1 =======================
2 A Linux CD-ROM standard
3 =======================
5 :Author: David van Leeuwen <david@ElseWare.cistron.nl>
7 :Updated by: Erik Andersen (andersee@debian.org)
8 :Updated by: Jens Axboe (axboe@image.dk)
14 Linux is probably the Unix-like operating system that supports
15 the widest variety of hardware devices. The reasons for this are
18 - The large list of hardware devices available for the many platforms
19 that Linux now supports (i.e., i386-PCs, Sparc Suns, etc.)
20 - The open design of the operating system, such that anybody can write a
22 - There is plenty of source code around as examples of how to write a driver.
24 The openness of Linux, and the many different types of available
25 hardware has allowed Linux to support many different hardware devices.
26 Unfortunately, the very openness that has allowed Linux to support
27 all these different devices has also allowed the behavior of each
28 device driver to differ significantly from one device to another.
29 This divergence of behavior has been very significant for CD-ROM
30 devices; the way a particular drive reacts to a `standard` *ioctl()*
31 call varies greatly from one device driver to another. To avoid making
32 their drivers totally inconsistent, the writers of Linux CD-ROM
33 drivers generally created new device drivers by understanding, copying,
34 and then changing an existing one. Unfortunately, this practice did not
35 maintain uniform behavior across all the Linux CD-ROM drivers.
37 This document describes an effort to establish Uniform behavior across
38 all the different CD-ROM device drivers for Linux. This document also
39 defines the various *ioctl()'s*, and how the low-level CD-ROM device
40 drivers should implement them. Currently (as of the Linux 2.1.\ *x*
41 development kernels) several low-level CD-ROM device drivers, including
42 both IDE/ATAPI and SCSI, now use this Uniform interface.
44 When the CD-ROM was developed, the interface between the CD-ROM drive
45 and the computer was not specified in the standards. As a result, many
46 different CD-ROM interfaces were developed. Some of them had their
47 own proprietary design (Sony, Mitsumi, Panasonic, Philips), other
48 manufacturers adopted an existing electrical interface and changed
49 the functionality (CreativeLabs/SoundBlaster, Teac, Funai) or simply
50 adapted their drives to one or more of the already existing electrical
51 interfaces (Aztech, Sanyo, Funai, Vertos, Longshine, Optics Storage and
52 most of the `NoName` manufacturers). In cases where a new drive really
53 brought its own interface or used its own command set and flow control
54 scheme, either a separate driver had to be written, or an existing
55 driver had to be enhanced. History has delivered us CD-ROM support for
56 many of these different interfaces. Nowadays, almost all new CD-ROM
57 drives are either IDE/ATAPI or SCSI, and it is very unlikely that any
58 manufacturer will create a new interface. Even finding drives for the
59 old proprietary interfaces is getting difficult.
61 When (in the 1.3.70's) I looked at the existing software interface,
62 which was expressed through `cdrom.h`, it appeared to be a rather wild
63 set of commands and data formats [#f1]_. It seemed that many
64 features of the software interface had been added to accommodate the
65 capabilities of a particular drive, in an *ad hoc* manner. More
66 importantly, it appeared that the behavior of the `standard` commands
67 was different for most of the different drivers: e. g., some drivers
68 close the tray if an *open()* call occurs when the tray is open, while
69 others do not. Some drivers lock the door upon opening the device, to
70 prevent an incoherent file system, but others don't, to allow software
71 ejection. Undoubtedly, the capabilities of the different drives vary,
72 but even when two drives have the same capability their drivers'
73 behavior was usually different.
76 I cannot recollect what kernel version I looked at, then,
77 presumably 1.2.13 and 1.3.34 --- the latest kernel that I was
78 indirectly involved in.
80 I decided to start a discussion on how to make all the Linux CD-ROM
81 drivers behave more uniformly. I began by contacting the developers of
82 the many CD-ROM drivers found in the Linux kernel. Their reactions
83 encouraged me to write the Uniform CD-ROM Driver which this document is
84 intended to describe. The implementation of the Uniform CD-ROM Driver is
85 in the file `cdrom.c`. This driver is intended to be an additional software
86 layer that sits on top of the low-level device drivers for each CD-ROM drive.
87 By adding this additional layer, it is possible to have all the different
88 CD-ROM devices behave **exactly** the same (insofar as the underlying
91 The goal of the Uniform CD-ROM Driver is **not** to alienate driver developers
92 whohave not yet taken steps to support this effort. The goal of Uniform CD-ROM
93 Driver is simply to give people writing application programs for CD-ROM drives
94 **one** Linux CD-ROM interface with consistent behavior for all
95 CD-ROM devices. In addition, this also provides a consistent interface
96 between the low-level device driver code and the Linux kernel. Care
97 is taken that 100% compatibility exists with the data structures and
98 programmer's interface defined in `cdrom.h`. This guide was written to
99 help CD-ROM driver developers adapt their code to use the Uniform CD-ROM
100 Driver code defined in `cdrom.c`.
102 Personally, I think that the most important hardware interfaces are
103 the IDE/ATAPI drives and, of course, the SCSI drives, but as prices
104 of hardware drop continuously, it is also likely that people may have
105 more than one CD-ROM drive, possibly of mixed types. It is important
106 that these drives behave in the same way. In December 1994, one of the
107 cheapest CD-ROM drives was a Philips cm206, a double-speed proprietary
108 drive. In the months that I was busy writing a Linux driver for it,
109 proprietary drives became obsolete and IDE/ATAPI drives became the
110 standard. At the time of the last update to this document (November
111 1997) it is becoming difficult to even **find** anything less than a
112 16 speed CD-ROM drive, and 24 speed drives are common.
116 Standardizing through another software level
117 ============================================
119 At the time this document was conceived, all drivers directly
120 implemented the CD-ROM *ioctl()* calls through their own routines. This
121 led to the danger of different drivers forgetting to do important things
122 like checking that the user was giving the driver valid data. More
123 importantly, this led to the divergence of behavior, which has already
126 For this reason, the Uniform CD-ROM Driver was created to enforce consistent
127 CD-ROM drive behavior, and to provide a common set of services to the various
128 low-level CD-ROM device drivers. The Uniform CD-ROM Driver now provides another
129 software-level, that separates the *ioctl()* and *open()* implementation
130 from the actual hardware implementation. Note that this effort has
131 made few changes which will affect a user's application programs. The
132 greatest change involved moving the contents of the various low-level
133 CD-ROM drivers\' header files to the kernel's cdrom directory. This was
134 done to help ensure that the user is only presented with only one cdrom
135 interface, the interface defined in `cdrom.h`.
137 CD-ROM drives are specific enough (i. e., different from other
138 block-devices such as floppy or hard disc drives), to define a set
139 of common **CD-ROM device operations**, *<cdrom-device>_dops*.
140 These operations are different from the classical block-device file
141 operations, *<block-device>_fops*.
143 The routines for the Uniform CD-ROM Driver interface level are implemented
144 in the file `cdrom.c`. In this file, the Uniform CD-ROM Driver interfaces
145 with the kernel as a block device by registering the following general
146 *struct file_operations*::
148 struct file_operations cdrom_fops = {
150 block _read , /∗ read—general block-dev read ∗/
151 block _write, /∗ write—general block-dev write ∗/
154 cdrom_ioctl, /∗ ioctl ∗/
156 cdrom_open, /∗ open ∗/
157 cdrom_release, /∗ release ∗/
160 cdrom_media_changed, /∗ media change ∗/
161 NULL /∗ revalidate ∗/
164 Every active CD-ROM device shares this *struct*. The routines
165 declared above are all implemented in `cdrom.c`, since this file is the
166 place where the behavior of all CD-ROM-devices is defined and
167 standardized. The actual interface to the various types of CD-ROM
168 hardware is still performed by various low-level CD-ROM-device
169 drivers. These routines simply implement certain **capabilities**
170 that are common to all CD-ROM (and really, all removable-media
173 Registration of a low-level CD-ROM device driver is now done through
174 the general routines in `cdrom.c`, not through the Virtual File System
175 (VFS) any more. The interface implemented in `cdrom.c` is carried out
176 through two general structures that contain information about the
177 capabilities of the driver, and the specific drives on which the
178 driver operates. The structures are:
181 This structure contains information about the low-level driver for a
182 CD-ROM device. This structure is conceptually connected to the major
183 number of the device (although some drivers may have different
184 major numbers, as is the case for the IDE driver).
187 This structure contains information about a particular CD-ROM drive,
188 such as its device name, speed, etc. This structure is conceptually
189 connected to the minor number of the device.
191 Registering a particular CD-ROM drive with the Uniform CD-ROM Driver
192 is done by the low-level device driver though a call to::
194 register_cdrom(struct cdrom_device_info * <device>_info)
196 The device information structure, *<device>_info*, contains all the
197 information needed for the kernel to interface with the low-level
198 CD-ROM device driver. One of the most important entries in this
199 structure is a pointer to the *cdrom_device_ops* structure of the
202 The device operations structure, *cdrom_device_ops*, contains a list
203 of pointers to the functions which are implemented in the low-level
204 device driver. When `cdrom.c` accesses a CD-ROM device, it does it
205 through the functions in this structure. It is impossible to know all
206 the capabilities of future CD-ROM drives, so it is expected that this
207 list may need to be expanded from time to time as new technologies are
208 developed. For example, CD-R and CD-R/W drives are beginning to become
209 popular, and support will soon need to be added for them. For now, the
210 current *struct* is::
212 struct cdrom_device_ops {
213 int (*open)(struct cdrom_device_info *, int)
214 void (*release)(struct cdrom_device_info *);
215 int (*drive_status)(struct cdrom_device_info *, int);
216 unsigned int (*check_events)(struct cdrom_device_info *,
218 int (*media_changed)(struct cdrom_device_info *, int);
219 int (*tray_move)(struct cdrom_device_info *, int);
220 int (*lock_door)(struct cdrom_device_info *, int);
221 int (*select_speed)(struct cdrom_device_info *, int);
222 int (*select_disc)(struct cdrom_device_info *, int);
223 int (*get_last_session) (struct cdrom_device_info *,
224 struct cdrom_multisession *);
225 int (*get_mcn)(struct cdrom_device_info *, struct cdrom_mcn *);
226 int (*reset)(struct cdrom_device_info *);
227 int (*audio_ioctl)(struct cdrom_device_info *,
228 unsigned int, void *);
229 const int capability; /* capability flags */
230 int (*generic_packet)(struct cdrom_device_info *,
231 struct packet_command *);
234 When a low-level device driver implements one of these capabilities,
235 it should add a function pointer to this *struct*. When a particular
236 function is not implemented, however, this *struct* should contain a
237 NULL instead. The *capability* flags specify the capabilities of the
238 CD-ROM hardware and/or low-level CD-ROM driver when a CD-ROM drive
239 is registered with the Uniform CD-ROM Driver.
241 Note that most functions have fewer parameters than their
242 *blkdev_fops* counterparts. This is because very little of the
243 information in the structures *inode* and *file* is used. For most
244 drivers, the main parameter is the *struct* *cdrom_device_info*, from
245 which the major and minor number can be extracted. (Most low-level
246 CD-ROM drivers don't even look at the major and minor number though,
247 since many of them only support one device.) This will be available
248 through *dev* in *cdrom_device_info* described below.
250 The drive-specific, minor-like information that is registered with
251 `cdrom.c`, currently contains the following fields::
253 struct cdrom_device_info {
254 const struct cdrom_device_ops * ops; /* device operations for this major */
255 struct list_head list; /* linked list of all device_info */
256 struct gendisk * disk; /* matching block layer disk */
257 void * handle; /* driver-dependent data */
259 int mask; /* mask of capability: disables them */
260 int speed; /* maximum speed for reading data */
261 int capacity; /* number of discs in a jukebox */
263 unsigned int options:30; /* options flags */
264 unsigned mc_flags:2; /* media-change buffer flags */
265 unsigned int vfs_events; /* cached events for vfs path */
266 unsigned int ioctl_events; /* cached events for ioctl path */
267 int use_count; /* number of times device is opened */
268 char name[20]; /* name of the device type */
270 __u8 sanyo_slot : 2; /* Sanyo 3-CD changer support */
271 __u8 keeplocked : 1; /* CDROM_LOCKDOOR status */
272 __u8 reserved : 5; /* not used yet */
273 int cdda_method; /* see CDDA_* flags */
274 __u8 last_sense; /* saves last sense key */
275 __u8 media_written; /* dirty flag, DVD+RW bookkeeping */
276 unsigned short mmc3_profile; /* current MMC3 profile */
277 int for_data; /* unknown:TBD */
278 int (*exit)(struct cdrom_device_info *);/* unknown:TBD */
279 int mrw_mode_page; /* which MRW mode page is in use */
282 Using this *struct*, a linked list of the registered minor devices is
283 built, using the *next* field. The device number, the device operations
284 struct and specifications of properties of the drive are stored in this
287 The *mask* flags can be used to mask out some of the capabilities listed
288 in *ops->capability*, if a specific drive doesn't support a feature
289 of the driver. The value *speed* specifies the maximum head-rate of the
290 drive, measured in units of normal audio speed (176kB/sec raw data or
291 150kB/sec file system data). The parameters are declared *const*
292 because they describe properties of the drive, which don't change after
295 A few registers contain variables local to the CD-ROM drive. The
296 flags *options* are used to specify how the general CD-ROM routines
297 should behave. These various flags registers should provide enough
298 flexibility to adapt to the different users' wishes (and **not** the
299 `arbitrary` wishes of the author of the low-level device driver, as is
300 the case in the old scheme). The register *mc_flags* is used to buffer
301 the information from *media_changed()* to two separate queues. Other
302 data that is specific to a minor drive, can be accessed through *handle*,
303 which can point to a data structure specific to the low-level driver.
304 The fields *use_count*, *next*, *options* and *mc_flags* need not be
307 The intermediate software layer that `cdrom.c` forms will perform some
308 additional bookkeeping. The use count of the device (the number of
309 processes that have the device opened) is registered in *use_count*. The
310 function *cdrom_ioctl()* will verify the appropriate user-memory regions
311 for read and write, and in case a location on the CD is transferred,
312 it will `sanitize` the format by making requests to the low-level
313 drivers in a standard format, and translating all formats between the
314 user-software and low level drivers. This relieves much of the drivers'
315 memory checking and format checking and translation. Also, the necessary
316 structures will be declared on the program stack.
318 The implementation of the functions should be as defined in the
319 following sections. Two functions **must** be implemented, namely
320 *open()* and *release()*. Other functions may be omitted, their
321 corresponding capability flags will be cleared upon registration.
322 Generally, a function returns zero on success and negative on error. A
323 function call should return only after the command has completed, but of
324 course waiting for the device should not use processor time.
328 int open(struct cdrom_device_info *cdi, int purpose)
330 *Open()* should try to open the device for a specific *purpose*, which
333 - Open for reading data, as done by `mount()` (2), or the
334 user commands `dd` or `cat`.
335 - Open for *ioctl* commands, as done by audio-CD playing programs.
337 Notice that any strategic code (closing tray upon *open()*, etc.) is
338 done by the calling routine in `cdrom.c`, so the low-level routine
339 should only be concerned with proper initialization, such as spinning
344 void release(struct cdrom_device_info *cdi)
346 Device-specific actions should be taken such as spinning down the device.
347 However, strategic actions such as ejection of the tray, or unlocking
348 the door, should be left over to the general routine *cdrom_release()*.
349 This is the only function returning type *void*.
351 .. _cdrom_drive_status:
355 int drive_status(struct cdrom_device_info *cdi, int slot_nr)
357 The function *drive_status*, if implemented, should provide
358 information on the status of the drive (not the status of the disc,
359 which may or may not be in the drive). If the drive is not a changer,
360 *slot_nr* should be ignored. In `cdrom.h` the possibilities are listed::
363 CDS_NO_INFO /* no information available */
364 CDS_NO_DISC /* no disc is inserted, tray is closed */
365 CDS_TRAY_OPEN /* tray is opened */
366 CDS_DRIVE_NOT_READY /* something is wrong, tray is moving? */
367 CDS_DISC_OK /* a disc is loaded and everything is fine */
371 int media_changed(struct cdrom_device_info *cdi, int disc_nr)
373 This function is very similar to the original function in $struct
374 file_operations*. It returns 1 if the medium of the device *cdi->dev*
375 has changed since the last call, and 0 otherwise. The parameter
376 *disc_nr* identifies a specific slot in a juke-box, it should be
377 ignored for single-disc drives. Note that by `re-routing` this
378 function through *cdrom_media_changed()*, we can implement separate
379 queues for the VFS and a new *ioctl()* function that can report device
380 changes to software (e. g., an auto-mounting daemon).
384 int tray_move(struct cdrom_device_info *cdi, int position)
386 This function, if implemented, should control the tray movement. (No
387 other function should control this.) The parameter *position* controls
388 the desired direction of movement:
393 This function returns 0 upon success, and a non-zero value upon
394 error. Note that if the tray is already in the desired position, no
395 action need be taken, and the return value should be 0.
399 int lock_door(struct cdrom_device_info *cdi, int lock)
401 This function (and no other code) controls locking of the door, if the
402 drive allows this. The value of *lock* controls the desired locking
405 - 0 Unlock door, manual opening is allowed
406 - 1 Lock door, tray cannot be ejected manually
408 This function returns 0 upon success, and a non-zero value upon
409 error. Note that if the door is already in the requested state, no
410 action need be taken, and the return value should be 0.
414 int select_speed(struct cdrom_device_info *cdi, int speed)
416 Some CD-ROM drives are capable of changing their head-speed. There
417 are several reasons for changing the speed of a CD-ROM drive. Badly
418 pressed CD-ROM s may benefit from less-than-maximum head rate. Modern
419 CD-ROM drives can obtain very high head rates (up to *24x* is
420 common). It has been reported that these drives can make reading
421 errors at these high speeds, reducing the speed can prevent data loss
422 in these circumstances. Finally, some of these drives can
423 make an annoyingly loud noise, which a lower speed may reduce.
425 This function specifies the speed at which data is read or audio is
426 played back. The value of *speed* specifies the head-speed of the
427 drive, measured in units of standard cdrom speed (176kB/sec raw data
428 or 150kB/sec file system data). So to request that a CD-ROM drive
429 operate at 300kB/sec you would call the CDROM_SELECT_SPEED *ioctl*
430 with *speed=2*. The special value `0` means `auto-selection`, i. e.,
431 maximum data-rate or real-time audio rate. If the drive doesn't have
432 this `auto-selection` capability, the decision should be made on the
433 current disc loaded and the return value should be positive. A negative
434 return value indicates an error.
438 int select_disc(struct cdrom_device_info *cdi, int number)
440 If the drive can store multiple discs (a juke-box) this function
441 will perform disc selection. It should return the number of the
442 selected disc on success, a negative value on error. Currently, only
443 the ide-cd driver supports this functionality.
447 int get_last_session(struct cdrom_device_info *cdi,
448 struct cdrom_multisession *ms_info)
450 This function should implement the old corresponding *ioctl()*. For
451 device *cdi->dev*, the start of the last session of the current disc
452 should be returned in the pointer argument *ms_info*. Note that
453 routines in `cdrom.c` have sanitized this argument: its requested
454 format will **always** be of the type *CDROM_LBA* (linear block
455 addressing mode), whatever the calling software requested. But
456 sanitization goes even further: the low-level implementation may
457 return the requested information in *CDROM_MSF* format if it wishes so
458 (setting the *ms_info->addr_format* field appropriately, of
459 course) and the routines in `cdrom.c` will make the transformation if
460 necessary. The return value is 0 upon success.
464 int get_mcn(struct cdrom_device_info *cdi,
465 struct cdrom_mcn *mcn)
467 Some discs carry a `Media Catalog Number` (MCN), also called
468 `Universal Product Code` (UPC). This number should reflect the number
469 that is generally found in the bar-code on the product. Unfortunately,
470 the few discs that carry such a number on the disc don't even use the
471 same format. The return argument to this function is a pointer to a
472 pre-declared memory region of type *struct cdrom_mcn*. The MCN is
473 expected as a 13-character string, terminated by a null-character.
477 int reset(struct cdrom_device_info *cdi)
479 This call should perform a hard-reset on the drive (although in
480 circumstances that a hard-reset is necessary, a drive may very well not
481 listen to commands anymore). Preferably, control is returned to the
482 caller only after the drive has finished resetting. If the drive is no
483 longer listening, it may be wise for the underlying low-level cdrom
488 int audio_ioctl(struct cdrom_device_info *cdi,
489 unsigned int cmd, void *arg)
491 Some of the CD-ROM-\ *ioctl()*\ 's defined in `cdrom.h` can be
492 implemented by the routines described above, and hence the function
493 *cdrom_ioctl* will use those. However, most *ioctl()*\ 's deal with
494 audio-control. We have decided to leave these to be accessed through a
495 single function, repeating the arguments *cmd* and *arg*. Note that
496 the latter is of type *void*, rather than *unsigned long int*.
497 The routine *cdrom_ioctl()* does do some useful things,
498 though. It sanitizes the address format type to *CDROM_MSF* (Minutes,
499 Seconds, Frames) for all audio calls. It also verifies the memory
500 location of *arg*, and reserves stack-memory for the argument. This
501 makes implementation of the *audio_ioctl()* much simpler than in the
502 old driver scheme. For example, you may look up the function
503 *cm206_audio_ioctl()* `cm206.c` that should be updated with
506 An unimplemented ioctl should return *-ENOSYS*, but a harmless request
507 (e. g., *CDROMSTART*) may be ignored by returning 0 (success). Other
508 errors should be according to the standards, whatever they are. When
509 an error is returned by the low-level driver, the Uniform CD-ROM Driver
510 tries whenever possible to return the error code to the calling program.
511 (We may decide to sanitize the return value in *cdrom_ioctl()* though, in
512 order to guarantee a uniform interface to the audio-player software.)
516 int dev_ioctl(struct cdrom_device_info *cdi,
517 unsigned int cmd, unsigned long arg)
519 Some *ioctl()'s* seem to be specific to certain CD-ROM drives. That is,
520 they are introduced to service some capabilities of certain drives. In
521 fact, there are 6 different *ioctl()'s* for reading data, either in some
522 particular kind of format, or audio data. Not many drives support
523 reading audio tracks as data, I believe this is because of protection
524 of copyrights of artists. Moreover, I think that if audio-tracks are
525 supported, it should be done through the VFS and not via *ioctl()'s*. A
526 problem here could be the fact that audio-frames are 2352 bytes long,
527 so either the audio-file-system should ask for 75264 bytes at once
528 (the least common multiple of 512 and 2352), or the drivers should
529 bend their backs to cope with this incoherence (to which I would be
530 opposed). Furthermore, it is very difficult for the hardware to find
531 the exact frame boundaries, since there are no synchronization headers
532 in audio frames. Once these issues are resolved, this code should be
533 standardized in `cdrom.c`.
535 Because there are so many *ioctl()'s* that seem to be introduced to
536 satisfy certain drivers [#f2]_, any non-standard *ioctl()*\ s
537 are routed through the call *dev_ioctl()*. In principle, `private`
538 *ioctl()*\ 's should be numbered after the device's major number, and not
539 the general CD-ROM *ioctl* number, `0x53`. Currently the
540 non-supported *ioctl()'s* are:
542 CDROMREADMODE1, CDROMREADMODE2, CDROMREADAUDIO, CDROMREADRAW,
543 CDROMREADCOOKED, CDROMSEEK, CDROMPLAY-BLK and CDROM-READALL
547 Is there software around that actually uses these? I'd be interested!
549 .. _cdrom_capabilities:
554 Instead of just implementing some *ioctl* calls, the interface in
555 `cdrom.c` supplies the possibility to indicate the **capabilities**
556 of a CD-ROM drive. This can be done by ORing any number of
557 capability-constants that are defined in `cdrom.h` at the registration
558 phase. Currently, the capabilities are any of::
560 CDC_CLOSE_TRAY /* can close tray by software control */
561 CDC_OPEN_TRAY /* can open tray */
562 CDC_LOCK /* can lock and unlock the door */
563 CDC_SELECT_SPEED /* can select speed, in units of * sim*150 ,kB/s */
564 CDC_SELECT_DISC /* drive is juke-box */
565 CDC_MULTI_SESSION /* can read sessions *> rm1* */
566 CDC_MCN /* can read Media Catalog Number */
567 CDC_MEDIA_CHANGED /* can report if disc has changed */
568 CDC_PLAY_AUDIO /* can perform audio-functions (play, pause, etc) */
569 CDC_RESET /* hard reset device */
570 CDC_IOCTLS /* driver has non-standard ioctls */
571 CDC_DRIVE_STATUS /* driver implements drive status */
573 The capability flag is declared *const*, to prevent drivers from
574 accidentally tampering with the contents. The capability fags actually
575 inform `cdrom.c` of what the driver can do. If the drive found
576 by the driver does not have the capability, is can be masked out by
577 the *cdrom_device_info* variable *mask*. For instance, the SCSI CD-ROM
578 driver has implemented the code for loading and ejecting CD-ROM's, and
579 hence its corresponding flags in *capability* will be set. But a SCSI
580 CD-ROM drive might be a caddy system, which can't load the tray, and
581 hence for this drive the *cdrom_device_info* struct will have set
582 the *CDC_CLOSE_TRAY* bit in *mask*.
584 In the file `cdrom.c` you will encounter many constructions of the type::
586 if (cdo->capability & ∼cdi->mask & CDC _⟨capability⟩) ...
588 There is no *ioctl* to set the mask... The reason is that
589 I think it is better to control the **behavior** rather than the
595 A final flag register controls the **behavior** of the CD-ROM
596 drives, in order to satisfy different users' wishes, hopefully
597 independently of the ideas of the respective author who happened to
598 have made the drive's support available to the Linux community. The
599 current behavior options are::
601 CDO_AUTO_CLOSE /* try to close tray upon device open() */
602 CDO_AUTO_EJECT /* try to open tray on last device close() */
603 CDO_USE_FFLAGS /* use file_pointer->f_flags to indicate purpose for open() */
604 CDO_LOCK /* try to lock door if device is opened */
605 CDO_CHECK_TYPE /* ensure disc type is data if opened for data */
607 The initial value of this register is
608 `CDO_AUTO_CLOSE | CDO_USE_FFLAGS | CDO_LOCK`, reflecting my own view on user
609 interface and software standards. Before you protest, there are two
610 new *ioctl()'s* implemented in `cdrom.c`, that allow you to control the
611 behavior by software. These are::
613 CDROM_SET_OPTIONS /* set options specified in (int)arg */
614 CDROM_CLEAR_OPTIONS /* clear options specified in (int)arg */
616 One option needs some more explanation: *CDO_USE_FFLAGS*. In the next
617 newsection we explain what the need for this option is.
619 A software package `setcd`, available from the Debian distribution
620 and `sunsite.unc.edu`, allows user level control of these flags.
623 The need to know the purpose of opening the CD-ROM device
624 =========================================================
626 Traditionally, Unix devices can be used in two different `modes`,
627 either by reading/writing to the device file, or by issuing
628 controlling commands to the device, by the device's *ioctl()*
629 call. The problem with CD-ROM drives, is that they can be used for
630 two entirely different purposes. One is to mount removable
631 file systems, CD-ROM's, the other is to play audio CD's. Audio commands
632 are implemented entirely through *ioctl()\'s*, presumably because the
633 first implementation (SUN?) has been such. In principle there is
634 nothing wrong with this, but a good control of the `CD player` demands
635 that the device can **always** be opened in order to give the
636 *ioctl* commands, regardless of the state the drive is in.
638 On the other hand, when used as a removable-media disc drive (what the
639 original purpose of CD-ROM s is) we would like to make sure that the
640 disc drive is ready for operation upon opening the device. In the old
641 scheme, some CD-ROM drivers don't do any integrity checking, resulting
642 in a number of i/o errors reported by the VFS to the kernel when an
643 attempt for mounting a CD-ROM on an empty drive occurs. This is not a
644 particularly elegant way to find out that there is no CD-ROM inserted;
645 it more-or-less looks like the old IBM-PC trying to read an empty floppy
646 drive for a couple of seconds, after which the system complains it
647 can't read from it. Nowadays we can **sense** the existence of a
648 removable medium in a drive, and we believe we should exploit that
649 fact. An integrity check on opening of the device, that verifies the
650 availability of a CD-ROM and its correct type (data), would be
653 These two ways of using a CD-ROM drive, principally for data and
654 secondarily for playing audio discs, have different demands for the
655 behavior of the *open()* call. Audio use simply wants to open the
656 device in order to get a file handle which is needed for issuing
657 *ioctl* commands, while data use wants to open for correct and
658 reliable data transfer. The only way user programs can indicate what
659 their *purpose* of opening the device is, is through the *flags*
660 parameter (see `open(2)`). For CD-ROM devices, these flags aren't
661 implemented (some drivers implement checking for write-related flags,
662 but this is not strictly necessary if the device file has correct
663 permission flags). Most option flags simply don't make sense to
664 CD-ROM devices: *O_CREAT*, *O_NOCTTY*, *O_TRUNC*, *O_APPEND*, and
665 *O_SYNC* have no meaning to a CD-ROM.
667 We therefore propose to use the flag *O_NONBLOCK* to indicate
668 that the device is opened just for issuing *ioctl*
669 commands. Strictly, the meaning of *O_NONBLOCK* is that opening and
670 subsequent calls to the device don't cause the calling process to
671 wait. We could interpret this as don't wait until someone has
672 inserted some valid data-CD-ROM. Thus, our proposal of the
673 implementation for the *open()* call for CD-ROM s is:
675 - If no other flags are set than *O_RDONLY*, the device is opened
676 for data transfer, and the return value will be 0 only upon successful
677 initialization of the transfer. The call may even induce some actions
678 on the CD-ROM, such as closing the tray.
679 - If the option flag *O_NONBLOCK* is set, opening will always be
680 successful, unless the whole device doesn't exist. The drive will take
681 no actions whatsoever.
683 And what about standards?
684 -------------------------
686 You might hesitate to accept this proposal as it comes from the
687 Linux community, and not from some standardizing institute. What
688 about SUN, SGI, HP and all those other Unix and hardware vendors?
689 Well, these companies are in the lucky position that they generally
690 control both the hardware and software of their supported products,
691 and are large enough to set their own standard. They do not have to
692 deal with a dozen or more different, competing hardware
693 configurations\ [#f3]_.
697 Incidentally, I think that SUN's approach to mounting CD-ROM s is very
698 good in origin: under Solaris a volume-daemon automatically mounts a
699 newly inserted CD-ROM under `/cdrom/*<volume-name>*`.
701 In my opinion they should have pushed this
702 further and have **every** CD-ROM on the local area network be
703 mounted at the similar location, i. e., no matter in which particular
704 machine you insert a CD-ROM, it will always appear at the same
705 position in the directory tree, on every system. When I wanted to
706 implement such a user-program for Linux, I came across the
707 differences in behavior of the various drivers, and the need for an
708 *ioctl* informing about media changes.
710 We believe that using *O_NONBLOCK* to indicate that a device is being opened
711 for *ioctl* commands only can be easily introduced in the Linux
712 community. All the CD-player authors will have to be informed, we can
713 even send in our own patches to the programs. The use of *O_NONBLOCK*
714 has most likely no influence on the behavior of the CD-players on
715 other operating systems than Linux. Finally, a user can always revert
716 to old behavior by a call to
717 *ioctl(file_descriptor, CDROM_CLEAR_OPTIONS, CDO_USE_FFLAGS)*.
719 The preferred strategy of *open()*
720 ----------------------------------
722 The routines in `cdrom.c` are designed in such a way that run-time
723 configuration of the behavior of CD-ROM devices (of **any** type)
724 can be carried out, by the *CDROM_SET/CLEAR_OPTIONS* *ioctls*. Thus, various
725 modes of operation can be set:
727 `CDO_AUTO_CLOSE | CDO_USE_FFLAGS | CDO_LOCK`
728 This is the default setting. (With *CDO_CHECK_TYPE* it will be better, in
729 the future.) If the device is not yet opened by any other process, and if
730 the device is being opened for data (*O_NONBLOCK* is not set) and the
731 tray is found to be open, an attempt to close the tray is made. Then,
732 it is verified that a disc is in the drive and, if *CDO_CHECK_TYPE* is
733 set, that it contains tracks of type `data mode 1`. Only if all tests
734 are passed is the return value zero. The door is locked to prevent file
735 system corruption. If the drive is opened for audio (*O_NONBLOCK* is
736 set), no actions are taken and a value of 0 will be returned.
738 `CDO_AUTO_CLOSE | CDO_AUTO_EJECT | CDO_LOCK`
739 This mimics the behavior of the current sbpcd-driver. The option flags are
740 ignored, the tray is closed on the first open, if necessary. Similarly,
741 the tray is opened on the last release, i. e., if a CD-ROM is unmounted,
742 it is automatically ejected, such that the user can replace it.
744 We hope that these option can convince everybody (both driver
745 maintainers and user program developers) to adopt the new CD-ROM
746 driver scheme and option flag interpretation.
748 Description of routines in `cdrom.c`
749 ====================================
751 Only a few routines in `cdrom.c` are exported to the drivers. In this
752 new section we will discuss these, as well as the functions that `take
753 over' the CD-ROM interface to the kernel. The header file belonging
754 to `cdrom.c` is called `cdrom.h`. Formerly, some of the contents of this
755 file were placed in the file `ucdrom.h`, but this file has now been
756 merged back into `cdrom.h`.
760 struct file_operations cdrom_fops
762 The contents of this structure were described in cdrom_api_.
763 A pointer to this structure is assigned to the *fops* field
764 of the *struct gendisk*.
768 int register_cdrom(struct cdrom_device_info *cdi)
770 This function is used in about the same way one registers *cdrom_fops*
771 with the kernel, the device operations and information structures,
772 as described in cdrom_api_, should be registered with the
773 Uniform CD-ROM Driver::
775 register_cdrom(&<device>_info);
778 This function returns zero upon success, and non-zero upon
779 failure. The structure *<device>_info* should have a pointer to the
780 driver's *<device>_dops*, as in::
782 struct cdrom_device_info <device>_info = {
787 Note that a driver must have one static structure, *<device>_dops*, while
788 it may have as many structures *<device>_info* as there are minor devices
789 active. *Register_cdrom()* builds a linked list from these.
794 void unregister_cdrom(struct cdrom_device_info *cdi)
796 Unregistering device *cdi* with minor number *MINOR(cdi->dev)* removes
797 the minor device from the list. If it was the last registered minor for
798 the low-level driver, this disconnects the registered device-operation
799 routines from the CD-ROM interface. This function returns zero upon
800 success, and non-zero upon failure.
804 int cdrom_open(struct inode * ip, struct file * fp)
806 This function is not called directly by the low-level drivers, it is
807 listed in the standard *cdrom_fops*. If the VFS opens a file, this
808 function becomes active. A strategy is implemented in this routine,
809 taking care of all capabilities and options that are set in the
810 *cdrom_device_ops* connected to the device. Then, the program flow is
811 transferred to the device_dependent *open()* call.
815 void cdrom_release(struct inode *ip, struct file *fp)
817 This function implements the reverse-logic of *cdrom_open()*, and then
818 calls the device-dependent *release()* routine. When the use-count has
819 reached 0, the allocated buffers are flushed by calls to *sync_dev(dev)*
820 and *invalidate_buffers(dev)*.
827 int cdrom_ioctl(struct inode *ip, struct file *fp,
828 unsigned int cmd, unsigned long arg)
830 This function handles all the standard *ioctl* requests for CD-ROM
831 devices in a uniform way. The different calls fall into three
832 categories: *ioctl()'s* that can be directly implemented by device
833 operations, ones that are routed through the call *audio_ioctl()*, and
834 the remaining ones, that are presumable device-dependent. Generally, a
835 negative return value indicates an error.
837 Directly implemented *ioctl()'s*
838 --------------------------------
840 The following `old` CD-ROM *ioctl()*\ 's are implemented by directly
841 calling device-operations in *cdrom_device_ops*, if implemented and
845 Requests the last session on a CD-ROM.
851 If *arg\not=0*, set behavior to auto-close (close
852 tray on first open) and auto-eject (eject on last release), otherwise
853 set behavior to non-moving on *open()* and *release()* calls.
855 Get the Media Catalog Number from a CD.
857 *Ioctl*s routed through *audio_ioctl()*
858 ---------------------------------------
860 The following set of *ioctl()'s* are all implemented through a call to
861 the *cdrom_fops* function *audio_ioctl()*. Memory checks and
862 allocation are performed in *cdrom_ioctl()*, and also sanitization of
863 address format (*CDROM_LBA*/*CDROM_MSF*) is done.
866 Get sub-channel data in argument *arg* of type
867 `struct cdrom_subchnl *`.
869 Read Table of Contents header, in *arg* of type
870 `struct cdrom_tochdr *`.
872 Read a Table of Contents entry in *arg* and specified by *arg*
873 of type `struct cdrom_tocentry *`.
875 Play audio fragment specified in Minute, Second, Frame format,
876 delimited by *arg* of type `struct cdrom_msf *`.
878 Play audio fragment in track-index format delimited by *arg*
879 of type `struct cdrom_ti *`.
881 Set volume specified by *arg* of type `struct cdrom_volctrl *`.
883 Read volume into by *arg* of type `struct cdrom_volctrl *`.
887 Stop playback of audio fragment.
889 Pause playback of audio fragment.
893 New *ioctl()'s* in `cdrom.c`
894 ----------------------------
896 The following *ioctl()'s* have been introduced to allow user programs to
897 control the behavior of individual CD-ROM devices. New *ioctl*
898 commands can be identified by the underscores in their names.
901 Set options specified by *arg*. Returns the option flag register
902 after modification. Use *arg = \rm0* for reading the current flags.
903 `CDROM_CLEAR_OPTIONS`
904 Clear options specified by *arg*. Returns the option flag register
907 Select head-rate speed of disc specified as by *arg* in units
908 of standard cdrom speed (176\,kB/sec raw data or
909 150kB/sec file system data). The value 0 means `auto-select`,
910 i. e., play audio discs at real time and data discs at maximum speed.
911 The value *arg* is checked against the maximum head rate of the
912 drive found in the *cdrom_dops*.
914 Select disc numbered *arg* from a juke-box.
916 First disc is numbered 0. The number *arg* is checked against the
917 maximum number of discs in the juke-box found in the *cdrom_dops*.
918 `CDROM_MEDIA_CHANGED`
919 Returns 1 if a disc has been changed since the last call.
920 Note that calls to *cdrom_media_changed* by the VFS are treated
921 by an independent queue, so both mechanisms will detect a
922 media change once. For juke-boxes, an extra argument *arg*
923 specifies the slot for which the information is given. The special
924 value *CDSL_CURRENT* requests that information about the currently
925 selected slot be returned.
927 Returns the status of the drive by a call to
928 *drive_status()*. Return values are defined in cdrom_drive_status_.
929 Note that this call doesn't return information on the
930 current playing activity of the drive; this can be polled through
931 an *ioctl* call to *CDROMSUBCHNL*. For juke-boxes, an extra argument
932 *arg* specifies the slot for which (possibly limited) information is
933 given. The special value *CDSL_CURRENT* requests that information
934 about the currently selected slot be returned.
936 Returns the type of the disc currently in the drive.
937 It should be viewed as a complement to *CDROM_DRIVE_STATUS*.
938 This *ioctl* can provide *some* information about the current
939 disc that is inserted in the drive. This functionality used to be
940 implemented in the low level drivers, but is now carried out
941 entirely in Uniform CD-ROM Driver.
943 The history of development of the CD's use as a carrier medium for
944 various digital information has lead to many different disc types.
945 This *ioctl* is useful only in the case that CDs have \emph {only
946 one} type of data on them. While this is often the case, it is
947 also very common for CDs to have some tracks with data, and some
948 tracks with audio. Because this is an existing interface, rather
949 than fixing this interface by changing the assumptions it was made
950 under, thereby breaking all user applications that use this
951 function, the Uniform CD-ROM Driver implements this *ioctl* as
952 follows: If the CD in question has audio tracks on it, and it has
953 absolutely no CD-I, XA, or data tracks on it, it will be reported
954 as *CDS_AUDIO*. If it has both audio and data tracks, it will
955 return *CDS_MIXED*. If there are no audio tracks on the disc, and
956 if the CD in question has any CD-I tracks on it, it will be
957 reported as *CDS_XA_2_2*. Failing that, if the CD in question
958 has any XA tracks on it, it will be reported as *CDS_XA_2_1*.
959 Finally, if the CD in question has any data tracks on it,
960 it will be reported as a data CD (*CDS_DATA_1*).
962 This *ioctl* can return::
964 CDS_NO_INFO /* no information available */
965 CDS_NO_DISC /* no disc is inserted, or tray is opened */
966 CDS_AUDIO /* Audio disc (2352 audio bytes/frame) */
967 CDS_DATA_1 /* data disc, mode 1 (2048 user bytes/frame) */
968 CDS_XA_2_1 /* mixed data (XA), mode 2, form 1 (2048 user bytes) */
969 CDS_XA_2_2 /* mixed data (XA), mode 2, form 1 (2324 user bytes) */
970 CDS_MIXED /* mixed audio/data disc */
972 For some information concerning frame layout of the various disc
973 types, see a recent version of `cdrom.h`.
975 `CDROM_CHANGER_NSLOTS`
976 Returns the number of slots in a juke-box.
979 `CDROM_GET_CAPABILITY`
980 Returns the *capability* flags for the drive. Refer to section
981 cdrom_capabilities_ for more information on these flags.
983 Locks the door of the drive. `arg == 0` unlocks the door,
984 any other value locks it.
986 Turns on debugging info. Only root is allowed to do this.
987 Same semantics as CDROM_LOCKDOOR.
990 Device dependent *ioctl()'s*
991 ----------------------------
993 Finally, all other *ioctl()'s* are passed to the function *dev_ioctl()*,
994 if implemented. No memory allocation or verification is carried out.
996 How to update your driver
997 =========================
999 - Make a backup of your current driver.
1000 - Get hold of the files `cdrom.c` and `cdrom.h`, they should be in
1001 the directory tree that came with this documentation.
1002 - Make sure you include `cdrom.h`.
1003 - Change the 3rd argument of *register_blkdev* from `&<your-drive>_fops`
1005 - Just after that line, add the following to register with the Uniform
1008 register_cdrom(&<your-drive>_info);*
1010 Similarly, add a call to *unregister_cdrom()* at the appropriate place.
1011 - Copy an example of the device-operations *struct* to your
1012 source, e. g., from `cm206.c` *cm206_dops*, and change all
1013 entries to names corresponding to your driver, or names you just
1014 happen to like. If your driver doesn't support a certain function,
1015 make the entry *NULL*. At the entry *capability* you should list all
1016 capabilities your driver currently supports. If your driver
1017 has a capability that is not listed, please send me a message.
1018 - Copy the *cdrom_device_info* declaration from the same example
1019 driver, and modify the entries according to your needs. If your
1020 driver dynamically determines the capabilities of the hardware, this
1021 structure should also be declared dynamically.
1022 - Implement all functions in your `<device>_dops` structure,
1023 according to prototypes listed in `cdrom.h`, and specifications given
1024 in cdrom_api_. Most likely you have already implemented
1025 the code in a large part, and you will almost certainly need to adapt the
1026 prototype and return values.
1027 - Rename your `<device>_ioctl()` function to *audio_ioctl* and
1028 change the prototype a little. Remove entries listed in the first
1029 part in cdrom_ioctl_, if your code was OK, these are
1030 just calls to the routines you adapted in the previous step.
1031 - You may remove all remaining memory checking code in the
1032 *audio_ioctl()* function that deals with audio commands (these are
1033 listed in the second part of cdrom_ioctl_. There is no
1034 need for memory allocation either, so most *case*s in the *switch*
1035 statement look similar to::
1037 case CDROMREADTOCENTRY:
1038 get_toc_entry\bigl((struct cdrom_tocentry *) arg);
1040 - All remaining *ioctl* cases must be moved to a separate
1041 function, *<device>_ioctl*, the device-dependent *ioctl()'s*. Note that
1042 memory checking and allocation must be kept in this code!
1043 - Change the prototypes of *<device>_open()* and
1044 *<device>_release()*, and remove any strategic code (i. e., tray
1045 movement, door locking, etc.).
1046 - Try to recompile the drivers. We advise you to use modules, both
1047 for `cdrom.o` and your driver, as debugging is much easier this
1053 Thanks to all the people involved. First, Erik Andersen, who has
1054 taken over the torch in maintaining `cdrom.c` and integrating much
1055 CD-ROM-related code in the 2.1-kernel. Thanks to Scott Snyder and
1056 Gerd Knorr, who were the first to implement this interface for SCSI
1057 and IDE-CD drivers and added many ideas for extension of the data
1058 structures relative to kernel~2.0. Further thanks to Heiko Eißfeldt,
1059 Thomas Quinot, Jon Tombs, Ken Pizzini, Eberhard Mönkeberg and Andrew Kroll,
1060 the Linux CD-ROM device driver developers who were kind
1061 enough to give suggestions and criticisms during the writing. Finally
1062 of course, I want to thank Linus Torvalds for making this possible in
projects / linux-2.6-microblaze.git / blob