PageRenderTime 61ms CodeModel.GetById 21ms RepoModel.GetById 0ms app.codeStats 0ms

TeX | 1022 lines | 911 code | 108 blank | 3 comment | 96 complexity | 42333639ad8a0d0b53214b136f611432 MD5 | raw file
Possible License(s): GPL-2.0, LGPL-2.0, AGPL-1.0
  1. \documentclass{article}
  2. \def\version{$Id: cdrom-standard.tex,v 1.9 1997/12/28 15:42:49 david Exp $}
  3. \newcommand{\newsection}[1]{\newpage\section{#1}}
  4. \evensidemargin=0pt
  5. \oddsidemargin=0pt
  6. \topmargin=-\headheight \advance\topmargin by -\headsep
  7. \textwidth=15.99cm \textheight=24.62cm % normal A4, 1'' margin
  8. \def\linux{{\sc Linux}}
  9. \def\cdrom{{\sc cd-rom}}
  10. \def\UCD{{\sc Uniform cd-rom Driver}}
  11. \def\cdromc{{\tt {cdrom.c}}}
  12. \def\cdromh{{\tt {cdrom.h}}}
  13. \def\fo{\sl} % foreign words
  14. \def\ie{{\fo i.e.}}
  15. \def\eg{{\fo e.g.}}
  16. \everymath{\it} \everydisplay{\it}
  17. \catcode `\_=\active \def_{\_\penalty100 }
  18. \catcode`\<=\active \def<#1>{{\langle\hbox{\rm#1}\rangle}}
  19. \begin{document}
  20. \title{A \linux\ \cdrom\ standard}
  21. \author{David van Leeuwen\\{\normalsize\tt}
  22. \\{\footnotesize updated by Erik Andersen {\tt(}}
  23. \\{\footnotesize updated by Jens Axboe {\tt(}}}
  24. \date{12 March 1999}
  25. \maketitle
  26. \newsection{Introduction}
  27. \linux\ is probably the Unix-like operating system that supports
  28. the widest variety of hardware devices. The reasons for this are
  29. presumably
  30. \begin{itemize}
  31. \item
  32. The large list of hardware devices available for the many platforms
  33. that \linux\ now supports (\ie, i386-PCs, Sparc Suns, etc.)
  34. \item
  35. The open design of the operating system, such that anybody can write a
  36. driver for \linux.
  37. \item
  38. There is plenty of source code around as examples of how to write a driver.
  39. \end{itemize}
  40. The openness of \linux, and the many different types of available
  41. hardware has allowed \linux\ to support many different hardware devices.
  42. Unfortunately, the very openness that has allowed \linux\ to support
  43. all these different devices has also allowed the behavior of each
  44. device driver to differ significantly from one device to another.
  45. This divergence of behavior has been very significant for \cdrom\
  46. devices; the way a particular drive reacts to a `standard' $ioctl()$
  47. call varies greatly from one device driver to another. To avoid making
  48. their drivers totally inconsistent, the writers of \linux\ \cdrom\
  49. drivers generally created new device drivers by understanding, copying,
  50. and then changing an existing one. Unfortunately, this practice did not
  51. maintain uniform behavior across all the \linux\ \cdrom\ drivers.
  52. This document describes an effort to establish Uniform behavior across
  53. all the different \cdrom\ device drivers for \linux. This document also
  54. defines the various $ioctl$s, and how the low-level \cdrom\ device
  55. drivers should implement them. Currently (as of the \linux\ 2.1.$x$
  56. development kernels) several low-level \cdrom\ device drivers, including
  57. both IDE/ATAPI and SCSI, now use this Uniform interface.
  58. When the \cdrom\ was developed, the interface between the \cdrom\ drive
  59. and the computer was not specified in the standards. As a result, many
  60. different \cdrom\ interfaces were developed. Some of them had their
  61. own proprietary design (Sony, Mitsumi, Panasonic, Philips), other
  62. manufacturers adopted an existing electrical interface and changed
  63. the functionality (CreativeLabs/SoundBlaster, Teac, Funai) or simply
  64. adapted their drives to one or more of the already existing electrical
  65. interfaces (Aztech, Sanyo, Funai, Vertos, Longshine, Optics Storage and
  66. most of the `NoName' manufacturers). In cases where a new drive really
  67. brought its own interface or used its own command set and flow control
  68. scheme, either a separate driver had to be written, or an existing
  69. driver had to be enhanced. History has delivered us \cdrom\ support for
  70. many of these different interfaces. Nowadays, almost all new \cdrom\
  71. drives are either IDE/ATAPI or SCSI, and it is very unlikely that any
  72. manufacturer will create a new interface. Even finding drives for the
  73. old proprietary interfaces is getting difficult.
  74. When (in the 1.3.70's) I looked at the existing software interface,
  75. which was expressed through \cdromh, it appeared to be a rather wild
  76. set of commands and data formats.\footnote{I cannot recollect what
  77. kernel version I looked at, then, presumably 1.2.13 and 1.3.34---the
  78. latest kernel that I was indirectly involved in.} It seemed that many
  79. features of the software interface had been added to accommodate the
  80. capabilities of a particular drive, in an {\fo ad hoc\/} manner. More
  81. importantly, it appeared that the behavior of the `standard' commands
  82. was different for most of the different drivers: \eg, some drivers
  83. close the tray if an $open()$ call occurs when the tray is open, while
  84. others do not. Some drivers lock the door upon opening the device, to
  85. prevent an incoherent file system, but others don't, to allow software
  86. ejection. Undoubtedly, the capabilities of the different drives vary,
  87. but even when two drives have the same capability their drivers'
  88. behavior was usually different.
  89. I decided to start a discussion on how to make all the \linux\ \cdrom\
  90. drivers behave more uniformly. I began by contacting the developers of
  91. the many \cdrom\ drivers found in the \linux\ kernel. Their reactions
  92. encouraged me to write the \UCD\ which this document is intended to
  93. describe. The implementation of the \UCD\ is in the file \cdromc. This
  94. driver is intended to be an additional software layer that sits on top
  95. of the low-level device drivers for each \cdrom\ drive. By adding this
  96. additional layer, it is possible to have all the different \cdrom\
  97. devices behave {\em exactly\/} the same (insofar as the underlying
  98. hardware will allow).
  99. The goal of the \UCD\ is {\em not\/} to alienate driver developers who
  100. have not yet taken steps to support this effort. The goal of \UCD\ is
  101. simply to give people writing application programs for \cdrom\ drives
  102. {\em one\/} \linux\ \cdrom\ interface with consistent behavior for all
  103. \cdrom\ devices. In addition, this also provides a consistent interface
  104. between the low-level device driver code and the \linux\ kernel. Care
  105. is taken that 100\,\% compatibility exists with the data structures and
  106. programmer's interface defined in \cdromh. This guide was written to
  107. help \cdrom\ driver developers adapt their code to use the \UCD\ code
  108. defined in \cdromc.
  109. Personally, I think that the most important hardware interfaces are
  110. the IDE/ATAPI drives and, of course, the SCSI drives, but as prices
  111. of hardware drop continuously, it is also likely that people may have
  112. more than one \cdrom\ drive, possibly of mixed types. It is important
  113. that these drives behave in the same way. In December 1994, one of the
  114. cheapest \cdrom\ drives was a Philips cm206, a double-speed proprietary
  115. drive. In the months that I was busy writing a \linux\ driver for it,
  116. proprietary drives became obsolete and IDE/ATAPI drives became the
  117. standard. At the time of the last update to this document (November
  118. 1997) it is becoming difficult to even {\em find} anything less than a
  119. 16 speed \cdrom\ drive, and 24 speed drives are common.
  120. \newsection{Standardizing through another software level}
  121. \label{cdrom.c}
  122. At the time this document was conceived, all drivers directly
  123. implemented the \cdrom\ $ioctl()$ calls through their own routines. This
  124. led to the danger of different drivers forgetting to do important things
  125. like checking that the user was giving the driver valid data. More
  126. importantly, this led to the divergence of behavior, which has already
  127. been discussed.
  128. For this reason, the \UCD\ was created to enforce consistent \cdrom\
  129. drive behavior, and to provide a common set of services to the various
  130. low-level \cdrom\ device drivers. The \UCD\ now provides another
  131. software-level, that separates the $ioctl()$ and $open()$ implementation
  132. from the actual hardware implementation. Note that this effort has
  133. made few changes which will affect a user's application programs. The
  134. greatest change involved moving the contents of the various low-level
  135. \cdrom\ drivers' header files to the kernel's cdrom directory. This was
  136. done to help ensure that the user is only presented with only one cdrom
  137. interface, the interface defined in \cdromh.
  138. \cdrom\ drives are specific enough (\ie, different from other
  139. block-devices such as floppy or hard disc drives), to define a set
  140. of common {\em \cdrom\ device operations}, $<cdrom-device>_dops$.
  141. These operations are different from the classical block-device file
  142. operations, $<block-device>_fops$.
  143. The routines for the \UCD\ interface level are implemented in the file
  144. \cdromc. In this file, the \UCD\ interfaces with the kernel as a block
  145. device by registering the following general $struct\ file_operations$:
  146. $$
  147. \halign{$#$\ \hfil&$#$\ \hfil&$/*$ \rm# $*/$\hfil\cr
  148. struct& file_operations\ cdrom_fops = \{\hidewidth\cr
  149. &NULL, & lseek \cr
  150. &block_read, & read---general block-dev read \cr
  151. &block_write, & write---general block-dev write \cr
  152. &NULL, & readdir \cr
  153. &NULL, & select \cr
  154. &cdrom_ioctl, & ioctl \cr
  155. &NULL, & mmap \cr
  156. &cdrom_open, & open \cr
  157. &cdrom_release, & release \cr
  158. &NULL, & fsync \cr
  159. &NULL, & fasync \cr
  160. &cdrom_media_changed, & media change \cr
  161. &NULL & revalidate \cr
  162. \};\cr
  163. }
  164. $$
  165. Every active \cdrom\ device shares this $struct$. The routines
  166. declared above are all implemented in \cdromc, since this file is the
  167. place where the behavior of all \cdrom-devices is defined and
  168. standardized. The actual interface to the various types of \cdrom\
  169. hardware is still performed by various low-level \cdrom-device
  170. drivers. These routines simply implement certain {\em capabilities\/}
  171. that are common to all \cdrom\ (and really, all removable-media
  172. devices).
  173. Registration of a low-level \cdrom\ device driver is now done through
  174. the general routines in \cdromc, not through the Virtual File System
  175. (VFS) any more. The interface implemented in \cdromc\ 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:
  179. \begin{description}
  180. \item[$cdrom_device_ops$]
  181. This structure contains information about the low-level driver for a
  182. \cdrom\ 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).
  185. \item[$cdrom_device_info$]
  186. This structure contains information about a particular \cdrom\ drive,
  187. such as its device name, speed, etc. This structure is conceptually
  188. connected to the minor number of the device.
  189. \end{description}
  190. Registering a particular \cdrom\ drive with the \UCD\ is done by the
  191. low-level device driver though a call to:
  192. $$register_cdrom(struct\ cdrom_device_info * <device>_info)
  193. $$
  194. The device information structure, $<device>_info$, contains all the
  195. information needed for the kernel to interface with the low-level
  196. \cdrom\ device driver. One of the most important entries in this
  197. structure is a pointer to the $cdrom_device_ops$ structure of the
  198. low-level driver.
  199. The device operations structure, $cdrom_device_ops$, contains a list
  200. of pointers to the functions which are implemented in the low-level
  201. device driver. When \cdromc\ accesses a \cdrom\ device, it does it
  202. through the functions in this structure. It is impossible to know all
  203. the capabilities of future \cdrom\ drives, so it is expected that this
  204. list may need to be expanded from time to time as new technologies are
  205. developed. For example, CD-R and CD-R/W drives are beginning to become
  206. popular, and support will soon need to be added for them. For now, the
  207. current $struct$ is:
  208. $$
  209. \halign{$#$\ \hfil&$#$\ \hfil&\hbox to 10em{$#$\hss}&
  210. $/*$ \rm# $*/$\hfil\cr
  211. struct& cdrom_device_ops\ \{ \hidewidth\cr
  212. &int& (* open)(struct\ cdrom_device_info *, int)\cr
  213. &void& (* release)(struct\ cdrom_device_info *);\cr
  214. &int& (* drive_status)(struct\ cdrom_device_info *, int);\cr
  215. &int& (* media_changed)(struct\ cdrom_device_info *, int);\cr
  216. &int& (* tray_move)(struct\ cdrom_device_info *, int);\cr
  217. &int& (* lock_door)(struct\ cdrom_device_info *, int);\cr
  218. &int& (* select_speed)(struct\ cdrom_device_info *, int);\cr
  219. &int& (* select_disc)(struct\ cdrom_device_info *, int);\cr
  220. &int& (* get_last_session) (struct\ cdrom_device_info *,
  221. struct\ cdrom_multisession *{});\cr
  222. &int& (* get_mcn)(struct\ cdrom_device_info *, struct\ cdrom_mcn *{});\cr
  223. &int& (* reset)(struct\ cdrom_device_info *);\cr
  224. &int& (* audio_ioctl)(struct\ cdrom_device_info *, unsigned\ int,
  225. void *{});\cr
  226. &int& (* dev_ioctl)(struct\ cdrom_device_info *, unsigned\ int,
  227. unsigned\ long);\cr
  228. \noalign{\medskip}
  229. &const\ int& capability;& capability flags \cr
  230. &int& n_minors;& number of active minor devices \cr
  231. \};\cr
  232. }
  233. $$
  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. \cdrom\ hardware and/or low-level \cdrom\ driver when a \cdrom\ drive
  239. is registered with the \UCD. The value $n_minors$ should be a positive
  240. value indicating the number of minor devices that are supported by
  241. the low-level device driver, normally~1. Although these two variables
  242. are `informative' rather than `operational,' they are included in
  243. $cdrom_device_ops$ because they describe the capability of the {\em
  244. driver\/} rather than the {\em drive}. Nomenclature has always been
  245. difficult in computer programming.
  246. Note that most functions have fewer parameters than their
  247. $blkdev_fops$ counterparts. This is because very little of the
  248. information in the structures $inode$ and $file$ is used. For most
  249. drivers, the main parameter is the $struct$ $cdrom_device_info$, from
  250. which the major and minor number can be extracted. (Most low-level
  251. \cdrom\ drivers don't even look at the major and minor number though,
  252. since many of them only support one device.) This will be available
  253. through $dev$ in $cdrom_device_info$ described below.
  254. The drive-specific, minor-like information that is registered with
  255. \cdromc, currently contains the following fields:
  256. $$
  257. \halign{$#$\ \hfil&$#$\ \hfil&\hbox to 10em{$#$\hss}&
  258. $/*$ \rm# $*/$\hfil\cr
  259. struct& cdrom_device_info\ \{ \hidewidth\cr
  260. & struct\ cdrom_device_ops *& ops;& device operations for this major\cr
  261. & struct\ cdrom_device_info *& next;& next device_info for this major\cr
  262. & void *& handle;& driver-dependent data\cr
  263. \noalign{\medskip}
  264. & kdev_t& dev;& device number (incorporates minor)\cr
  265. & int& mask;& mask of capability: disables them \cr
  266. & int& speed;& maximum speed for reading data \cr
  267. & int& capacity;& number of discs in a jukebox \cr
  268. \noalign{\medskip}
  269. &int& options : 30;& options flags \cr
  270. &unsigned& mc_flags : 2;& media-change buffer flags \cr
  271. & int& use_count;& number of times device is opened\cr
  272. & char& name[20];& name of the device type\cr
  273. \}\cr
  274. }$$
  275. Using this $struct$, a linked list of the registered minor devices is
  276. built, using the $next$ field. The device number, the device operations
  277. struct and specifications of properties of the drive are stored in this
  278. structure.
  279. The $mask$ flags can be used to mask out some of the capabilities listed
  280. in $ops\to capability$, if a specific drive doesn't support a feature
  281. of the driver. The value $speed$ specifies the maximum head-rate of the
  282. drive, measured in units of normal audio speed (176\,kB/sec raw data or
  283. 150\,kB/sec file system data). The value $n_discs$ should reflect the
  284. number of discs the drive can hold simultaneously, if it is designed
  285. as a juke-box, or otherwise~1. The parameters are declared $const$
  286. because they describe properties of the drive, which don't change after
  287. registration.
  288. A few registers contain variables local to the \cdrom\ drive. The
  289. flags $options$ are used to specify how the general \cdrom\ routines
  290. should behave. These various flags registers should provide enough
  291. flexibility to adapt to the different users' wishes (and {\em not\/} the
  292. `arbitrary' wishes of the author of the low-level device driver, as is
  293. the case in the old scheme). The register $mc_flags$ is used to buffer
  294. the information from $media_changed()$ to two separate queues. Other
  295. data that is specific to a minor drive, can be accessed through $handle$,
  296. which can point to a data structure specific to the low-level driver.
  297. The fields $use_count$, $next$, $options$ and $mc_flags$ need not be
  298. initialized.
  299. The intermediate software layer that \cdromc\ forms will perform some
  300. additional bookkeeping. The use count of the device (the number of
  301. processes that have the device opened) is registered in $use_count$. The
  302. function $cdrom_ioctl()$ will verify the appropriate user-memory regions
  303. for read and write, and in case a location on the CD is transferred,
  304. it will `sanitize' the format by making requests to the low-level
  305. drivers in a standard format, and translating all formats between the
  306. user-software and low level drivers. This relieves much of the drivers'
  307. memory checking and format checking and translation. Also, the necessary
  308. structures will be declared on the program stack.
  309. The implementation of the functions should be as defined in the
  310. following sections. Two functions {\em must\/} be implemented, namely
  311. $open()$ and $release()$. Other functions may be omitted, their
  312. corresponding capability flags will be cleared upon registration.
  313. Generally, a function returns zero on success and negative on error. A
  314. function call should return only after the command has completed, but of
  315. course waiting for the device should not use processor time.
  316. \subsection{$Int\ open(struct\ cdrom_device_info * cdi, int\ purpose)$}
  317. $Open()$ should try to open the device for a specific $purpose$, which
  318. can be either:
  319. \begin{itemize}
  320. \item[0] Open for reading data, as done by {\tt {mount()}} (2), or the
  321. user commands {\tt {dd}} or {\tt {cat}}.
  322. \item[1] Open for $ioctl$ commands, as done by audio-CD playing
  323. programs.
  324. \end{itemize}
  325. Notice that any strategic code (closing tray upon $open()$, etc.)\ is
  326. done by the calling routine in \cdromc, so the low-level routine
  327. should only be concerned with proper initialization, such as spinning
  328. up the disc, etc. % and device-use count
  329. \subsection{$Void\ release(struct\ cdrom_device_info * cdi)$}
  330. Device-specific actions should be taken such as spinning down the device.
  331. However, strategic actions such as ejection of the tray, or unlocking
  332. the door, should be left over to the general routine $cdrom_release()$.
  333. This is the only function returning type $void$.
  334. \subsection{$Int\ drive_status(struct\ cdrom_device_info * cdi, int\ slot_nr)$}
  335. \label{drive status}
  336. The function $drive_status$, if implemented, should provide
  337. information on the status of the drive (not the status of the disc,
  338. which may or may not be in the drive). If the drive is not a changer,
  339. $slot_nr$ should be ignored. In \cdromh\ the possibilities are listed:
  340. $$
  341. \halign{$#$\ \hfil&$/*$ \rm# $*/$\hfil\cr
  342. CDS_NO_INFO& no information available\cr
  343. CDS_NO_DISC& no disc is inserted, tray is closed\cr
  344. CDS_TRAY_OPEN& tray is opened\cr
  345. CDS_DRIVE_NOT_READY& something is wrong, tray is moving?\cr
  346. CDS_DISC_OK& a disc is loaded and everything is fine\cr
  347. }
  348. $$
  349. \subsection{$Int\ media_changed(struct\ cdrom_device_info * cdi, int\ disc_nr)$}
  350. This function is very similar to the original function in $struct\
  351. file_operations$. It returns 1 if the medium of the device $cdi\to
  352. dev$ has changed since the last call, and 0 otherwise. The parameter
  353. $disc_nr$ identifies a specific slot in a juke-box, it should be
  354. ignored for single-disc drives. Note that by `re-routing' this
  355. function through $cdrom_media_changed()$, we can implement separate
  356. queues for the VFS and a new $ioctl()$ function that can report device
  357. changes to software (\eg, an auto-mounting daemon).
  358. \subsection{$Int\ tray_move(struct\ cdrom_device_info * cdi, int\ position)$}
  359. This function, if implemented, should control the tray movement. (No
  360. other function should control this.) The parameter $position$ controls
  361. the desired direction of movement:
  362. \begin{itemize}
  363. \item[0] Close tray
  364. \item[1] Open tray
  365. \end{itemize}
  366. This function returns 0 upon success, and a non-zero value upon
  367. error. Note that if the tray is already in the desired position, no
  368. action need be taken, and the return value should be 0.
  369. \subsection{$Int\ lock_door(struct\ cdrom_device_info * cdi, int\ lock)$}
  370. This function (and no other code) controls locking of the door, if the
  371. drive allows this. The value of $lock$ controls the desired locking
  372. state:
  373. \begin{itemize}
  374. \item[0] Unlock door, manual opening is allowed
  375. \item[1] Lock door, tray cannot be ejected manually
  376. \end{itemize}
  377. This function returns 0 upon success, and a non-zero value upon
  378. error. Note that if the door is already in the requested state, no
  379. action need be taken, and the return value should be 0.
  380. \subsection{$Int\ select_speed(struct\ cdrom_device_info * cdi, int\ speed)$}
  381. Some \cdrom\ drives are capable of changing their head-speed. There
  382. are several reasons for changing the speed of a \cdrom\ drive. Badly
  383. pressed \cdrom s may benefit from less-than-maximum head rate. Modern
  384. \cdrom\ drives can obtain very high head rates (up to $24\times$ is
  385. common). It has been reported that these drives can make reading
  386. errors at these high speeds, reducing the speed can prevent data loss
  387. in these circumstances. Finally, some of these drives can
  388. make an annoyingly loud noise, which a lower speed may reduce. %Finally,
  389. %although the audio-low-pass filters probably aren't designed for it,
  390. %more than real-time playback of audio might be used for high-speed
  391. %copying of audio tracks.
  392. This function specifies the speed at which data is read or audio is
  393. played back. The value of $speed$ specifies the head-speed of the
  394. drive, measured in units of standard cdrom speed (176\,kB/sec raw data
  395. or 150\,kB/sec file system data). So to request that a \cdrom\ drive
  396. operate at 300\,kB/sec you would call the CDROM_SELECT_SPEED $ioctl$
  397. with $speed=2$. The special value `0' means `auto-selection', \ie,
  398. maximum data-rate or real-time audio rate. If the drive doesn't have
  399. this `auto-selection' capability, the decision should be made on the
  400. current disc loaded and the return value should be positive. A negative
  401. return value indicates an error.
  402. \subsection{$Int\ select_disc(struct\ cdrom_device_info * cdi, int\ number)$}
  403. If the drive can store multiple discs (a juke-box) this function
  404. will perform disc selection. It should return the number of the
  405. selected disc on success, a negative value on error. Currently, only
  406. the ide-cd driver supports this functionality.
  407. \subsection{$Int\ get_last_session(struct\ cdrom_device_info * cdi, struct\
  408. cdrom_multisession * ms_info)$}
  409. This function should implement the old corresponding $ioctl()$. For
  410. device $cdi\to dev$, the start of the last session of the current disc
  411. should be returned in the pointer argument $ms_info$. Note that
  412. routines in \cdromc\ have sanitized this argument: its requested
  413. format will {\em always\/} be of the type $CDROM_LBA$ (linear block
  414. addressing mode), whatever the calling software requested. But
  415. sanitization goes even further: the low-level implementation may
  416. return the requested information in $CDROM_MSF$ format if it wishes so
  417. (setting the $ms_info\rightarrow addr_format$ field appropriately, of
  418. course) and the routines in \cdromc\ will make the transformation if
  419. necessary. The return value is 0 upon success.
  420. \subsection{$Int\ get_mcn(struct\ cdrom_device_info * cdi, struct\
  421. cdrom_mcn * mcn)$}
  422. Some discs carry a `Media Catalog Number' (MCN), also called
  423. `Universal Product Code' (UPC). This number should reflect the number
  424. that is generally found in the bar-code on the product. Unfortunately,
  425. the few discs that carry such a number on the disc don't even use the
  426. same format. The return argument to this function is a pointer to a
  427. pre-declared memory region of type $struct\ cdrom_mcn$. The MCN is
  428. expected as a 13-character string, terminated by a null-character.
  429. \subsection{$Int\ reset(struct\ cdrom_device_info * cdi)$}
  430. This call should perform a hard-reset on the drive (although in
  431. circumstances that a hard-reset is necessary, a drive may very well not
  432. listen to commands anymore). Preferably, control is returned to the
  433. caller only after the drive has finished resetting. If the drive is no
  434. longer listening, it may be wise for the underlying low-level cdrom
  435. driver to time out.
  436. \subsection{$Int\ audio_ioctl(struct\ cdrom_device_info * cdi, unsigned\
  437. int\ cmd, void * arg)$}
  438. Some of the \cdrom-$ioctl$s defined in \cdromh\ can be
  439. implemented by the routines described above, and hence the function
  440. $cdrom_ioctl$ will use those. However, most $ioctl$s deal with
  441. audio-control. We have decided to leave these to be accessed through a
  442. single function, repeating the arguments $cmd$ and $arg$. Note that
  443. the latter is of type $void*{}$, rather than $unsigned\ long\
  444. int$. The routine $cdrom_ioctl()$ does do some useful things,
  445. though. It sanitizes the address format type to $CDROM_MSF$ (Minutes,
  446. Seconds, Frames) for all audio calls. It also verifies the memory
  447. location of $arg$, and reserves stack-memory for the argument. This
  448. makes implementation of the $audio_ioctl()$ much simpler than in the
  449. old driver scheme. For example, you may look up the function
  450. $cm206_audio_ioctl()$ in {\tt {cm206.c}} that should be updated with
  451. this documentation.
  452. An unimplemented ioctl should return $-ENOSYS$, but a harmless request
  453. (\eg, $CDROMSTART$) may be ignored by returning 0 (success). Other
  454. errors should be according to the standards, whatever they are. When
  455. an error is returned by the low-level driver, the \UCD\ tries whenever
  456. possible to return the error code to the calling program. (We may decide
  457. to sanitize the return value in $cdrom_ioctl()$ though, in order to
  458. guarantee a uniform interface to the audio-player software.)
  459. \subsection{$Int\ dev_ioctl(struct\ cdrom_device_info * cdi, unsigned\ int\
  460. cmd, unsigned\ long\ arg)$}
  461. Some $ioctl$s seem to be specific to certain \cdrom\ drives. That is,
  462. they are introduced to service some capabilities of certain drives. In
  463. fact, there are 6 different $ioctl$s for reading data, either in some
  464. particular kind of format, or audio data. Not many drives support
  465. reading audio tracks as data, I believe this is because of protection
  466. of copyrights of artists. Moreover, I think that if audio-tracks are
  467. supported, it should be done through the VFS and not via $ioctl$s. A
  468. problem here could be the fact that audio-frames are 2352 bytes long,
  469. so either the audio-file-system should ask for 75264 bytes at once
  470. (the least common multiple of 512 and 2352), or the drivers should
  471. bend their backs to cope with this incoherence (to which I would be
  472. opposed). Furthermore, it is very difficult for the hardware to find
  473. the exact frame boundaries, since there are no synchronization headers
  474. in audio frames. Once these issues are resolved, this code should be
  475. standardized in \cdromc.
  476. Because there are so many $ioctl$s that seem to be introduced to
  477. satisfy certain drivers,\footnote{Is there software around that
  478. actually uses these? I'd be interested!} any `non-standard' $ioctl$s
  479. are routed through the call $dev_ioctl()$. In principle, `private'
  480. $ioctl$s should be numbered after the device's major number, and not
  481. the general \cdrom\ $ioctl$ number, {\tt {0x53}}. Currently the
  482. non-supported $ioctl$s are: {\it CDROMREADMODE1, CDROMREADMODE2,
  485. \subsection{\cdrom\ capabilities}
  486. \label{capability}
  487. Instead of just implementing some $ioctl$ calls, the interface in
  488. \cdromc\ supplies the possibility to indicate the {\em capabilities\/}
  489. of a \cdrom\ drive. This can be done by ORing any number of
  490. capability-constants that are defined in \cdromh\ at the registration
  491. phase. Currently, the capabilities are any of:
  492. $$
  493. \halign{$#$\ \hfil&$/*$ \rm# $*/$\hfil\cr
  494. CDC_CLOSE_TRAY& can close tray by software control\cr
  495. CDC_OPEN_TRAY& can open tray\cr
  496. CDC_LOCK& can lock and unlock the door\cr
  497. CDC_SELECT_SPEED& can select speed, in units of $\sim$150\,kB/s\cr
  498. CDC_SELECT_DISC& drive is juke-box\cr
  499. CDC_MULTI_SESSION& can read sessions $>\rm1$\cr
  500. CDC_MCN& can read Media Catalog Number\cr
  501. CDC_MEDIA_CHANGED& can report if disc has changed\cr
  502. CDC_PLAY_AUDIO& can perform audio-functions (play, pause, etc)\cr
  503. CDC_RESET& hard reset device\cr
  504. CDC_IOCTLS& driver has non-standard ioctls\cr
  505. CDC_DRIVE_STATUS& driver implements drive status\cr
  506. }
  507. $$
  508. The capability flag is declared $const$, to prevent drivers from
  509. accidentally tampering with the contents. The capability fags actually
  510. inform \cdromc\ of what the driver can do. If the drive found
  511. by the driver does not have the capability, is can be masked out by
  512. the $cdrom_device_info$ variable $mask$. For instance, the SCSI \cdrom\
  513. driver has implemented the code for loading and ejecting \cdrom's, and
  514. hence its corresponding flags in $capability$ will be set. But a SCSI
  515. \cdrom\ drive might be a caddy system, which can't load the tray, and
  516. hence for this drive the $cdrom_device_info$ struct will have set
  517. the $CDC_CLOSE_TRAY$ bit in $mask$.
  518. In the file \cdromc\ you will encounter many constructions of the type
  519. $$\it
  520. if\ (cdo\rightarrow capability \mathrel\& \mathord{\sim} cdi\rightarrow mask
  521. \mathrel{\&} CDC_<capability>) \ldots
  522. $$
  523. There is no $ioctl$ to set the mask\dots The reason is that
  524. I think it is better to control the {\em behavior\/} rather than the
  525. {\em capabilities}.
  526. \subsection{Options}
  527. A final flag register controls the {\em behavior\/} of the \cdrom\
  528. drives, in order to satisfy different users' wishes, hopefully
  529. independently of the ideas of the respective author who happened to
  530. have made the drive's support available to the \linux\ community. The
  531. current behavior options are:
  532. $$
  533. \halign{$#$\ \hfil&$/*$ \rm# $*/$\hfil\cr
  534. CDO_AUTO_CLOSE& try to close tray upon device $open()$\cr
  535. CDO_AUTO_EJECT& try to open tray on last device $close()$\cr
  536. CDO_USE_FFLAGS& use $file_pointer\rightarrow f_flags$ to indicate
  537. purpose for $open()$\cr
  538. CDO_LOCK& try to lock door if device is opened\cr
  539. CDO_CHECK_TYPE& ensure disc type is data if opened for data\cr
  540. }
  541. $$
  542. The initial value of this register is $CDO_AUTO_CLOSE \mathrel|
  543. CDO_USE_FFLAGS \mathrel| CDO_LOCK$, reflecting my own view on user
  544. interface and software standards. Before you protest, there are two
  545. new $ioctl$s implemented in \cdromc, that allow you to control the
  546. behavior by software. These are:
  547. $$
  548. \halign{$#$\ \hfil&$/*$ \rm# $*/$\hfil\cr
  549. CDROM_SET_OPTIONS& set options specified in $(int)\ arg$\cr
  550. CDROM_CLEAR_OPTIONS& clear options specified in $(int)\ arg$\cr
  551. }
  552. $$
  553. One option needs some more explanation: $CDO_USE_FFLAGS$. In the next
  554. newsection we explain what the need for this option is.
  555. A software package {\tt setcd}, available from the Debian distribution
  556. and {\tt}, allows user level control of these flags.
  557. \newsection{The need to know the purpose of opening the \cdrom\ device}
  558. Traditionally, Unix devices can be used in two different `modes',
  559. either by reading/writing to the device file, or by issuing
  560. controlling commands to the device, by the device's $ioctl()$
  561. call. The problem with \cdrom\ drives, is that they can be used for
  562. two entirely different purposes. One is to mount removable
  563. file systems, \cdrom s, the other is to play audio CD's. Audio commands
  564. are implemented entirely through $ioctl$s, presumably because the
  565. first implementation (SUN?) has been such. In principle there is
  566. nothing wrong with this, but a good control of the `CD player' demands
  567. that the device can {\em always\/} be opened in order to give the
  568. $ioctl$ commands, regardless of the state the drive is in.
  569. On the other hand, when used as a removable-media disc drive (what the
  570. original purpose of \cdrom s is) we would like to make sure that the
  571. disc drive is ready for operation upon opening the device. In the old
  572. scheme, some \cdrom\ drivers don't do any integrity checking, resulting
  573. in a number of i/o errors reported by the VFS to the kernel when an
  574. attempt for mounting a \cdrom\ on an empty drive occurs. This is not a
  575. particularly elegant way to find out that there is no \cdrom\ inserted;
  576. it more-or-less looks like the old IBM-PC trying to read an empty floppy
  577. drive for a couple of seconds, after which the system complains it
  578. can't read from it. Nowadays we can {\em sense\/} the existence of a
  579. removable medium in a drive, and we believe we should exploit that
  580. fact. An integrity check on opening of the device, that verifies the
  581. availability of a \cdrom\ and its correct type (data), would be
  582. desirable.
  583. These two ways of using a \cdrom\ drive, principally for data and
  584. secondarily for playing audio discs, have different demands for the
  585. behavior of the $open()$ call. Audio use simply wants to open the
  586. device in order to get a file handle which is needed for issuing
  587. $ioctl$ commands, while data use wants to open for correct and
  588. reliable data transfer. The only way user programs can indicate what
  589. their {\em purpose\/} of opening the device is, is through the $flags$
  590. parameter (see {\tt {open(2)}}). For \cdrom\ devices, these flags aren't
  591. implemented (some drivers implement checking for write-related flags,
  592. but this is not strictly necessary if the device file has correct
  593. permission flags). Most option flags simply don't make sense to
  594. \cdrom\ devices: $O_CREAT$, $O_NOCTTY$, $O_TRUNC$, $O_APPEND$, and
  595. $O_SYNC$ have no meaning to a \cdrom.
  596. We therefore propose to use the flag $O_NONBLOCK$ to indicate
  597. that the device is opened just for issuing $ioctl$
  598. commands. Strictly, the meaning of $O_NONBLOCK$ is that opening and
  599. subsequent calls to the device don't cause the calling process to
  600. wait. We could interpret this as ``don't wait until someone has
  601. inserted some valid data-\cdrom.'' Thus, our proposal of the
  602. implementation for the $open()$ call for \cdrom s is:
  603. \begin{itemize}
  604. \item If no other flags are set than $O_RDONLY$, the device is opened
  605. for data transfer, and the return value will be 0 only upon successful
  606. initialization of the transfer. The call may even induce some actions
  607. on the \cdrom, such as closing the tray.
  608. \item If the option flag $O_NONBLOCK$ is set, opening will always be
  609. successful, unless the whole device doesn't exist. The drive will take
  610. no actions whatsoever.
  611. \end{itemize}
  612. \subsection{And what about standards?}
  613. You might hesitate to accept this proposal as it comes from the
  614. \linux\ community, and not from some standardizing institute. What
  615. about SUN, SGI, HP and all those other Unix and hardware vendors?
  616. Well, these companies are in the lucky position that they generally
  617. control both the hardware and software of their supported products,
  618. and are large enough to set their own standard. They do not have to
  619. deal with a dozen or more different, competing hardware
  620. configurations.\footnote{Incidentally, I think that SUN's approach to
  621. mounting \cdrom s is very good in origin: under Solaris a
  622. volume-daemon automatically mounts a newly inserted \cdrom\ under {\tt
  623. {/cdrom/$<volume-name>$/}}. In my opinion they should have pushed this
  624. further and have {\em every\/} \cdrom\ on the local area network be
  625. mounted at the similar location, \ie, no matter in which particular
  626. machine you insert a \cdrom, it will always appear at the same
  627. position in the directory tree, on every system. When I wanted to
  628. implement such a user-program for \linux, I came across the
  629. differences in behavior of the various drivers, and the need for an
  630. $ioctl$ informing about media changes.}
  631. We believe that using $O_NONBLOCK$ to indicate that a device is being opened
  632. for $ioctl$ commands only can be easily introduced in the \linux\
  633. community. All the CD-player authors will have to be informed, we can
  634. even send in our own patches to the programs. The use of $O_NONBLOCK$
  635. has most likely no influence on the behavior of the CD-players on
  636. other operating systems than \linux. Finally, a user can always revert
  637. to old behavior by a call to $ioctl(file_descriptor, CDROM_CLEAR_OPTIONS,
  638. CDO_USE_FFLAGS)$.
  639. \subsection{The preferred strategy of $open()$}
  640. The routines in \cdromc\ are designed in such a way that run-time
  641. configuration of the behavior of \cdrom\ devices (of {\em any\/} type)
  642. can be carried out, by the $CDROM_SET/CLEAR_OPTIONS$ $ioctls$. Thus, various
  643. modes of operation can be set:
  644. \begin{description}
  645. \item[$CDO_AUTO_CLOSE \mathrel| CDO_USE_FFLAGS \mathrel| CDO_LOCK$] This
  646. is the default setting. (With $CDO_CHECK_TYPE$ it will be better, in the
  647. future.) If the device is not yet opened by any other process, and if
  648. the device is being opened for data ($O_NONBLOCK$ is not set) and the
  649. tray is found to be open, an attempt to close the tray is made. Then,
  650. it is verified that a disc is in the drive and, if $CDO_CHECK_TYPE$ is
  651. set, that it contains tracks of type `data mode 1.' Only if all tests
  652. are passed is the return value zero. The door is locked to prevent file
  653. system corruption. If the drive is opened for audio ($O_NONBLOCK$ is
  654. set), no actions are taken and a value of 0 will be returned.
  655. \item[$CDO_AUTO_CLOSE \mathrel| CDO_AUTO_EJECT \mathrel| CDO_LOCK$] This
  656. mimics the behavior of the current sbpcd-driver. The option flags are
  657. ignored, the tray is closed on the first open, if necessary. Similarly,
  658. the tray is opened on the last release, \ie, if a \cdrom\ is unmounted,
  659. it is automatically ejected, such that the user can replace it.
  660. \end{description}
  661. We hope that these option can convince everybody (both driver
  662. maintainers and user program developers) to adopt the new \cdrom\
  663. driver scheme and option flag interpretation.
  664. \newsection{Description of routines in \cdromc}
  665. Only a few routines in \cdromc\ are exported to the drivers. In this
  666. new section we will discuss these, as well as the functions that `take
  667. over' the \cdrom\ interface to the kernel. The header file belonging
  668. to \cdromc\ is called \cdromh. Formerly, some of the contents of this
  669. file were placed in the file {\tt {ucdrom.h}}, but this file has now been
  670. merged back into \cdromh.
  671. \subsection{$Struct\ file_operations\ cdrom_fops$}
  672. The contents of this structure were described in section~\ref{cdrom.c}.
  673. A pointer to this structure is assigned to the $fops$ field
  674. of the $struct gendisk$.
  675. \subsection{$Int\ register_cdrom( struct\ cdrom_device_info\ * cdi)$}
  676. This function is used in about the same way one registers $cdrom_fops$
  677. with the kernel, the device operations and information structures,
  678. as described in section~\ref{cdrom.c}, should be registered with the
  679. \UCD:
  680. $$
  681. register_cdrom(\&<device>_info));
  682. $$
  683. This function returns zero upon success, and non-zero upon
  684. failure. The structure $<device>_info$ should have a pointer to the
  685. driver's $<device>_dops$, as in
  686. $$
  687. \vbox{\halign{&$#$\hfil\cr
  688. struct\ &cdrom_device_info\ <device>_info = \{\cr
  689. & <device>_dops;\cr
  690. &\ldots\cr
  691. \}\cr
  692. }}$$
  693. Note that a driver must have one static structure, $<device>_dops$, while
  694. it may have as many structures $<device>_info$ as there are minor devices
  695. active. $Register_cdrom()$ builds a linked list from these.
  696. \subsection{$Void\ unregister_cdrom(struct\ cdrom_device_info * cdi)$}
  697. Unregistering device $cdi$ with minor number $MINOR(cdi\to dev)$ removes
  698. the minor device from the list. If it was the last registered minor for
  699. the low-level driver, this disconnects the registered device-operation
  700. routines from the \cdrom\ interface. This function returns zero upon
  701. success, and non-zero upon failure.
  702. \subsection{$Int\ cdrom_open(struct\ inode * ip, struct\ file * fp)$}
  703. This function is not called directly by the low-level drivers, it is
  704. listed in the standard $cdrom_fops$. If the VFS opens a file, this
  705. function becomes active. A strategy is implemented in this routine,
  706. taking care of all capabilities and options that are set in the
  707. $cdrom_device_ops$ connected to the device. Then, the program flow is
  708. transferred to the device_dependent $open()$ call.
  709. \subsection{$Void\ cdrom_release(struct\ inode *ip, struct\ file
  710. *fp)$}
  711. This function implements the reverse-logic of $cdrom_open()$, and then
  712. calls the device-dependent $release()$ routine. When the use-count has
  713. reached 0, the allocated buffers are flushed by calls to $sync_dev(dev)$
  714. and $invalidate_buffers(dev)$.
  715. \subsection{$Int\ cdrom_ioctl(struct\ inode *ip, struct\ file *fp,
  716. unsigned\ int\ cmd, unsigned\ long\ arg)$}
  717. \label{cdrom-ioctl}
  718. This function handles all the standard $ioctl$ requests for \cdrom\
  719. devices in a uniform way. The different calls fall into three
  720. categories: $ioctl$s that can be directly implemented by device
  721. operations, ones that are routed through the call $audio_ioctl()$, and
  722. the remaining ones, that are presumable device-dependent. Generally, a
  723. negative return value indicates an error.
  724. \subsubsection{Directly implemented $ioctl$s}
  725. \label{ioctl-direct}
  726. The following `old' \cdrom-$ioctl$s are implemented by directly
  727. calling device-operations in $cdrom_device_ops$, if implemented and
  728. not masked:
  729. \begin{description}
  730. \item[CDROMMULTISESSION] Requests the last session on a \cdrom.
  731. \item[CDROMEJECT] Open tray.
  732. \item[CDROMCLOSETRAY] Close tray.
  733. \item[CDROMEJECT_SW] If $arg\not=0$, set behavior to auto-close (close
  734. tray on first open) and auto-eject (eject on last release), otherwise
  735. set behavior to non-moving on $open()$ and $release()$ calls.
  736. \item[CDROM_GET_MCN] Get the Media Catalog Number from a CD.
  737. \end{description}
  738. \subsubsection{$Ioctl$s routed through $audio_ioctl()$}
  739. \label{ioctl-audio}
  740. The following set of $ioctl$s are all implemented through a call to
  741. the $cdrom_fops$ function $audio_ioctl()$. Memory checks and
  742. allocation are performed in $cdrom_ioctl()$, and also sanitization of
  743. address format ($CDROM_LBA$/$CDROM_MSF$) is done.
  744. \begin{description}
  745. \item[CDROMSUBCHNL] Get sub-channel data in argument $arg$ of type $struct\
  746. cdrom_subchnl *{}$.
  747. \item[CDROMREADTOCHDR] Read Table of Contents header, in $arg$ of type
  748. $struct\ cdrom_tochdr *{}$.
  749. \item[CDROMREADTOCENTRY] Read a Table of Contents entry in $arg$ and
  750. specified by $arg$ of type $struct\ cdrom_tocentry *{}$.
  751. \item[CDROMPLAYMSF] Play audio fragment specified in Minute, Second,
  752. Frame format, delimited by $arg$ of type $struct\ cdrom_msf *{}$.
  753. \item[CDROMPLAYTRKIND] Play audio fragment in track-index format
  754. delimited by $arg$ of type $struct\ \penalty-1000 cdrom_ti *{}$.
  755. \item[CDROMVOLCTRL] Set volume specified by $arg$ of type $struct\
  756. cdrom_volctrl *{}$.
  757. \item[CDROMVOLREAD] Read volume into by $arg$ of type $struct\
  758. cdrom_volctrl *{}$.
  759. \item[CDROMSTART] Spin up disc.
  760. \item[CDROMSTOP] Stop playback of audio fragment.
  761. \item[CDROMPAUSE] Pause playback of audio fragment.
  762. \item[CDROMRESUME] Resume playing.
  763. \end{description}
  764. \subsubsection{New $ioctl$s in \cdromc}
  765. The following $ioctl$s have been introduced to allow user programs to
  766. control the behavior of individual \cdrom\ devices. New $ioctl$
  767. commands can be identified by the underscores in their names.
  768. \begin{description}
  769. \item[CDROM_SET_OPTIONS] Set options specified by $arg$. Returns the
  770. option flag register after modification. Use $arg = \rm0$ for reading
  771. the current flags.
  772. \item[CDROM_CLEAR_OPTIONS] Clear options specified by $arg$. Returns
  773. the option flag register after modification.
  774. \item[CDROM_SELECT_SPEED] Select head-rate speed of disc specified as
  775. by $arg$ in units of standard cdrom speed (176\,kB/sec raw data or
  776. 150\,kB/sec file system data). The value 0 means `auto-select', \ie,
  777. play audio discs at real time and data discs at maximum speed. The value
  778. $arg$ is checked against the maximum head rate of the drive found in the
  779. $cdrom_dops$.
  780. \item[CDROM_SELECT_DISC] Select disc numbered $arg$ from a juke-box.
  781. First disc is numbered 0. The number $arg$ is checked against the
  782. maximum number of discs in the juke-box found in the $cdrom_dops$.
  783. \item[CDROM_MEDIA_CHANGED] Returns 1 if a disc has been changed since
  784. the last call. Note that calls to $cdrom_media_changed$ by the VFS
  785. are treated by an independent queue, so both mechanisms will detect
  786. a media change once. For juke-boxes, an extra argument $arg$
  787. specifies the slot for which the information is given. The special
  788. value $CDSL_CURRENT$ requests that information about the currently
  789. selected slot be returned.
  790. \item[CDROM_DRIVE_STATUS] Returns the status of the drive by a call to
  791. $drive_status()$. Return values are defined in section~\ref{drive
  792. status}. Note that this call doesn't return information on the
  793. current playing activity of the drive; this can be polled through an
  794. $ioctl$ call to $CDROMSUBCHNL$. For juke-boxes, an extra argument
  795. $arg$ specifies the slot for which (possibly limited) information is
  796. given. The special value $CDSL_CURRENT$ requests that information
  797. about the currently selected slot be returned.
  798. \item[CDROM_DISC_STATUS] Returns the type of the disc currently in the
  799. drive. It should be viewed as a complement to $CDROM_DRIVE_STATUS$.
  800. This $ioctl$ can provide \emph {some} information about the current
  801. disc that is inserted in the drive. This functionality used to be
  802. implemented in the low level drivers, but is now carried out
  803. entirely in \UCD.
  804. The history of development of the CD's use as a carrier medium for
  805. various digital information has lead to many different disc types.
  806. This $ioctl$ is useful only in the case that CDs have \emph {only
  807. one} type of data on them. While this is often the case, it is
  808. also very common for CDs to have some tracks with data, and some
  809. tracks with audio. Because this is an existing interface, rather
  810. than fixing this interface by changing the assumptions it was made
  811. under, thereby breaking all user applications that use this
  812. function, the \UCD\ implements this $ioctl$ as follows: If the CD in
  813. question has audio tracks on it, and it has absolutely no CD-I, XA,
  814. or data tracks on it, it will be reported as $CDS_AUDIO$. If it has
  815. both audio and data tracks, it will return $CDS_MIXED$. If there
  816. are no audio tracks on the disc, and if the CD in question has any
  817. CD-I tracks on it, it will be reported as $CDS_XA_2_2$. Failing
  818. that, if the CD in question has any XA tracks on it, it will be
  819. reported as $CDS_XA_2_1$. Finally, if the CD in question has any
  820. data tracks on it, it will be reported as a data CD ($CDS_DATA_1$).
  821. This $ioctl$ can return:
  822. $$
  823. \halign{$#$\ \hfil&$/*$ \rm# $*/$\hfil\cr
  824. CDS_NO_INFO& no information available\cr
  825. CDS_NO_DISC& no disc is inserted, or tray is opened\cr
  826. CDS_AUDIO& Audio disc (2352 audio bytes/frame)\cr
  827. CDS_DATA_1& data disc, mode 1 (2048 user bytes/frame)\cr
  828. CDS_XA_2_1& mixed data (XA), mode 2, form 1 (2048 user bytes)\cr
  829. CDS_XA_2_2& mixed data (XA), mode 2, form 1 (2324 user bytes)\cr
  830. CDS_MIXED& mixed audio/data disc\cr
  831. }
  832. $$
  833. For some information concerning frame layout of the various disc
  834. types, see a recent version of \cdromh.
  835. \item[CDROM_CHANGER_NSLOTS] Returns the number of slots in a
  836. juke-box.
  837. \item[CDROMRESET] Reset the drive.
  838. \item[CDROM_GET_CAPABILITY] Returns the $capability$ flags for the
  839. drive. Refer to section \ref{capability} for more information on
  840. these flags.
  841. \item[CDROM_LOCKDOOR] Locks the door of the drive. $arg == \rm0$
  842. unlocks the door, any other value locks it.
  843. \item[CDROM_DEBUG] Turns on debugging info. Only root is allowed
  844. to do this. Same semantics as CDROM_LOCKDOOR.
  845. \end{description}
  846. \subsubsection{Device dependent $ioctl$s}
  847. Finally, all other $ioctl$s are passed to the function $dev_ioctl()$,
  848. if implemented. No memory allocation or verification is carried out.
  849. \newsection{How to update your driver}
  850. \begin{enumerate}
  851. \item Make a backup of your current driver.
  852. \item Get hold of the files \cdromc\ and \cdromh, they should be in
  853. the directory tree that came with this documentation.
  854. \item Make sure you include \cdromh.
  855. \item Change the 3rd argument of $register_blkdev$ from
  856. $\&<your-drive>_fops$ to $\&cdrom_fops$.
  857. \item Just after that line, add the following to register with the \UCD:
  858. $$register_cdrom(\&<your-drive>_info);$$
  859. Similarly, add a call to $unregister_cdrom()$ at the appropriate place.
  860. \item Copy an example of the device-operations $struct$ to your
  861. source, \eg, from {\tt {cm206.c}} $cm206_dops$, and change all
  862. entries to names corresponding to your driver, or names you just
  863. happen to like. If your driver doesn't support a certain function,
  864. make the entry $NULL$. At the entry $capability$ you should list all
  865. capabilities your driver currently supports. If your driver
  866. has a capability that is not listed, please send me a message.
  867. \item Copy the $cdrom_device_info$ declaration from the same example
  868. driver, and modify the entries according to your needs. If your
  869. driver dynamically determines the capabilities of the hardware, this
  870. structure should also be declared dynamically.
  871. \item Implement all functions in your $<device>_dops$ structure,
  872. according to prototypes listed in \cdromh, and specifications given
  873. in section~\ref{cdrom.c}. Most likely you have already implemented
  874. the code in a large part, and you will almost certainly need to adapt the
  875. prototype and return values.
  876. \item Rename your $<device>_ioctl()$ function to $audio_ioctl$ and
  877. change the prototype a little. Remove entries listed in the first
  878. part in section~\ref{cdrom-ioctl}, if your code was OK, these are
  879. just calls to the routines you adapted in the previous step.
  880. \item You may remove all remaining memory checking code in the
  881. $audio_ioctl()$ function that deals with audio commands (these are
  882. listed in the second part of section~\ref{cdrom-ioctl}). There is no
  883. need for memory allocation either, so most $case$s in the $switch$
  884. statement look similar to:
  885. $$
  886. case\ CDROMREADTOCENTRY\colon get_toc_entry\bigl((struct\
  887. cdrom_tocentry *{})\ arg\bigr);
  888. $$
  889. \item All remaining $ioctl$ cases must be moved to a separate
  890. function, $<device>_ioctl$, the device-dependent $ioctl$s. Note that
  891. memory checking and allocation must be kept in this code!
  892. \item Change the prototypes of $<device>_open()$ and
  893. $<device>_release()$, and remove any strategic code (\ie, tray
  894. movement, door locking, etc.).
  895. \item Try to recompile the drivers. We advise you to use modules, both
  896. for {\tt {cdrom.o}} and your driver, as debugging is much easier this
  897. way.
  898. \end{enumerate}
  899. \newsection{Thanks}
  900. Thanks to all the people involved. First, Erik Andersen, who has
  901. taken over the torch in maintaining \cdromc\ and integrating much
  902. \cdrom-related code in the 2.1-kernel. Thanks to Scott Snyder and
  903. Gerd Knorr, who were the first to implement this interface for SCSI
  904. and IDE-CD drivers and added many ideas for extension of the data
  905. structures relative to kernel~2.0. Further thanks to Heiko Ei{\sz}feldt,
  906. Thomas Quinot, Jon Tombs, Ken Pizzini, Eberhard M\"onkeberg and Andrew
  907. Kroll, the \linux\ \cdrom\ device driver developers who were kind
  908. enough to give suggestions and criticisms during the writing. Finally
  909. of course, I want to thank Linus Torvalds for making this possible in
  910. the first place.
  911. \vfill
  912. $ \version\ $
  913. \eject
  914. \end{document}