| blob | 83c88cae1eda446a12562a34f2c10f1a3f8cf5c2 |
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.
7 ---------------------------
9 What: dev->power.power_state
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>
20 ---------------------------
22 What: Video4Linux API 1 ioctls and video_decoder.h from Video devices.
23 When: December 2008
24 Files: include/linux/video_decoder.h include/linux/videodev.h
25 Check: include/linux/video_decoder.h include/linux/videodev.h
26 Why: V4L1 AP1 was replaced by V4L2 API during migration from 2.4 to 2.6
27 series. The old API have lots of drawbacks and don't provide enough
28 means to work with all video and audio standards. The newer API is
29 already available on the main drivers and should be used instead.
30 Newer drivers should use v4l_compat_translate_ioctl function to handle
31 old calls, replacing to newer ones.
32 Decoder iocts are using internally to allow video drivers to
33 communicate with video decoders. This should also be improved to allow
34 V4L2 calls being translated into compatible internal ioctls.
35 Compatibility ioctls will be provided, for a while, via
36 v4l1-compat module.
37 Who: Mauro Carvalho Chehab <mchehab@infradead.org>
39 ---------------------------
41 What: old tuner-3036 i2c driver
42 When: 2.6.28
43 Why: This driver is for VERY old i2c-over-parallel port teletext receiver
44 boxes. Rather then spending effort on converting this driver to V4L2,
45 and since it is extremely unlikely that anyone still uses one of these
46 devices, it was decided to drop it.
47 Who: Hans Verkuil <hverkuil@xs4all.nl>
48 Mauro Carvalho Chehab <mchehab@infradead.org>
50 ---------------------------
52 What: V4L2 dpc7146 driver
53 When: 2.6.28
54 Why: Old driver for the dpc7146 demonstration board that is no longer
55 relevant. The last time this was tested on actual hardware was
56 probably around 2002. Since this is a driver for a demonstration
57 board the decision was made to remove it rather than spending a
58 lot of effort continually updating this driver to stay in sync
59 with the latest internal V4L2 or I2C API.
60 Who: Hans Verkuil <hverkuil@xs4all.nl>
61 Mauro Carvalho Chehab <mchehab@infradead.org>
63 ---------------------------
65 What: PCMCIA control ioctl (needed for pcmcia-cs [cardmgr, cardctl])
66 When: November 2005
67 Files: drivers/pcmcia/: pcmcia_ioctl.c
68 Why: With the 16-bit PCMCIA subsystem now behaving (almost) like a
69 normal hotpluggable bus, and with it using the default kernel
70 infrastructure (hotplug, driver core, sysfs) keeping the PCMCIA
71 control ioctl needed by cardmgr and cardctl from pcmcia-cs is
72 unnecessary, and makes further cleanups and integration of the
73 PCMCIA subsystem into the Linux kernel device driver model more
74 difficult. The features provided by cardmgr and cardctl are either
75 handled by the kernel itself now or are available in the new
76 pcmciautils package available at
77 http://kernel.org/pub/linux/utils/kernel/pcmcia/
78 Who: Dominik Brodowski <linux@brodo.de>
80 ---------------------------
82 What: sys_sysctl
83 When: September 2010
84 Option: CONFIG_SYSCTL_SYSCALL
85 Why: The same information is available in a more convenient from
86 /proc/sys, and none of the sysctl variables appear to be
87 important performance wise.
89 Binary sysctls are a long standing source of subtle kernel
90 bugs and security issues.
92 When I looked several months ago all I could find after
93 searching several distributions were 5 user space programs and
94 glibc (which falls back to /proc/sys) using this syscall.
96 The man page for sysctl(2) documents it as unusable for user
97 space programs.
99 sysctl(2) is not generally ABI compatible to a 32bit user
100 space application on a 64bit and a 32bit kernel.
102 For the last several months the policy has been no new binary
103 sysctls and no one has put forward an argument to use them.
105 Binary sysctls issues seem to keep happening appearing so
106 properly deprecating them (with a warning to user space) and a
107 2 year grace warning period will mean eventually we can kill
108 them and end the pain.
110 In the mean time individual binary sysctls can be dealt with
111 in a piecewise fashion.
113 Who: Eric Biederman <ebiederm@xmission.com>
115 ---------------------------
117 What: remove EXPORT_SYMBOL(kernel_thread)
118 When: August 2006
119 Files: arch/*/kernel/*_ksyms.c
120 Check: kernel_thread
121 Why: kernel_thread is a low-level implementation detail. Drivers should
122 use the <linux/kthread.h> API instead which shields them from
123 implementation details and provides a higherlevel interface that
124 prevents bugs and code duplication
125 Who: Christoph Hellwig <hch@lst.de>
127 ---------------------------
129 What: eepro100 network driver
130 When: January 2007
131 Why: replaced by the e100 driver
132 Who: Adrian Bunk <bunk@stusta.de>
134 ---------------------------
136 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
137 (temporary transition config option provided until then)
138 The transition config option will also be removed at the same time.
139 When: before 2.6.19
140 Why: Unused symbols are both increasing the size of the kernel binary
141 and are often a sign of "wrong API"
142 Who: Arjan van de Ven <arjan@linux.intel.com>
144 ---------------------------
146 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
147 When: October 2008
148 Why: The stacking of class devices makes these values misleading and
149 inconsistent.
150 Class devices should not carry any of these properties, and bus
151 devices have SUBSYTEM and DRIVER as a replacement.
152 Who: Kay Sievers <kay.sievers@suse.de>
154 ---------------------------
156 What: ACPI procfs interface
157 When: July 2008
158 Why: ACPI sysfs conversion should be finished by January 2008.
159 ACPI procfs interface will be removed in July 2008 so that
160 there is enough time for the user space to catch up.
161 Who: Zhang Rui <rui.zhang@intel.com>
163 ---------------------------
165 What: /proc/acpi/button
166 When: August 2007
167 Why: /proc/acpi/button has been replaced by events to the input layer
168 since 2.6.20.
169 Who: Len Brown <len.brown@intel.com>
171 ---------------------------
173 What: /proc/acpi/event
174 When: February 2008
175 Why: /proc/acpi/event has been replaced by events via the input layer
176 and netlink since 2.6.23.
177 Who: Len Brown <len.brown@intel.com>
179 ---------------------------
181 What: libata spindown skipping and warning
182 When: Dec 2008
183 Why: Some halt(8) implementations synchronize caches for and spin
184 down libata disks because libata didn't use to spin down disk on
185 system halt (only synchronized caches).
186 Spin down on system halt is now implemented. sysfs node
187 /sys/class/scsi_disk/h:c:i:l/manage_start_stop is present if
188 spin down support is available.
189 Because issuing spin down command to an already spun down disk
190 makes some disks spin up just to spin down again, libata tracks
191 device spindown status to skip the extra spindown command and
192 warn about it.
193 This is to give userspace tools the time to get updated and will
194 be removed after userspace is reasonably updated.
195 Who: Tejun Heo <htejun@gmail.com>
197 ---------------------------
199 What: i386/x86_64 bzImage symlinks
200 When: April 2010
202 Why: The i386/x86_64 merge provides a symlink to the old bzImage
203 location so not yet updated user space tools, e.g. package
204 scripts, do not break.
205 Who: Thomas Gleixner <tglx@linutronix.de>
207 ---------------------------
209 What (Why):
210 - include/linux/netfilter_ipv4/ipt_TOS.h ipt_tos.h header files
211 (superseded by xt_TOS/xt_tos target & match)
213 - "forwarding" header files like ipt_mac.h in
214 include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/
216 - xt_CONNMARK match revision 0
217 (superseded by xt_CONNMARK match revision 1)
219 - xt_MARK target revisions 0 and 1
220 (superseded by xt_MARK match revision 2)
222 - xt_connmark match revision 0
223 (superseded by xt_connmark match revision 1)
225 - xt_conntrack match revision 0
226 (superseded by xt_conntrack match revision 1)
228 - xt_iprange match revision 0,
229 include/linux/netfilter_ipv4/ipt_iprange.h
230 (superseded by xt_iprange match revision 1)
232 - xt_mark match revision 0
233 (superseded by xt_mark match revision 1)
235 When: January 2009 or Linux 2.7.0, whichever comes first
236 Why: Superseded by newer revisions or modules
237 Who: Jan Engelhardt <jengelh@computergmbh.de>
239 ---------------------------
241 What: b43 support for firmware revision < 410
242 When: July 2008
243 Why: The support code for the old firmware hurts code readability/maintainability
244 and slightly hurts runtime performance. Bugfixes for the old firmware
245 are not provided by Broadcom anymore.
246 Who: Michael Buesch <mb@bu3sch.de>
248 ---------------------------
250 What: init_mm export
251 When: 2.6.26
252 Why: Not used in-tree. The current out-of-tree users used it to
253 work around problems in the CPA code which should be resolved
254 by now. One usecase was described to provide verification code
255 of the CPA operation. That's a good idea in general, but such
256 code / infrastructure should be in the kernel and not in some
257 out-of-tree driver.
258 Who: Thomas Gleixner <tglx@linutronix.de>
260 ----------------------------
262 What: usedac i386 kernel parameter
263 When: 2.6.27
264 Why: replaced by allowdac and no dac combination
265 Who: Glauber Costa <gcosta@redhat.com>
267 ---------------------------
269 What: old style serial driver for ColdFire (CONFIG_SERIAL_COLDFIRE)
270 When: 2.6.28
271 Why: This driver still uses the old interface and has been replaced
272 by CONFIG_SERIAL_MCF.
273 Who: Sebastian Siewior <sebastian@breakpoint.cc>
275 ---------------------------
277 What: /sys/o2cb symlink
278 When: January 2010
279 Why: /sys/fs/o2cb is the proper location for this information - /sys/o2cb
280 exists as a symlink for backwards compatibility for old versions of
281 ocfs2-tools. 2 years should be sufficient time to phase in new versions
282 which know to look in /sys/fs/o2cb.
283 Who: ocfs2-devel@oss.oracle.com
285 ---------------------------
287 What: SCTP_GET_PEER_ADDRS_NUM_OLD, SCTP_GET_PEER_ADDRS_OLD,
288 SCTP_GET_LOCAL_ADDRS_NUM_OLD, SCTP_GET_LOCAL_ADDRS_OLD
289 When: June 2009
290 Why: A newer version of the options have been introduced in 2005 that
291 removes the limitions of the old API. The sctp library has been
292 converted to use these new options at the same time. Any user
293 space app that directly uses the old options should convert to using
294 the new options.
295 Who: Vlad Yasevich <vladislav.yasevich@hp.com>
297 ---------------------------
299 What: CONFIG_THERMAL_HWMON
300 When: January 2009
301 Why: This option was introduced just to allow older lm-sensors userspace
302 to keep working over the upgrade to 2.6.26. At the scheduled time of
303 removal fixed lm-sensors (2.x or 3.x) should be readily available.
304 Who: Rene Herman <rene.herman@gmail.com>
306 ---------------------------
308 What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
309 (in net/core/net-sysfs.c)
310 When: After the only user (hal) has seen a release with the patches
311 for enough time, probably some time in 2010.
312 Why: Over 1K .text/.data size reduction, data is available in other
313 ways (ioctls)
314 Who: Johannes Berg <johannes@sipsolutions.net>
316 ---------------------------
318 What: CONFIG_NF_CT_ACCT
319 When: 2.6.29
320 Why: Accounting can now be enabled/disabled without kernel recompilation.
321 Currently used only to set a default value for a feature that is also
322 controlled by a kernel/module/sysfs/sysctl parameter.
323 Who: Krzysztof Piotr Oledzki <ole@ans.pl>
325 ---------------------------
327 What: ide-scsi (BLK_DEV_IDESCSI)
328 When: 2.6.29
329 Why: The 2.6 kernel supports direct writing to ide CD drives, which
330 eliminates the need for ide-scsi. The new method is more
331 efficient in every way.
332 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>