| blob | c23955404bf5d5056f7ddf448ea52415b6d3adfd |
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: old NCR53C9x driver
23 When: October 2007
24 Why: Replaced by the much better esp_scsi driver. Actual low-level
25 driver can be ported over almost trivially.
26 Who: David Miller <davem@davemloft.net>
27 Christoph Hellwig <hch@lst.de>
29 ---------------------------
31 What: Video4Linux API 1 ioctls and video_decoder.h from Video devices.
32 When: December 2008
33 Files: include/linux/video_decoder.h include/linux/videodev.h
34 Check: include/linux/video_decoder.h include/linux/videodev.h
35 Why: V4L1 AP1 was replaced by V4L2 API during migration from 2.4 to 2.6
36 series. The old API have lots of drawbacks and don't provide enough
37 means to work with all video and audio standards. The newer API is
38 already available on the main drivers and should be used instead.
39 Newer drivers should use v4l_compat_translate_ioctl function to handle
40 old calls, replacing to newer ones.
41 Decoder iocts are using internally to allow video drivers to
42 communicate with video decoders. This should also be improved to allow
43 V4L2 calls being translated into compatible internal ioctls.
44 Compatibility ioctls will be provided, for a while, via
45 v4l1-compat module.
46 Who: Mauro Carvalho Chehab <mchehab@infradead.org>
48 ---------------------------
50 What: old tuner-3036 i2c driver
51 When: 2.6.28
52 Why: This driver is for VERY old i2c-over-parallel port teletext receiver
53 boxes. Rather then spending effort on converting this driver to V4L2,
54 and since it is extremely unlikely that anyone still uses one of these
55 devices, it was decided to drop it.
56 Who: Hans Verkuil <hverkuil@xs4all.nl>
57 Mauro Carvalho Chehab <mchehab@infradead.org>
59 ---------------------------
61 What: V4L2 dpc7146 driver
62 When: 2.6.28
63 Why: Old driver for the dpc7146 demonstration board that is no longer
64 relevant. The last time this was tested on actual hardware was
65 probably around 2002. Since this is a driver for a demonstration
66 board the decision was made to remove it rather than spending a
67 lot of effort continually updating this driver to stay in sync
68 with the latest internal V4L2 or I2C API.
69 Who: Hans Verkuil <hverkuil@xs4all.nl>
70 Mauro Carvalho Chehab <mchehab@infradead.org>
72 ---------------------------
74 What: PCMCIA control ioctl (needed for pcmcia-cs [cardmgr, cardctl])
75 When: November 2005
76 Files: drivers/pcmcia/: pcmcia_ioctl.c
77 Why: With the 16-bit PCMCIA subsystem now behaving (almost) like a
78 normal hotpluggable bus, and with it using the default kernel
79 infrastructure (hotplug, driver core, sysfs) keeping the PCMCIA
80 control ioctl needed by cardmgr and cardctl from pcmcia-cs is
81 unnecessary, and makes further cleanups and integration of the
82 PCMCIA subsystem into the Linux kernel device driver model more
83 difficult. The features provided by cardmgr and cardctl are either
84 handled by the kernel itself now or are available in the new
85 pcmciautils package available at
86 http://kernel.org/pub/linux/utils/kernel/pcmcia/
87 Who: Dominik Brodowski <linux@brodo.de>
89 ---------------------------
91 What: sys_sysctl
92 When: September 2010
93 Option: CONFIG_SYSCTL_SYSCALL
94 Why: The same information is available in a more convenient from
95 /proc/sys, and none of the sysctl variables appear to be
96 important performance wise.
98 Binary sysctls are a long standing source of subtle kernel
99 bugs and security issues.
101 When I looked several months ago all I could find after
102 searching several distributions were 5 user space programs and
103 glibc (which falls back to /proc/sys) using this syscall.
105 The man page for sysctl(2) documents it as unusable for user
106 space programs.
108 sysctl(2) is not generally ABI compatible to a 32bit user
109 space application on a 64bit and a 32bit kernel.
111 For the last several months the policy has been no new binary
112 sysctls and no one has put forward an argument to use them.
114 Binary sysctls issues seem to keep happening appearing so
115 properly deprecating them (with a warning to user space) and a
116 2 year grace warning period will mean eventually we can kill
117 them and end the pain.
119 In the mean time individual binary sysctls can be dealt with
120 in a piecewise fashion.
122 Who: Eric Biederman <ebiederm@xmission.com>
124 ---------------------------
126 What: remove EXPORT_SYMBOL(kernel_thread)
127 When: August 2006
128 Files: arch/*/kernel/*_ksyms.c
129 Check: kernel_thread
130 Why: kernel_thread is a low-level implementation detail. Drivers should
131 use the <linux/kthread.h> API instead which shields them from
132 implementation details and provides a higherlevel interface that
133 prevents bugs and code duplication
134 Who: Christoph Hellwig <hch@lst.de>
136 ---------------------------
138 What: eepro100 network driver
139 When: January 2007
140 Why: replaced by the e100 driver
141 Who: Adrian Bunk <bunk@stusta.de>
143 ---------------------------
145 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
146 (temporary transition config option provided until then)
147 The transition config option will also be removed at the same time.
148 When: before 2.6.19
149 Why: Unused symbols are both increasing the size of the kernel binary
150 and are often a sign of "wrong API"
151 Who: Arjan van de Ven <arjan@linux.intel.com>
153 ---------------------------
155 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
156 When: October 2008
157 Why: The stacking of class devices makes these values misleading and
158 inconsistent.
159 Class devices should not carry any of these properties, and bus
160 devices have SUBSYTEM and DRIVER as a replacement.
161 Who: Kay Sievers <kay.sievers@suse.de>
163 ---------------------------
165 What: ACPI procfs interface
166 When: July 2008
167 Why: ACPI sysfs conversion should be finished by January 2008.
168 ACPI procfs interface will be removed in July 2008 so that
169 there is enough time for the user space to catch up.
170 Who: Zhang Rui <rui.zhang@intel.com>
172 ---------------------------
174 What: /proc/acpi/button
175 When: August 2007
176 Why: /proc/acpi/button has been replaced by events to the input layer
177 since 2.6.20.
178 Who: Len Brown <len.brown@intel.com>
180 ---------------------------
182 What: /proc/acpi/event
183 When: February 2008
184 Why: /proc/acpi/event has been replaced by events via the input layer
185 and netlink since 2.6.23.
186 Who: Len Brown <len.brown@intel.com>
188 ---------------------------
190 What: libata spindown skipping and warning
191 When: Dec 2008
192 Why: Some halt(8) implementations synchronize caches for and spin
193 down libata disks because libata didn't use to spin down disk on
194 system halt (only synchronized caches).
195 Spin down on system halt is now implemented. sysfs node
196 /sys/class/scsi_disk/h:c:i:l/manage_start_stop is present if
197 spin down support is available.
198 Because issuing spin down command to an already spun down disk
199 makes some disks spin up just to spin down again, libata tracks
200 device spindown status to skip the extra spindown command and
201 warn about it.
202 This is to give userspace tools the time to get updated and will
203 be removed after userspace is reasonably updated.
204 Who: Tejun Heo <htejun@gmail.com>
206 ---------------------------
208 What: The arch/ppc and include/asm-ppc directories
209 When: Jun 2008
210 Why: The arch/powerpc tree is the merged architecture for ppc32 and ppc64
211 platforms. Currently there are efforts underway to port the remaining
212 arch/ppc platforms to the merged tree. New submissions to the arch/ppc
213 tree have been frozen with the 2.6.22 kernel release and that tree will
214 remain in bug-fix only mode until its scheduled removal. Platforms
215 that are not ported by June 2008 will be removed due to the lack of an
216 interested maintainer.
217 Who: linuxppc-dev@ozlabs.org
219 ---------------------------
221 What: i386/x86_64 bzImage symlinks
222 When: April 2010
224 Why: The i386/x86_64 merge provides a symlink to the old bzImage
225 location so not yet updated user space tools, e.g. package
226 scripts, do not break.
227 Who: Thomas Gleixner <tglx@linutronix.de>
229 ---------------------------
231 What (Why):
232 - include/linux/netfilter_ipv4/ipt_TOS.h ipt_tos.h header files
233 (superseded by xt_TOS/xt_tos target & match)
235 - "forwarding" header files like ipt_mac.h in
236 include/linux/netfilter_ipv4/ and include/linux/netfilter_ipv6/
238 - xt_CONNMARK match revision 0
239 (superseded by xt_CONNMARK match revision 1)
241 - xt_MARK target revisions 0 and 1
242 (superseded by xt_MARK match revision 2)
244 - xt_connmark match revision 0
245 (superseded by xt_connmark match revision 1)
247 - xt_conntrack match revision 0
248 (superseded by xt_conntrack match revision 1)
250 - xt_iprange match revision 0,
251 include/linux/netfilter_ipv4/ipt_iprange.h
252 (superseded by xt_iprange match revision 1)
254 - xt_mark match revision 0
255 (superseded by xt_mark match revision 1)
257 When: January 2009 or Linux 2.7.0, whichever comes first
258 Why: Superseded by newer revisions or modules
259 Who: Jan Engelhardt <jengelh@computergmbh.de>
261 ---------------------------
263 What: b43 support for firmware revision < 410
264 When: July 2008
265 Why: The support code for the old firmware hurts code readability/maintainability
266 and slightly hurts runtime performance. Bugfixes for the old firmware
267 are not provided by Broadcom anymore.
268 Who: Michael Buesch <mb@bu3sch.de>
270 ---------------------------
272 What: init_mm export
273 When: 2.6.26
274 Why: Not used in-tree. The current out-of-tree users used it to
275 work around problems in the CPA code which should be resolved
276 by now. One usecase was described to provide verification code
277 of the CPA operation. That's a good idea in general, but such
278 code / infrastructure should be in the kernel and not in some
279 out-of-tree driver.
280 Who: Thomas Gleixner <tglx@linutronix.de>
282 ----------------------------
284 What: usedac i386 kernel parameter
285 When: 2.6.27
286 Why: replaced by allowdac and no dac combination
287 Who: Glauber Costa <gcosta@redhat.com>
289 ---------------------------
291 What: old style serial driver for ColdFire (CONFIG_SERIAL_COLDFIRE)
292 When: 2.6.28
293 Why: This driver still uses the old interface and has been replaced
294 by CONFIG_SERIAL_MCF.
295 Who: Sebastian Siewior <sebastian@breakpoint.cc>
297 ---------------------------
299 What: /sys/o2cb symlink
300 When: January 2010
301 Why: /sys/fs/o2cb is the proper location for this information - /sys/o2cb
302 exists as a symlink for backwards compatibility for old versions of
303 ocfs2-tools. 2 years should be sufficient time to phase in new versions
304 which know to look in /sys/fs/o2cb.
305 Who: ocfs2-devel@oss.oracle.com
307 ---------------------------
309 What: SCTP_GET_PEER_ADDRS_NUM_OLD, SCTP_GET_PEER_ADDRS_OLD,
310 SCTP_GET_LOCAL_ADDRS_NUM_OLD, SCTP_GET_LOCAL_ADDRS_OLD
311 When: June 2009
312 Why: A newer version of the options have been introduced in 2005 that
313 removes the limitions of the old API. The sctp library has been
314 converted to use these new options at the same time. Any user
315 space app that directly uses the old options should convert to using
316 the new options.
317 Who: Vlad Yasevich <vladislav.yasevich@hp.com>
319 ---------------------------
321 What: CONFIG_THERMAL_HWMON
322 When: January 2009
323 Why: This option was introduced just to allow older lm-sensors userspace
324 to keep working over the upgrade to 2.6.26. At the scheduled time of
325 removal fixed lm-sensors (2.x or 3.x) should be readily available.
326 Who: Rene Herman <rene.herman@gmail.com>
328 ---------------------------
330 What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
331 (in net/core/net-sysfs.c)
332 When: After the only user (hal) has seen a release with the patches
333 for enough time, probably some time in 2010.
334 Why: Over 1K .text/.data size reduction, data is available in other
335 ways (ioctls)
336 Who: Johannes Berg <johannes@sipsolutions.net>
338 ---------------------------
340 What: CONFIG_NF_CT_ACCT
341 When: 2.6.29
342 Why: Accounting can now be enabled/disabled without kernel recompilation.
343 Currently used only to set a default value for a feature that is also
344 controlled by a kernel/module/sysfs/sysctl parameter.
345 Who: Krzysztof Piotr Oledzki <ole@ans.pl>