feature-removal-schedule.txt 12 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315
  1. The following is a list of files and features that are going to be
  2. removed in the kernel source tree. Every entry should contain what
  3. exactly is going away, why it is happening, and who is going to be doing
  4. the work. When the feature is removed from the kernel, it should also
  5. be removed from this file.
  6. ---------------------------
  7. What: /sys/devices/.../power/state
  8. dev->power.power_state
  9. dpm_runtime_{suspend,resume)()
  10. When: July 2007
  11. Why: Broken design for runtime control over driver power states, confusing
  12. driver-internal runtime power management with: mechanisms to support
  13. system-wide sleep state transitions; event codes that distinguish
  14. different phases of swsusp "sleep" transitions; and userspace policy
  15. inputs. This framework was never widely used, and most attempts to
  16. use it were broken. Drivers should instead be exposing domain-specific
  17. interfaces either to kernel or to userspace.
  18. Who: Pavel Machek <pavel@suse.cz>
  19. ---------------------------
  20. What: RAW driver (CONFIG_RAW_DRIVER)
  21. When: December 2005
  22. Why: declared obsolete since kernel 2.6.3
  23. O_DIRECT can be used instead
  24. Who: Adrian Bunk <bunk@stusta.de>
  25. ---------------------------
  26. What: raw1394: requests of type RAW1394_REQ_ISO_SEND, RAW1394_REQ_ISO_LISTEN
  27. When: June 2007
  28. Why: Deprecated in favour of the more efficient and robust rawiso interface.
  29. Affected are applications which use the deprecated part of libraw1394
  30. (raw1394_iso_write, raw1394_start_iso_write, raw1394_start_iso_rcv,
  31. raw1394_stop_iso_rcv) or bypass libraw1394.
  32. Who: Dan Dennedy <dan@dennedy.org>, Stefan Richter <stefanr@s5r6.in-berlin.de>
  33. ---------------------------
  34. What: Video4Linux API 1 ioctls and video_decoder.h from Video devices.
  35. When: December 2006
  36. Why: V4L1 AP1 was replaced by V4L2 API. during migration from 2.4 to 2.6
  37. series. The old API have lots of drawbacks and don't provide enough
  38. means to work with all video and audio standards. The newer API is
  39. already available on the main drivers and should be used instead.
  40. Newer drivers should use v4l_compat_translate_ioctl function to handle
  41. old calls, replacing to newer ones.
  42. Decoder iocts are using internally to allow video drivers to
  43. communicate with video decoders. This should also be improved to allow
  44. V4L2 calls being translated into compatible internal ioctls.
  45. Who: Mauro Carvalho Chehab <mchehab@brturbo.com.br>
  46. ---------------------------
  47. What: PCMCIA control ioctl (needed for pcmcia-cs [cardmgr, cardctl])
  48. When: November 2005
  49. Files: drivers/pcmcia/: pcmcia_ioctl.c
  50. Why: With the 16-bit PCMCIA subsystem now behaving (almost) like a
  51. normal hotpluggable bus, and with it using the default kernel
  52. infrastructure (hotplug, driver core, sysfs) keeping the PCMCIA
  53. control ioctl needed by cardmgr and cardctl from pcmcia-cs is
  54. unnecessary, and makes further cleanups and integration of the
  55. PCMCIA subsystem into the Linux kernel device driver model more
  56. difficult. The features provided by cardmgr and cardctl are either
  57. handled by the kernel itself now or are available in the new
  58. pcmciautils package available at
  59. http://kernel.org/pub/linux/utils/kernel/pcmcia/
  60. Who: Dominik Brodowski <linux@brodo.de>
  61. ---------------------------
  62. What: remove EXPORT_SYMBOL(kernel_thread)
  63. When: August 2006
  64. Files: arch/*/kernel/*_ksyms.c
  65. Why: kernel_thread is a low-level implementation detail. Drivers should
  66. use the <linux/kthread.h> API instead which shields them from
  67. implementation details and provides a higherlevel interface that
  68. prevents bugs and code duplication
  69. Who: Christoph Hellwig <hch@lst.de>
  70. ---------------------------
  71. What: CONFIG_FORCED_INLINING
  72. When: June 2006
  73. Why: Config option is there to see if gcc is good enough. (in january
  74. 2006). If it is, the behavior should just be the default. If it's not,
  75. the option should just go away entirely.
  76. Who: Arjan van de Ven
  77. ---------------------------
  78. What: eepro100 network driver
  79. When: January 2007
  80. Why: replaced by the e100 driver
  81. Who: Adrian Bunk <bunk@stusta.de>
  82. ---------------------------
  83. What: drivers depending on OSS_OBSOLETE_DRIVER
  84. When: options in 2.6.20, code in 2.6.22
  85. Why: OSS drivers with ALSA replacements
  86. Who: Adrian Bunk <bunk@stusta.de>
  87. ---------------------------
  88. What: pci_module_init(driver)
  89. When: January 2007
  90. Why: Is replaced by pci_register_driver(pci_driver).
  91. Who: Richard Knutsson <ricknu-0@student.ltu.se> and Greg Kroah-Hartman <gregkh@suse.de>
  92. ---------------------------
  93. What: Usage of invalid timevals in setitimer
  94. When: March 2007
  95. Why: POSIX requires to validate timevals in the setitimer call. This
  96. was never done by Linux. The invalid (e.g. negative timevals) were
  97. silently converted to more or less random timeouts and intervals.
  98. Until the removal a per boot limited number of warnings is printed
  99. and the timevals are sanitized.
  100. Who: Thomas Gleixner <tglx@linutronix.de>
  101. ---------------------------
  102. What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
  103. (temporary transition config option provided until then)
  104. The transition config option will also be removed at the same time.
  105. When: before 2.6.19
  106. Why: Unused symbols are both increasing the size of the kernel binary
  107. and are often a sign of "wrong API"
  108. Who: Arjan van de Ven <arjan@linux.intel.com>
  109. ---------------------------
  110. What: mount/umount uevents
  111. When: February 2007
  112. Why: These events are not correct, and do not properly let userspace know
  113. when a file system has been mounted or unmounted. Userspace should
  114. poll the /proc/mounts file instead to detect this properly.
  115. Who: Greg Kroah-Hartman <gregkh@suse.de>
  116. ---------------------------
  117. What: USB driver API moves to EXPORT_SYMBOL_GPL
  118. When: February 2008
  119. Files: include/linux/usb.h, drivers/usb/core/driver.c
  120. Why: The USB subsystem has changed a lot over time, and it has been
  121. possible to create userspace USB drivers using usbfs/libusb/gadgetfs
  122. that operate as fast as the USB bus allows. Because of this, the USB
  123. subsystem will not be allowing closed source kernel drivers to
  124. register with it, after this grace period is over. If anyone needs
  125. any help in converting their closed source drivers over to use the
  126. userspace filesystems, please contact the
  127. linux-usb-devel@lists.sourceforge.net mailing list, and the developers
  128. there will be glad to help you out.
  129. Who: Greg Kroah-Hartman <gregkh@suse.de>
  130. ---------------------------
  131. What: Interrupt only SA_* flags
  132. When: Januar 2007
  133. Why: The interrupt related SA_* flags are replaced by IRQF_* to move them
  134. out of the signal namespace.
  135. Who: Thomas Gleixner <tglx@linutronix.de>
  136. ---------------------------
  137. What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
  138. When: October 2008
  139. Why: The stacking of class devices makes these values misleading and
  140. inconsistent.
  141. Class devices should not carry any of these properties, and bus
  142. devices have SUBSYTEM and DRIVER as a replacement.
  143. Who: Kay Sievers <kay.sievers@suse.de>
  144. ---------------------------
  145. What: i2c-isa
  146. When: December 2006
  147. Why: i2c-isa is a non-sense and doesn't fit in the device driver
  148. model. Drivers relying on it are better implemented as platform
  149. drivers.
  150. Who: Jean Delvare <khali@linux-fr.org>
  151. ---------------------------
  152. What: i2c_adapter.dev
  153. i2c_adapter.list
  154. When: July 2007
  155. Why: Superfluous, given i2c_adapter.class_dev:
  156. * The "dev" was a stand-in for the physical device node that legacy
  157. drivers would not have; but now it's almost always present. Any
  158. remaining legacy drivers must upgrade (they now trigger warnings).
  159. * The "list" duplicates class device children.
  160. The delay in removing this is so upgraded lm_sensors and libsensors
  161. can get deployed. (Removal causes minor changes in the sysfs layout,
  162. notably the location of the adapter type name and parenting the i2c
  163. client hardware directly from their controller.)
  164. Who: Jean Delvare <khali@linux-fr.org>,
  165. David Brownell <dbrownell@users.sourceforge.net>
  166. ---------------------------
  167. What: drivers depending on OBSOLETE_OSS
  168. When: options in 2.6.22, code in 2.6.24
  169. Why: OSS drivers with ALSA replacements
  170. Who: Adrian Bunk <bunk@stusta.de>
  171. ---------------------------
  172. What: IPv4 only connection tracking/NAT/helpers
  173. When: 2.6.22
  174. Why: The new layer 3 independant connection tracking replaces the old
  175. IPv4 only version. After some stabilization of the new code the
  176. old one will be removed.
  177. Who: Patrick McHardy <kaber@trash.net>
  178. ---------------------------
  179. What: ACPI hooks (X86_SPEEDSTEP_CENTRINO_ACPI) in speedstep-centrino driver
  180. When: December 2006
  181. Why: Speedstep-centrino driver with ACPI hooks and acpi-cpufreq driver are
  182. functionally very much similar. They talk to ACPI in same way. Only
  183. difference between them is the way they do frequency transitions.
  184. One uses MSRs and the other one uses IO ports. Functionaliy of
  185. speedstep_centrino with ACPI hooks is now merged into acpi-cpufreq.
  186. That means one common driver will support all Intel Enhanced Speedstep
  187. capable CPUs. That means less confusion over name of
  188. speedstep-centrino driver (with that driver supposed to be used on
  189. non-centrino platforms). That means less duplication of code and
  190. less maintenance effort and no possibility of these two drivers
  191. going out of sync.
  192. Current users of speedstep_centrino with ACPI hooks are requested to
  193. switch over to acpi-cpufreq driver. speedstep-centrino will continue
  194. to work using older non-ACPI static table based scheme even after this
  195. date.
  196. Who: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
  197. ---------------------------
  198. What: /sys/firmware/acpi/namespace
  199. When: 2.6.21
  200. Why: The ACPI namespace is effectively the symbol list for
  201. the BIOS. The device names are completely arbitrary
  202. and have no place being exposed to user-space.
  203. For those interested in the BIOS ACPI namespace,
  204. the BIOS can be extracted and disassembled with acpidump
  205. and iasl as documented in the pmtools package here:
  206. http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/utils
  207. Who: Len Brown <len.brown@intel.com>
  208. ---------------------------
  209. What: ACPI procfs interface
  210. When: July 2007
  211. Why: After ACPI sysfs conversion, ACPI attributes will be duplicated
  212. in sysfs and the ACPI procfs interface should be removed.
  213. Who: Zhang Rui <rui.zhang@intel.com>
  214. ---------------------------
  215. What: /proc/acpi/button
  216. When: August 2007
  217. Why: /proc/acpi/button has been replaced by events to the input layer
  218. since 2.6.20.
  219. Who: Len Brown <len.brown@intel.com>
  220. ---------------------------
  221. What: sk98lin network driver
  222. When: July 2007
  223. Why: In kernel tree version of driver is unmaintained. Sk98lin driver
  224. replaced by the skge driver.
  225. Who: Stephen Hemminger <shemminger@osdl.org>
  226. ---------------------------
  227. What: Compaq touchscreen device emulation
  228. When: Oct 2007
  229. Files: drivers/input/tsdev.c
  230. Why: The code says it was obsolete when it was written in 2001.
  231. tslib is a userspace library which does anything tsdev can do and
  232. much more besides in userspace where this code belongs. There is no
  233. longer any need for tsdev and applications should have converted to
  234. use tslib by now.
  235. The name "tsdev" is also extremely confusing and lots of people have
  236. it loaded when they don't need/use it.
  237. Who: Richard Purdie <rpurdie@rpsys.net>
  238. ---------------------------
  239. What: Wireless extensions over netlink (CONFIG_NET_WIRELESS_RTNETLINK)
  240. When: with the merge of wireless-dev, 2.6.22 or later
  241. Why: The option/code is
  242. * not enabled on most kernels
  243. * not required by any userspace tools (except an experimental one,
  244. and even there only for some parts, others use ioctl)
  245. * pointless since wext is no longer evolving and the ioctl
  246. interface needs to be kept
  247. Who: Johannes Berg <johannes@sipsolutions.net>
  248. ---------------------------
  249. What: i8xx_tco watchdog driver
  250. When: in 2.6.22
  251. Why: the i8xx_tco watchdog driver has been replaced by the iTCO_wdt
  252. watchdog driver.
  253. Who: Wim Van Sebroeck <wim@iguana.be>
  254. ---------------------------