scsi_host_alloc()). Such instances are "owned"
190 by the mid-level. struct Scsi_Host instances are freed from
191 scsi_host_put() when the reference count hits zero.
193 Hot unplugging an HBA that controls a disk which is processing SCSI
194 commands on a mounted file system is an interesting situation. Reference
195 counting logic is being introduced into the mid level to cope with many
196 of the issues involved. See the section on reference counting below.
199 The hotplug concept may be extended to SCSI devices. Currently, when an
200 HBA is added, the scsi_scan_host() function causes a scan for SCSI devices
201 attached to the HBA's SCSI transport. On newer SCSI transports the HBA
202 may become aware of a new SCSI device _after_ the scan has completed.
203 An LLD can use this sequence to make the mid level aware of a SCSI device:
207 ===-------------------=========--------------------===------
208 scsi_add_device() ------+
211 slave_configure() [--> scsi_adjust_queue_depth()]
212 ------------------------------------------------------------
214 In a similar fashion, an LLD may become aware that a SCSI device has been
215 removed (unplugged) or the connection to it has been interrupted. Some
216 existing SCSI transports (e.g. SPI) may not become aware that a SCSI
217 device has been removed until a subsequent SCSI command fails which will
218 probably cause that device to be set offline by the mid level. An LLD that
219 detects the removal of a SCSI device can instigate its removal from
220 upper layers with this sequence:
222 SCSI DEVICE hot unplug
224 ===----------------------=========-----------------===------
225 scsi_remove_device() -------+
228 ------------------------------------------------------------
230 It may be useful for an LLD to keep track of struct scsi_device instances
231 (a pointer is passed as the parameter to slave_alloc() and
232 slave_configure() callbacks). Such instances are "owned" by the mid-level.
233 struct scsi_device instances are freed after slave_destroy().
236 Passive initialization model
237 ============================
238 These older LLDs include a file called "scsi_module.c" [yes the ".c" is a
239 little surprising] in their source code. For that file to work an
240 instance of struct scsi_host_template with the name "driver_template"
241 needs to be defined. Here is a typical code sequence used in this model:
242 static struct scsi_host_template driver_template = {
245 #include "scsi_module.c"
247 The scsi_module.c file contains two functions:
248 - init_this_scsi_driver() which is executed when the LLD is
249 initialized (i.e. boot time or module load time)
250 - exit_this_scsi_driver() which is executed when the LLD is shut
251 down (i.e. module unload time)
252 Note: since these functions are tagged with __init and __exit qualifiers
253 an LLD should not call them explicitly (since the kernel does that).
255 Here is an example of an initialization sequence when two hosts are
256 detected (so detect() returns 2) and the SCSI bus scan on each host
257 finds 1 SCSI device (and a second device does not respond).
260 ===----------------------=========-----------------===------
261 init_this_scsi_driver() ----+
263 detect() -----------------+
269 slave_configure() --> scsi_adjust_queue_depth()
277 ------------------------------------------------------------
279 The mid level invokes scsi_adjust_queue_depth() with tagged queuing off and
280 "cmd_per_lun" for that host as the queue length. These settings can be
281 overridden by a slave_configure() supplied by the LLD.
283 *** For scsi devices that the mid level tries to scan but do not
284 respond, a slave_alloc(), slave_destroy() pair is called.
286 Here is an LLD shutdown sequence:
289 ===----------------------=========-----------------===------
290 exit_this_scsi_driver() ----+
293 release() --> scsi_unregister()
296 release() --> scsi_unregister()
297 ------------------------------------------------------------
299 An LLD need not define slave_destroy() (i.e. it is optional).
301 The shortcoming of the "passive initialization model" is that host
302 registration and de-registration are (typically) tied to LLD initialization
303 and shutdown. Once the LLD is initialized then a new host that appears
304 (e.g. via hotplugging) cannot easily be added without a redundant
305 driver shutdown and re-initialization. It may be possible to write an LLD
306 that uses both initialization models.
311 The Scsi_Host structure has had reference counting infrastructure added.
312 This effectively spreads the ownership of struct Scsi_Host instances
313 across the various SCSI layers which use them. Previously such instances
314 were exclusively owned by the mid level. LLDs would not usually need to
315 directly manipulate these reference counts but there may be some cases
318 There are 3 reference counting functions of interest associated with
320 - scsi_host_alloc(): returns a pointer to new instance of struct
321 Scsi_Host which has its reference count ^^ set to 1
322 - scsi_host_get(): adds 1 to the reference count of the given instance
323 - scsi_host_put(): decrements 1 from the reference count of the given
324 instance. If the reference count reaches 0 then the given instance
327 The Scsi_device structure has had reference counting infrastructure added.
328 This effectively spreads the ownership of struct Scsi_device instances
329 across the various SCSI layers which use them. Previously such instances
330 were exclusively owned by the mid level. See the access functions declared
331 towards the end of include/scsi/scsi_device.h . If an LLD wants to keep
332 a copy of a pointer to a Scsi_device instance it should use scsi_device_get()
333 to bump its reference count. When it is finished with the pointer it can
334 use scsi_device_put() to decrement its reference count (and potentially
337 ^^ struct Scsi_Host actually has 2 reference counts which are manipulated
338 in parallel by these functions.
343 First, Linus Torvalds's thoughts on C coding style can be found in the
344 Documentation/CodingStyle file.
346 Next, there is a movement to "outlaw" typedefs introducing synonyms for
347 struct tags. Both can be still found in the SCSI subsystem, but
348 the typedefs have been moved to a single file, scsi_typedefs.h to
349 make their future removal easier, for example:
350 "typedef struct scsi_cmnd Scsi_Cmnd;"
352 Also, most C99 enhancements are encouraged to the extent they are supported
353 by the relevant gcc compilers. So C99 style structure and array
354 initializers are encouraged where appropriate. Don't go too far,
355 VLAs are not properly supported yet. An exception to this is the use of
356 "//" style comments; /*...*/ comments are still preferred in Linux.
358 Well written, tested and documented code, need not be re-formatted to
359 comply with the above conventions. For example, the aic7xxx driver
360 comes to Linux from FreeBSD and Adaptec's own labs. No doubt FreeBSD
361 and Adaptec have their own coding conventions.
364 Mid level supplied functions
365 ============================
366 These functions are supplied by the SCSI mid level for use by LLDs.
367 The names (i.e. entry points) of these functions are exported
368 so an LLD that is a module can access them. The kernel will
369 arrange for the SCSI mid level to be loaded and initialized before any LLD
370 is initialized. The functions below are listed alphabetically and their
371 names all start with "scsi_".
374 scsi_activate_tcq - turn on tag command queueing
375 scsi_add_device - creates new scsi device (lu) instance
376 scsi_add_host - perform sysfs registration and set up transport class
377 scsi_adjust_queue_depth - change the queue depth on a SCSI device
378 scsi_assign_lock - replace default host_lock with given lock
379 scsi_bios_ptable - return copy of block device's partition table
380 scsi_block_requests - prevent further commands being queued to given host
381 scsi_deactivate_tcq - turn off tag command queueing
382 scsi_host_alloc - return a new scsi_host instance whose refcount==1
383 scsi_host_get - increments Scsi_Host instance's refcount
384 scsi_host_put - decrements Scsi_Host instance's refcount (free if 0)
385 scsi_partsize - parse partition table into cylinders, heads + sectors
386 scsi_register - create and register a scsi host adapter instance.
387 scsi_remove_device - detach and remove a SCSI device
388 scsi_remove_host - detach and remove all SCSI devices owned by host
389 scsi_report_bus_reset - report scsi _bus_ reset observed
390 scsi_scan_host - scan SCSI bus
391 scsi_track_queue_full - track successive QUEUE_FULL events
392 scsi_unblock_requests - allow further commands to be queued to given host
393 scsi_unregister - [calls scsi_host_put()]
399 * scsi_activate_tcq - turn on tag command queueing ("ordered" task attribute)
400 * @sdev: device to turn on TCQ for
401 * @depth: queue depth
407 * Notes: Eventually, it is hoped depth would be the maximum depth
408 * the device could cope with and the real queue depth
409 * would be adjustable from 0 to depth.
411 * Defined (inline) in: include/scsi/scsi_tcq.h
413 void scsi_activate_tcq(struct scsi_device *sdev, int depth)
417 * scsi_add_device - creates new scsi device (lu) instance
418 * @shost: pointer to scsi host instance
419 * @channel: channel number (rarely other than 0)
420 * @id: target id number
421 * @lun: logical unit number
423 * Returns pointer to new struct scsi_device instance or
424 * ERR_PTR(-ENODEV) (or some other bent pointer) if something is
425 * wrong (e.g. no lu responds at given address)
429 * Notes: This call is usually performed internally during a scsi
430 * bus scan when an HBA is added (i.e. scsi_scan_host()). So it
431 * should only be called if the HBA becomes aware of a new scsi
432 * device (lu) after scsi_scan_host() has completed. If successful
433 * this call can lead to slave_alloc() and slave_configure() callbacks
436 * Defined in: drivers/scsi/scsi_scan.c
438 struct scsi_device * scsi_add_device(struct Scsi_Host *shost,
439 unsigned int channel,
440 unsigned int id, unsigned int lun)
444 * scsi_add_host - perform sysfs registration and set up transport class
445 * @shost: pointer to scsi host instance
446 * @dev: pointer to struct device of type scsi class
448 * Returns 0 on success, negative errno of failure (e.g. -ENOMEM)
452 * Notes: Only required in "hotplug initialization model" after a
453 * successful call to scsi_host_alloc(). This function does not
454 * scan the bus; this can be done by calling scsi_scan_host() or
455 * in some other transport-specific way. The LLD must set up
456 * the transport template before calling this function and may only
457 * access the transport class data after this function has been called.
459 * Defined in: drivers/scsi/hosts.c
461 int scsi_add_host(struct Scsi_Host *shost, struct device * dev)
465 * scsi_adjust_queue_depth - allow LLD to change queue depth on a SCSI device
466 * @sdev: pointer to SCSI device to change queue depth on
467 * @tagged: 0 - no tagged queuing
468 * MSG_SIMPLE_TAG - simple tagged queuing
469 * MSG_ORDERED_TAG - ordered tagged queuing
470 * @tags Number of tags allowed if tagged queuing enabled,
471 * or number of commands the LLD can queue up
472 * in non-tagged mode (as per cmd_per_lun).
478 * Notes: Can be invoked any time on a SCSI device controlled by this
479 * LLD. [Specifically during and after slave_configure() and prior to
480 * slave_destroy().] Can safely be invoked from interrupt code. Actual
481 * queue depth change may be delayed until the next command is being
482 * processed. See also scsi_activate_tcq() and scsi_deactivate_tcq().
484 * Defined in: drivers/scsi/scsi.c [see source code for more notes]
487 void scsi_adjust_queue_depth(struct scsi_device * sdev, int tagged,
492 * scsi_assign_lock - replace default host_lock with given lock
493 * @shost: a pointer to a scsi host instance
494 * @lock: pointer to lock to replace host_lock for this host
500 * Defined in: include/scsi/scsi_host.h .
502 void scsi_assign_lock(struct Scsi_Host *shost, spinlock_t *lock)
506 * scsi_bios_ptable - return copy of block device's partition table
507 * @dev: pointer to block device
509 * Returns pointer to partition table, or NULL for failure
513 * Notes: Caller owns memory returned (free with kfree() )
515 * Defined in: drivers/scsi/scsicam.c
517 unsigned char *scsi_bios_ptable(struct block_device *dev)
521 * scsi_block_requests - prevent further commands being queued to given host
523 * @shost: pointer to host to block commands on
529 * Notes: There is no timer nor any other means by which the requests
530 * get unblocked other than the LLD calling scsi_unblock_requests().
532 * Defined in: drivers/scsi/scsi_lib.c
534 void scsi_block_requests(struct Scsi_Host * shost)
538 * scsi_deactivate_tcq - turn off tag command queueing
539 * @sdev: device to turn off TCQ for
540 * @depth: queue depth (stored in sdev)
546 * Defined (inline) in: include/scsi/scsi_tcq.h
548 void scsi_deactivate_tcq(struct scsi_device *sdev, int depth)
552 * scsi_host_alloc - create a scsi host adapter instance and perform basic
554 * @sht: pointer to scsi host template
555 * @privsize: extra bytes to allocate in hostdata array (which is the
556 * last member of the returned Scsi_Host instance)
558 * Returns pointer to new Scsi_Host instance or NULL on failure
562 * Notes: When this call returns to the LLD, the SCSI bus scan on
563 * this host has _not_ yet been done.
564 * The hostdata array (by default zero length) is a per host scratch
565 * area for the LLD's exclusive use.
566 * Both associated refcounting objects have their refcount set to 1.
567 * Full registration (in sysfs) and a bus scan are performed later when
568 * scsi_add_host() and scsi_scan_host() are called.
570 * Defined in: drivers/scsi/hosts.c .
572 struct Scsi_Host * scsi_host_alloc(struct scsi_host_template * sht,
577 * scsi_host_get - increment Scsi_Host instance refcount
578 * @shost: pointer to struct Scsi_Host instance
582 * Might block: currently may block but may be changed to not block
584 * Notes: Actually increments the counts in two sub-objects
586 * Defined in: drivers/scsi/hosts.c
588 void scsi_host_get(struct Scsi_Host *shost)
592 * scsi_host_put - decrement Scsi_Host instance refcount, free if 0
593 * @shost: pointer to struct Scsi_Host instance
597 * Might block: currently may block but may be changed to not block
599 * Notes: Actually decrements the counts in two sub-objects. If the
600 * latter refcount reaches 0, the Scsi_Host instance is freed.
601 * The LLD need not worry exactly when the Scsi_Host instance is
602 * freed, it just shouldn't access the instance after it has balanced
603 * out its refcount usage.
605 * Defined in: drivers/scsi/hosts.c
607 void scsi_host_put(struct Scsi_Host *shost)
611 * scsi_partsize - parse partition table into cylinders, heads + sectors
612 * @buf: pointer to partition table
613 * @capacity: size of (total) disk in 512 byte sectors
614 * @cyls: outputs number of cylinders calculated via this pointer
615 * @hds: outputs number of heads calculated via this pointer
616 * @secs: outputs number of sectors calculated via this pointer
618 * Returns 0 on success, -1 on failure
622 * Notes: Caller owns memory returned (free with kfree() )
624 * Defined in: drivers/scsi/scsicam.c
626 int scsi_partsize(unsigned char *buf, unsigned long capacity,
627 unsigned int *cyls, unsigned int *hds, unsigned int *secs)
631 * scsi_register - create and register a scsi host adapter instance.
632 * @sht: pointer to scsi host template
633 * @privsize: extra bytes to allocate in hostdata array (which is the
634 * last member of the returned Scsi_Host instance)
636 * Returns pointer to new Scsi_Host instance or NULL on failure
640 * Notes: When this call returns to the LLD, the SCSI bus scan on
641 * this host has _not_ yet been done.
642 * The hostdata array (by default zero length) is a per host scratch
645 * Defined in: drivers/scsi/hosts.c .
647 struct Scsi_Host * scsi_register(struct scsi_host_template * sht,
652 * scsi_remove_device - detach and remove a SCSI device
653 * @sdev: a pointer to a scsi device instance
655 * Returns value: 0 on success, -EINVAL if device not attached
659 * Notes: If an LLD becomes aware that a scsi device (lu) has
660 * been removed but its host is still present then it can request
661 * the removal of that scsi device. If successful this call will
662 * lead to the slave_destroy() callback being invoked. sdev is an
663 * invalid pointer after this call.
665 * Defined in: drivers/scsi/scsi_sysfs.c .
667 int scsi_remove_device(struct scsi_device *sdev)
671 * scsi_remove_host - detach and remove all SCSI devices owned by host
672 * @shost: a pointer to a scsi host instance
674 * Returns value: 0 on success, 1 on failure (e.g. LLD busy ??)
678 * Notes: Should only be invoked if the "hotplug initialization
679 * model" is being used. It should be called _prior_ to
682 * Defined in: drivers/scsi/hosts.c .
684 int scsi_remove_host(struct Scsi_Host *shost)
688 * scsi_report_bus_reset - report scsi _bus_ reset observed
689 * @shost: a pointer to a scsi host involved
690 * @channel: channel (within) host on which scsi bus reset occurred
696 * Notes: This only needs to be called if the reset is one which
697 * originates from an unknown location. Resets originated by the
698 * mid level itself don't need to call this, but there should be
699 * no harm. The main purpose of this is to make sure that a
700 * CHECK_CONDITION is properly treated.
702 * Defined in: drivers/scsi/scsi_error.c .
704 void scsi_report_bus_reset(struct Scsi_Host * shost, int channel)
708 * scsi_scan_host - scan SCSI bus
709 * @shost: a pointer to a scsi host instance
713 * Notes: Should be called after scsi_add_host()
715 * Defined in: drivers/scsi/scsi_scan.c
717 void scsi_scan_host(struct Scsi_Host *shost)
721 * scsi_track_queue_full - track successive QUEUE_FULL events on given
722 * device to determine if and when there is a need
723 * to adjust the queue depth on the device.
724 * @sdev: pointer to SCSI device instance
725 * @depth: Current number of outstanding SCSI commands on this device,
726 * not counting the one returned as QUEUE_FULL.
728 * Returns 0 - no change needed
729 * >0 - adjust queue depth to this new depth
730 * -1 - drop back to untagged operation using host->cmd_per_lun
731 * as the untagged command depth
735 * Notes: LLDs may call this at any time and we will do "The Right
736 * Thing"; interrupt context safe.
738 * Defined in: drivers/scsi/scsi.c .
740 int scsi_track_queue_full(struct scsi_device *sdev, int depth)
744 * scsi_unblock_requests - allow further commands to be queued to given host
746 * @shost: pointer to host to unblock commands on
752 * Defined in: drivers/scsi/scsi_lib.c .
754 void scsi_unblock_requests(struct Scsi_Host * shost)
758 * scsi_unregister - unregister and free memory used by host instance
759 * @shp: pointer to scsi host instance to unregister.
765 * Notes: Should not be invoked if the "hotplug initialization
766 * model" is being used. Called internally by exit_this_scsi_driver()
767 * in the "passive initialization model". Hence a LLD has no need to
768 * call this function directly.
770 * Defined in: drivers/scsi/hosts.c .
772 void scsi_unregister(struct Scsi_Host * shp)
779 Interface functions are supplied (defined) by LLDs and their function
780 pointers are placed in an instance of struct scsi_host_template which
781 is passed to scsi_host_alloc() [or scsi_register() / init_this_scsi_driver()].
782 Some are mandatory. Interface functions should be declared static. The
783 accepted convention is that driver "xyz" will declare its slave_configure()
785 static int xyz_slave_configure(struct scsi_device * sdev);
786 and so forth for all interface functions listed below.
788 A pointer to this function should be placed in the 'slave_configure' member
789 of a "struct scsi_host_template" instance. A pointer to such an instance
790 should be passed to the mid level's scsi_host_alloc() [or scsi_register() /
791 init_this_scsi_driver()].
793 The interface functions are also described in the include/scsi/scsi_host.h
794 file immediately above their definition point in "struct scsi_host_template".
795 In some cases more detail is given in scsi_host.h than below.
797 The interface functions are listed below in alphabetical order.
800 bios_param - fetch head, sector, cylinder info for a disk
801 detect - detects HBAs this driver wants to control
802 eh_timed_out - notify the host that a command timer expired
803 eh_abort_handler - abort given command
804 eh_bus_reset_handler - issue SCSI bus reset
805 eh_device_reset_handler - issue SCSI device reset
806 eh_host_reset_handler - reset host (host bus adapter)
807 eh_strategy_handler - driver supplied alternate to scsi_unjam_host()
808 info - supply information about given host
809 ioctl - driver can respond to ioctls
810 proc_info - supports /proc/scsi/{driver_name}/{host_no}
811 queuecommand - queue scsi command, invoke 'done' on completion
812 release - release all resources associated with given host
813 slave_alloc - prior to any commands being sent to a new device
814 slave_configure - driver fine tuning for given device after attach
815 slave_destroy - given device is about to be shut down
821 * bios_param - fetch head, sector, cylinder info for a disk
822 * @sdev: pointer to scsi device context (defined in
823 * include/scsi/scsi_device.h)
824 * @bdev: pointer to block device context (defined in fs.h)
825 * @capacity: device size (in 512 byte sectors)
826 * @params: three element array to place output:
827 * params[0] number of heads (max 255)
828 * params[1] number of sectors (max 63)
829 * params[2] number of cylinders
831 * Return value is ignored
835 * Calling context: process (sd)
837 * Notes: an arbitrary geometry (based on READ CAPACITY) is used
838 * if this function is not provided. The params array is
839 * pre-initialized with made up values just in case this function
840 * doesn't output anything.
842 * Optionally defined in: LLD
844 int bios_param(struct scsi_device * sdev, struct block_device *bdev,
845 sector_t capacity, int params[3])
849 * detect - detects HBAs this driver wants to control
850 * @shtp: host template for this driver.
852 * Returns number of hosts this driver wants to control. 0 means no
853 * suitable hosts found.
857 * Calling context: process [invoked from init_this_scsi_driver()]
859 * Notes: First function called from the SCSI mid level on this
860 * driver. Upper level drivers (e.g. sd) may not (yet) be present.
861 * For each host found, this method should call scsi_register()
864 * Defined in: LLD (required if "passive initialization mode" is used,
865 * not invoked in "hotplug initialization mode")
867 int detect(struct scsi_host_template * shtp)
871 * eh_timed_out - The timer for the command has just fired
872 * @scp: identifies command timing out
876 * EH_HANDLED: I fixed the error, please complete the command
877 * EH_RESET_TIMER: I need more time, reset the timer and
878 * begin counting again
879 * EH_NOT_HANDLED Begin normal error recovery
884 * Calling context: interrupt
886 * Notes: This is to give the LLD an opportunity to do local recovery.
887 * This recovery is limited to determining if the outstanding command
888 * will ever complete. You may not abort and restart the command from
891 * Optionally defined in: LLD
893 int eh_timed_out(struct scsi_cmnd * scp)
897 * eh_abort_handler - abort command associated with scp
898 * @scp: identifies command to be aborted
900 * Returns SUCCESS if command aborted else FAILED
904 * Calling context: kernel thread
906 * Notes: Invoked from scsi_eh thread. No other commands will be
907 * queued on current host during eh.
909 * Optionally defined in: LLD
911 int eh_abort_handler(struct scsi_cmnd * scp)
915 * eh_bus_reset_handler - issue SCSI bus reset
916 * @scp: SCSI bus that contains this device should be reset
918 * Returns SUCCESS if command aborted else FAILED
922 * Calling context: kernel thread
924 * Notes: Invoked from scsi_eh thread. No other commands will be
925 * queued on current host during eh.
927 * Optionally defined in: LLD
929 int eh_bus_reset_handler(struct scsi_cmnd * scp)
933 * eh_device_reset_handler - issue SCSI device reset
934 * @scp: identifies SCSI device to be reset
936 * Returns SUCCESS if command aborted else FAILED
940 * Calling context: kernel thread
942 * Notes: Invoked from scsi_eh thread. No other commands will be
943 * queued on current host during eh.
945 * Optionally defined in: LLD
947 int eh_device_reset_handler(struct scsi_cmnd * scp)
951 * eh_host_reset_handler - reset host (host bus adapter)
952 * @scp: SCSI host that contains this device should be reset
954 * Returns SUCCESS if command aborted else FAILED
958 * Calling context: kernel thread
960 * Notes: Invoked from scsi_eh thread. No other commands will be
961 * queued on current host during eh.
962 * With the default eh_strategy in place, if none of the _abort_,
963 * _device_reset_, _bus_reset_ or this eh handler function are
964 * defined (or they all return FAILED) then the device in question
965 * will be set offline whenever eh is invoked.
967 * Optionally defined in: LLD
969 int eh_host_reset_handler(struct scsi_cmnd * scp)
973 * eh_strategy_handler - driver supplied alternate to scsi_unjam_host()
974 * @shp: host on which error has occurred
976 * Returns TRUE if host unjammed, else FALSE.
980 * Calling context: kernel thread
982 * Notes: Invoked from scsi_eh thread. LLD supplied alternate to
983 * scsi_unjam_host() found in scsi_error.c
985 * Optionally defined in: LLD
987 int eh_strategy_handler(struct Scsi_Host * shp)
991 * info - supply information about given host: driver name plus data
992 * to distinguish given host
993 * @shp: host to supply information about
995 * Return ASCII null terminated string. [This driver is assumed to
996 * manage the memory pointed to and maintain it, typically for the
997 * lifetime of this host.]
1001 * Calling context: process
1003 * Notes: Often supplies PCI or ISA information such as IO addresses
1004 * and interrupt numbers. If not supplied struct Scsi_Host::name used
1005 * instead. It is assumed the returned information fits on one line
1006 * (i.e. does not included embedded newlines).
1007 * The SCSI_IOCTL_PROBE_HOST ioctl yields the string returned by this
1008 * function (or struct Scsi_Host::name if this function is not
1010 * In a similar manner, init_this_scsi_driver() outputs to the console
1011 * each host's "info" (or name) for the driver it is registering.
1012 * Also if proc_info() is not supplied, the output of this function
1015 * Optionally defined in: LLD
1017 const char * info(struct Scsi_Host * shp)
1021 * ioctl - driver can respond to ioctls
1022 * @sdp: device that ioctl was issued for
1023 * @cmd: ioctl number
1024 * @arg: pointer to read or write data from. Since it points to
1025 * user space, should use appropriate kernel functions
1026 * (e.g. copy_from_user() ). In the Unix style this argument
1027 * can also be viewed as an unsigned long.
1029 * Returns negative "errno" value when there is a problem. 0 or a
1030 * positive value indicates success and is returned to the user space.
1034 * Calling context: process
1036 * Notes: The SCSI subsystem uses a "trickle down" ioctl model.
1037 * The user issues an ioctl() against an upper level driver
1038 * (e.g. /dev/sdc) and if the upper level driver doesn't recognize
1039 * the 'cmd' then it is passed to the SCSI mid level. If the SCSI
1040 * mid level does not recognize it, then the LLD that controls
1041 * the device receives the ioctl. According to recent Unix standards
1042 * unsupported ioctl() 'cmd' numbers should return -ENOTTY.
1044 * Optionally defined in: LLD
1046 int ioctl(struct scsi_device *sdp, int cmd, void *arg)
1050 * proc_info - supports /proc/scsi/{driver_name}/{host_no}
1051 * @buffer: anchor point to output to (0==writeto1_read0) or fetch from
1052 * (1==writeto1_read0).
1053 * @start: where "interesting" data is written to. Ignored when
1054 * 1==writeto1_read0.
1055 * @offset: offset within buffer 0==writeto1_read0 is actually
1056 * interested in. Ignored when 1==writeto1_read0 .
1057 * @length: maximum (or actual) extent of buffer
1058 * @host_no: host number of interest (struct Scsi_Host::host_no)
1059 * @writeto1_read0: 1 -> data coming from user space towards driver
1060 * (e.g. "echo some_string > /proc/scsi/xyz/2")
1061 * 0 -> user what data from this driver
1062 * (e.g. "cat /proc/scsi/xyz/2")
1064 * Returns length when 1==writeto1_read0. Otherwise number of chars
1065 * output to buffer past offset.
1069 * Calling context: process
1071 * Notes: Driven from scsi_proc.c which interfaces to proc_fs. proc_fs
1072 * support can now be configured out of the scsi subsystem.
1074 * Optionally defined in: LLD
1076 int proc_info(char * buffer, char ** start, off_t offset,
1077 int length, int host_no, int writeto1_read0)
1081 * queuecommand - queue scsi command, invoke 'done' on completion
1082 * @scp: pointer to scsi command object
1083 * @done: function pointer to be invoked on completion
1085 * Returns 0 on success.
1087 * If there's a failure, return either:
1089 * SCSI_MLQUEUE_DEVICE_BUSY if the device queue is full, or
1090 * SCSI_MLQUEUE_HOST_BUSY if the entire host queue is full
1092 * On both of these returns, the mid-layer will requeue the I/O
1094 * - if the return is SCSI_MLQUEUE_DEVICE_BUSY, only that particular
1095 * device will be paused, and it will be unpaused when a command to
1096 * the device returns (or after a brief delay if there are no more
1097 * outstanding commands to it). Commands to other devices continue
1098 * to be processed normally.
1100 * - if the return is SCSI_MLQUEUE_HOST_BUSY, all I/O to the host
1101 * is paused and will be unpaused when any command returns from
1102 * the host (or after a brief delay if there are no outstanding
1103 * commands to the host).
1105 * For compatibility with earlier versions of queuecommand, any
1106 * other return value is treated the same as
1107 * SCSI_MLQUEUE_HOST_BUSY.
1109 * Other types of errors that are detected immediately may be
1110 * flagged by setting scp->result to an appropriate value,
1111 * invoking the 'done' callback, and then returning 0 from this
1112 * function. If the command is not performed immediately (and the
1113 * LLD is starting (or will start) the given command) then this
1114 * function should place 0 in scp->result and return 0.
1116 * Command ownership. If the driver returns zero, it owns the
1117 * command and must take responsibility for ensuring the 'done'
1118 * callback is executed. Note: the driver may call done before
1119 * returning zero, but after it has called done, it may not
1120 * return any value other than zero. If the driver makes a
1121 * non-zero return, it must not execute the command's done
1122 * callback at any time.
1124 * Locks: struct Scsi_Host::host_lock held on entry (with "irqsave")
1125 * and is expected to be held on return.
1127 * Calling context: in interrupt (soft irq) or process context
1129 * Notes: This function should be relatively fast. Normally it will
1130 * not wait for IO to complete. Hence the 'done' callback is invoked
1131 * (often directly from an interrupt service routine) some time after
1132 * this function has returned. In some cases (e.g. pseudo adapter
1133 * drivers that manufacture the response to a SCSI INQUIRY)
1134 * the 'done' callback may be invoked before this function returns.
1135 * If the 'done' callback is not invoked within a certain period
1136 * the SCSI mid level will commence error processing.
1137 * If a status of CHECK CONDITION is placed in "result" when the
1138 * 'done' callback is invoked, then the LLD driver should
1139 * perform autosense and fill in the struct scsi_cmnd::sense_buffer
1140 * array. The scsi_cmnd::sense_buffer array is zeroed prior to
1141 * the mid level queuing a command to an LLD.
1145 int queuecommand(struct scsi_cmnd * scp,
1146 void (*done)(struct scsi_cmnd *))
1150 * release - release all resources associated with given host
1151 * @shp: host to be released.
1153 * Return value ignored (could soon be a function returning void).
1157 * Calling context: process
1159 * Notes: Invoked from scsi_module.c's exit_this_scsi_driver().
1160 * LLD's implementation of this function should call
1161 * scsi_unregister(shp) prior to returning.
1162 * Only needed for old-style host templates.
1164 * Defined in: LLD (required in "passive initialization model",
1165 * should not be defined in hotplug model)
1167 int release(struct Scsi_Host * shp)
1171 * slave_alloc - prior to any commands being sent to a new device
1172 * (i.e. just prior to scan) this call is made
1173 * @sdp: pointer to new device (about to be scanned)
1175 * Returns 0 if ok. Any other return is assumed to be an error and
1176 * the device is ignored.
1180 * Calling context: process
1182 * Notes: Allows the driver to allocate any resources for a device
1183 * prior to its initial scan. The corresponding scsi device may not
1184 * exist but the mid level is just about to scan for it (i.e. send
1185 * and INQUIRY command plus ...). If a device is found then
1186 * slave_configure() will be called while if a device is not found
1187 * slave_destroy() is called.
1188 * For more details see the include/scsi/scsi_host.h file.
1190 * Optionally defined in: LLD
1192 int slave_alloc(struct scsi_device *sdp)
1196 * slave_configure - driver fine tuning for given device just after it
1197 * has been first scanned (i.e. it responded to an
1199 * @sdp: device that has just been attached
1201 * Returns 0 if ok. Any other return is assumed to be an error and
1202 * the device is taken offline. [offline devices will _not_ have
1203 * slave_destroy() called on them so clean up resources.]
1207 * Calling context: process
1209 * Notes: Allows the driver to inspect the response to the initial
1210 * INQUIRY done by the scanning code and take appropriate action.
1211 * For more details see the include/scsi/scsi_host.h file.
1213 * Optionally defined in: LLD
1215 int slave_configure(struct scsi_device *sdp)
1219 * slave_destroy - given device is about to be shut down. All
1220 * activity has ceased on this device.
1221 * @sdp: device that is about to be shut down
1227 * Calling context: process
1229 * Notes: Mid level structures for given device are still in place
1230 * but are about to be torn down. Any per device resources allocated
1231 * by this driver for given device should be freed now. No further
1232 * commands will be sent for this sdp instance. [However the device
1233 * could be re-attached in the future in which case a new instance
1234 * of struct scsi_device would be supplied by future slave_alloc()
1235 * and slave_configure() calls.]
1237 * Optionally defined in: LLD
1239 void slave_destroy(struct scsi_device *sdp)
1245 struct scsi_host_template
1246 -------------------------
1247 There is one "struct scsi_host_template" instance per LLD ***. It is
1248 typically initialized as a file scope static in a driver's header file. That
1249 way members that are not explicitly initialized will be set to 0 or NULL.
1251 name - name of driver (may contain spaces, please limit to
1252 less than 80 characters)
1253 proc_name - name used in "/proc/scsi/<proc_name>/<host_no>" and
1254 by sysfs in one of its "drivers" directories. Hence
1255 "proc_name" should only contain characters acceptable
1256 to a Unix file name.
1257 (*queuecommand)() - primary callback that the mid level uses to inject
1258 SCSI commands into an LLD.
1259 The structure is defined and commented in include/scsi/scsi_host.h
1261 *** In extreme situations a single driver may have several instances
1262 if it controls several different classes of hardware (e.g. an LLD
1263 that handles both ISA and PCI cards and has a separate instance of
1264 struct scsi_host_template for each class).
1268 There is one struct Scsi_Host instance per host (HBA) that an LLD
1269 controls. The struct Scsi_Host structure has many members in common
1270 with "struct scsi_host_template". When a new struct Scsi_Host instance
1271 is created (in scsi_host_alloc() in hosts.c) those common members are
1272 initialized from the driver's struct scsi_host_template instance. Members
1274 host_no - system wide unique number that is used for identifying
1275 this host. Issued in ascending order from 0.
1276 can_queue - must be greater than 0; do not send more than can_queue
1277 commands to the adapter.
1278 this_id - scsi id of host (scsi initiator) or -1 if not known
1279 sg_tablesize - maximum scatter gather elements allowed by host.
1280 0 implies scatter gather not supported by host
1281 max_sectors - maximum number of sectors (usually 512 bytes) allowed
1282 in a single SCSI command. The default value of 0 leads
1283 to a setting of SCSI_DEFAULT_MAX_SECTORS (defined in
1284 scsi_host.h) which is currently set to 1024. So for a
1285 disk the maximum transfer size is 512 KB when max_sectors
1286 is not defined. Note that this size may not be sufficient
1287 for disk firmware uploads.
1288 cmd_per_lun - maximum number of commands that can be queued on devices
1289 controlled by the host. Overridden by LLD calls to
1290 scsi_adjust_queue_depth().
1291 unchecked_isa_dma - 1=>only use bottom 16 MB of ram (ISA DMA addressing
1292 restriction), 0=>can use full 32 bit (or better) DMA
1294 use_clustering - 1=>SCSI commands in mid level's queue can be merged,
1295 0=>disallow SCSI command merging
1296 hostt - pointer to driver's struct scsi_host_template from which
1297 this struct Scsi_Host instance was spawned
1298 hostt->proc_name - name of LLD. This is the driver name that sysfs uses
1299 transportt - pointer to driver's struct scsi_transport_template instance
1300 (if any). FC and SPI transports currently supported.
1301 sh_list - a double linked list of pointers to all struct Scsi_Host
1302 instances (currently ordered by ascending host_no)
1303 my_devices - a double linked list of pointers to struct scsi_device
1304 instances that belong to this host.
1305 hostdata[0] - area reserved for LLD at end of struct Scsi_Host. Size
1306 is set by the second argument (named 'xtr_bytes') to
1307 scsi_host_alloc() or scsi_register().
1309 The scsi_host structure is defined in include/scsi/scsi_host.h
1313 Generally, there is one instance of this structure for each SCSI logical unit
1314 on a host. Scsi devices connected to a host are uniquely identified by a
1315 channel number, target id and logical unit number (lun).
1316 The structure is defined in include/scsi/scsi_device.h
1320 Instances of this structure convey SCSI commands to the LLD and responses
1321 back to the mid level. The SCSI mid level will ensure that no more SCSI
1322 commands become queued against the LLD than are indicated by
1323 scsi_adjust_queue_depth() (or struct Scsi_Host::cmd_per_lun). There will
1324 be at least one instance of struct scsi_cmnd available for each SCSI device.
1325 Members of interest:
1326 cmnd - array containing SCSI command
1327 cmnd_len - length (in bytes) of SCSI command
1328 sc_data_direction - direction of data transfer in data phase. See
1329 "enum dma_data_direction" in include/linux/dma-mapping.h
1330 request_bufflen - number of data bytes to transfer (0 if no data phase)
1331 use_sg - ==0 -> no scatter gather list, hence transfer data
1332 to/from request_buffer
1333 - >0 -> scatter gather list (actually an array) in
1334 request_buffer with use_sg elements
1335 request_buffer - either contains data buffer or scatter gather list
1336 depending on the setting of use_sg. Scatter gather
1337 elements are defined by 'struct scatterlist' found
1338 in include/asm/scatterlist.h .
1339 done - function pointer that should be invoked by LLD when the
1340 SCSI command is completed (successfully or otherwise).
1341 Should only be called by an LLD if the LLD has accepted
1342 the command (i.e. queuecommand() returned or will return
1343 0). The LLD may invoke 'done' prior to queuecommand()
1345 result - should be set by LLD prior to calling 'done'. A value
1346 of 0 implies a successfully completed command (and all
1347 data (if any) has been transferred to or from the SCSI
1348 target device). 'result' is a 32 bit unsigned integer that
1349 can be viewed as 4 related bytes. The SCSI status value is
1350 in the LSB. See include/scsi/scsi.h status_byte(),
1351 msg_byte(), host_byte() and driver_byte() macros and
1353 sense_buffer - an array (maximum size: SCSI_SENSE_BUFFERSIZE bytes) that
1354 should be written when the SCSI status (LSB of 'result')
1355 is set to CHECK_CONDITION (2). When CHECK_CONDITION is
1356 set, if the top nibble of sense_buffer[0] has the value 7
1357 then the mid level will assume the sense_buffer array
1358 contains a valid SCSI sense buffer; otherwise the mid
1359 level will issue a REQUEST_SENSE SCSI command to
1360 retrieve the sense buffer. The latter strategy is error
1361 prone in the presence of command queuing so the LLD should
1362 always "auto-sense".
1363 device - pointer to scsi_device object that this command is
1365 resid - an LLD should set this signed integer to the requested
1366 transfer length (i.e. 'request_bufflen') less the number
1367 of bytes that are actually transferred. 'resid' is
1368 preset to 0 so an LLD can ignore it if it cannot detect
1369 underruns (overruns should be rare). If possible an LLD
1370 should set 'resid' prior to invoking 'done'. The most
1371 interesting case is data transfers from a SCSI target
1372 device device (i.e. READs) that underrun.
1373 underflow - LLD should place (DID_ERROR << 16) in 'result' if
1374 actual number of bytes transferred is less than this
1375 figure. Not many LLDs implement this check and some that
1376 do just output an error message to the log rather than
1377 report a DID_ERROR. Better for an LLD to implement
1380 The scsi_cmnd structure is defined in include/scsi/scsi_cmnd.h
1385 Each struct Scsi_Host instance has a spin_lock called struct
1386 Scsi_Host::default_lock which is initialized in scsi_host_alloc() [found in
1387 hosts.c]. Within the same function the struct Scsi_Host::host_lock pointer
1388 is initialized to point at default_lock with the scsi_assign_lock() function.
1389 Thereafter lock and unlock operations performed by the mid level use the
1390 struct Scsi_Host::host_lock pointer.
1392 LLDs can override the use of struct Scsi_Host::default_lock by
1393 using scsi_assign_lock(). The earliest opportunity to do this would
1394 be in the detect() function after it has invoked scsi_register(). It
1395 could be replaced by a coarser grain lock (e.g. per driver) or a
1396 lock of equal granularity (i.e. per host). Using finer grain locks
1397 (e.g. per SCSI device) may be possible by juggling locks in
1402 Autosense (or auto-sense) is defined in the SAM-2 document as "the
1403 automatic return of sense data to the application client coincident
1404 with the completion of a SCSI command" when a status of CHECK CONDITION
1405 occurs. LLDs should perform autosense. This should be done when the LLD
1406 detects a CHECK CONDITION status by either:
1407 a) instructing the SCSI protocol (e.g. SCSI Parallel Interface (SPI))
1408 to perform an extra data in phase on such responses
1409 b) or, the LLD issuing a REQUEST SENSE command itself
1411 Either way, when a status of CHECK CONDITION is detected, the mid level
1412 decides whether the LLD has performed autosense by checking struct
1413 scsi_cmnd::sense_buffer[0] . If this byte has an upper nibble of 7 (or 0xf)
1414 then autosense is assumed to have taken place. If it has another value (and
1415 this byte is initialized to 0 before each command) then the mid level will
1416 issue a REQUEST SENSE command.
1418 In the presence of queued commands the "nexus" that maintains sense
1419 buffer data from the command that failed until a following REQUEST SENSE
1420 may get out of synchronization. This is why it is best for the LLD
1421 to perform autosense.
1424 Changes since lk 2.4 series
1425 ===========================
1426 io_request_lock has been replaced by several finer grained locks. The lock
1427 relevant to LLDs is struct Scsi_Host::host_lock and there is
1430 The older error handling mechanism has been removed. This means the
1431 LLD interface functions abort() and reset() have been removed.
1432 The struct scsi_host_template::use_new_eh_code flag has been removed.
1434 In the 2.4 series the SCSI subsystem configuration descriptions were
1435 aggregated with the configuration descriptions from all other Linux
1436 subsystems in the Documentation/Configure.help file. In the 2.6 series,
1437 the SCSI subsystem now has its own (much smaller) drivers/scsi/Kconfig
1438 file that contains both configuration and help information.
1440 struct SHT has been renamed to struct scsi_host_template.
1442 Addition of the "hotplug initialization model" and many extra functions
1448 The following people have contributed to this document:
1449 Mike Anderson <andmike at us dot ibm dot com>
1450 James Bottomley <James dot Bottomley at steeleye dot com>
1451 Patrick Mansfield <patmans at us dot ibm dot com>
1452 Christoph Hellwig <hch at infradead dot org>
1453 Doug Ledford <dledford at redhat dot com>
1454 Andries Brouwer <Andries dot Brouwer at cwi dot nl>
1455 Randy Dunlap <rdunlap at xenotime dot net>
1456 Alan Stern <stern at rowland dot harvard dot edu>
1460 dgilbert at interlog dot com
projects / linux-2.6 / blob