tjh.dev
Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1The following is a list of files and features that are going to be
2removed in the kernel source tree. Every entry should contain what
3exactly is going away, why it is happening, and who is going to be doing
4the work. When the feature is removed from the kernel, it should also
5be removed from this file.
6
7---------------------------
8
9What: /sys/devices/.../power/state
10 dev->power.power_state
11 dpm_runtime_{suspend,resume)()
12When: July 2007
13Why: Broken design for runtime control over driver power states, confusing
14 driver-internal runtime power management with: mechanisms to support
15 system-wide sleep state transitions; event codes that distinguish
16 different phases of swsusp "sleep" transitions; and userspace policy
17 inputs. This framework was never widely used, and most attempts to
18 use it were broken. Drivers should instead be exposing domain-specific
19 interfaces either to kernel or to userspace.
20Who: Pavel Machek <pavel@suse.cz>
21
22---------------------------
23
24What: RAW driver (CONFIG_RAW_DRIVER)
25When: December 2005
26Why: declared obsolete since kernel 2.6.3
27 O_DIRECT can be used instead
28Who: Adrian Bunk <bunk@stusta.de>
29
30---------------------------
31
32What: raw1394: requests of type RAW1394_REQ_ISO_SEND, RAW1394_REQ_ISO_LISTEN
33When: November 2006
34Why: Deprecated in favour of the new ioctl-based rawiso interface, which is
35 more efficient. You should really be using libraw1394 for raw1394
36 access anyway.
37Who: Jody McIntyre <scjody@modernduck.com>
38
39---------------------------
40
41What: Video4Linux API 1 ioctls and video_decoder.h from Video devices.
42When: December 2006
43Why: V4L1 AP1 was replaced by V4L2 API. during migration from 2.4 to 2.6
44 series. The old API have lots of drawbacks and don't provide enough
45 means to work with all video and audio standards. The newer API is
46 already available on the main drivers and should be used instead.
47 Newer drivers should use v4l_compat_translate_ioctl function to handle
48 old calls, replacing to newer ones.
49 Decoder iocts are using internally to allow video drivers to
50 communicate with video decoders. This should also be improved to allow
51 V4L2 calls being translated into compatible internal ioctls.
52Who: Mauro Carvalho Chehab <mchehab@brturbo.com.br>
53
54---------------------------
55
56What: sys_sysctl
57When: January 2007
58Why: The same information is available through /proc/sys and that is the
59 interface user space prefers to use. And there do not appear to be
60 any existing user in user space of sys_sysctl. The additional
61 maintenance overhead of keeping a set of binary names gets
62 in the way of doing a good job of maintaining this interface.
63
64Who: Eric Biederman <ebiederm@xmission.com>
65
66---------------------------
67
68What: PCMCIA control ioctl (needed for pcmcia-cs [cardmgr, cardctl])
69When: November 2005
70Files: drivers/pcmcia/: pcmcia_ioctl.c
71Why: With the 16-bit PCMCIA subsystem now behaving (almost) like a
72 normal hotpluggable bus, and with it using the default kernel
73 infrastructure (hotplug, driver core, sysfs) keeping the PCMCIA
74 control ioctl needed by cardmgr and cardctl from pcmcia-cs is
75 unnecessary, and makes further cleanups and integration of the
76 PCMCIA subsystem into the Linux kernel device driver model more
77 difficult. The features provided by cardmgr and cardctl are either
78 handled by the kernel itself now or are available in the new
79 pcmciautils package available at
80 http://kernel.org/pub/linux/utils/kernel/pcmcia/
81Who: Dominik Brodowski <linux@brodo.de>
82
83---------------------------
84
85What: ip_queue and ip6_queue (old ipv4-only and ipv6-only netfilter queue)
86When: December 2005
87Why: This interface has been obsoleted by the new layer3-independent
88 "nfnetlink_queue". The Kernel interface is compatible, so the old
89 ip[6]tables "QUEUE" targets still work and will transparently handle
90 all packets into nfnetlink queue number 0. Userspace users will have
91 to link against API-compatible library on top of libnfnetlink_queue
92 instead of the current 'libipq'.
93Who: Harald Welte <laforge@netfilter.org>
94
95---------------------------
96
97What: remove EXPORT_SYMBOL(kernel_thread)
98When: August 2006
99Files: arch/*/kernel/*_ksyms.c
100Why: kernel_thread is a low-level implementation detail. Drivers should
101 use the <linux/kthread.h> API instead which shields them from
102 implementation details and provides a higherlevel interface that
103 prevents bugs and code duplication
104Who: Christoph Hellwig <hch@lst.de>
105
106---------------------------
107
108What: CONFIG_FORCED_INLINING
109When: June 2006
110Why: Config option is there to see if gcc is good enough. (in january
111 2006). If it is, the behavior should just be the default. If it's not,
112 the option should just go away entirely.
113Who: Arjan van de Ven
114
115---------------------------
116
117What: eepro100 network driver
118When: January 2007
119Why: replaced by the e100 driver
120Who: Adrian Bunk <bunk@stusta.de>
121
122---------------------------
123
124What: drivers depending on OSS_OBSOLETE_DRIVER
125When: options in 2.6.20, code in 2.6.22
126Why: OSS drivers with ALSA replacements
127Who: Adrian Bunk <bunk@stusta.de>
128
129---------------------------
130
131What: pci_module_init(driver)
132When: January 2007
133Why: Is replaced by pci_register_driver(pci_driver).
134Who: Richard Knutsson <ricknu-0@student.ltu.se> and Greg Kroah-Hartman <gregkh@suse.de>
135
136---------------------------
137
138What: Usage of invalid timevals in setitimer
139When: March 2007
140Why: POSIX requires to validate timevals in the setitimer call. This
141 was never done by Linux. The invalid (e.g. negative timevals) were
142 silently converted to more or less random timeouts and intervals.
143 Until the removal a per boot limited number of warnings is printed
144 and the timevals are sanitized.
145
146Who: Thomas Gleixner <tglx@linutronix.de>
147
148---------------------------
149
150What: I2C interface of the it87 driver
151When: January 2007
152Why: The ISA interface is faster and should be always available. The I2C
153 probing is also known to cause trouble in at least one case (see
154 bug #5889.)
155Who: Jean Delvare <khali@linux-fr.org>
156
157---------------------------
158
159What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
160 (temporary transition config option provided until then)
161 The transition config option will also be removed at the same time.
162When: before 2.6.19
163Why: Unused symbols are both increasing the size of the kernel binary
164 and are often a sign of "wrong API"
165Who: Arjan van de Ven <arjan@linux.intel.com>
166
167---------------------------
168
169What: mount/umount uevents
170When: February 2007
171Why: These events are not correct, and do not properly let userspace know
172 when a file system has been mounted or unmounted. Userspace should
173 poll the /proc/mounts file instead to detect this properly.
174Who: Greg Kroah-Hartman <gregkh@suse.de>
175
176---------------------------
177
178What: USB driver API moves to EXPORT_SYMBOL_GPL
179When: February 2008
180Files: include/linux/usb.h, drivers/usb/core/driver.c
181Why: The USB subsystem has changed a lot over time, and it has been
182 possible to create userspace USB drivers using usbfs/libusb/gadgetfs
183 that operate as fast as the USB bus allows. Because of this, the USB
184 subsystem will not be allowing closed source kernel drivers to
185 register with it, after this grace period is over. If anyone needs
186 any help in converting their closed source drivers over to use the
187 userspace filesystems, please contact the
188 linux-usb-devel@lists.sourceforge.net mailing list, and the developers
189 there will be glad to help you out.
190Who: Greg Kroah-Hartman <gregkh@suse.de>
191
192---------------------------
193
194What: find_trylock_page
195When: January 2007
196Why: The interface no longer has any callers left in the kernel. It
197 is an odd interface (compared with other find_*_page functions), in
198 that it does not take a refcount to the page, only the page lock.
199 It should be replaced with find_get_page or find_lock_page if possible.
200 This feature removal can be reevaluated if users of the interface
201 cannot cleanly use something else.
202Who: Nick Piggin <npiggin@suse.de>
203
204---------------------------
205
206What: Interrupt only SA_* flags
207When: Januar 2007
208Why: The interrupt related SA_* flags are replaced by IRQF_* to move them
209 out of the signal namespace.
210
211Who: Thomas Gleixner <tglx@linutronix.de>
212
213---------------------------
214
215What: i2c-ite and i2c-algo-ite drivers
216When: September 2006
217Why: These drivers never compiled since they were added to the kernel
218 tree 5 years ago. This feature removal can be reevaluated if
219 someone shows interest in the drivers, fixes them and takes over
220 maintenance.
221 http://marc.theaimsgroup.com/?l=linux-mips&m=115040510817448
222Who: Jean Delvare <khali@linux-fr.org>
223
224---------------------------
225
226What: Bridge netfilter deferred IPv4/IPv6 output hook calling
227When: January 2007
228Why: The deferred output hooks are a layering violation causing unusual
229 and broken behaviour on bridge devices. Examples of things they
230 break include QoS classifation using the MARK or CLASSIFY targets,
231 the IPsec policy match and connection tracking with VLANs on a
232 bridge. Their only use is to enable bridge output port filtering
233 within iptables with the physdev match, which can also be done by
234 combining iptables and ebtables using netfilter marks. Until it
235 will get removed the hook deferral is disabled by default and is
236 only enabled when needed.
237
238Who: Patrick McHardy <kaber@trash.net>
239
240---------------------------
241
242What: frame diverter
243When: November 2006
244Why: The frame diverter is included in most distribution kernels, but is
245 broken. It does not correctly handle many things:
246 - IPV6
247 - non-linear skb's
248 - network device RCU on removal
249 - input frames not correctly checked for protocol errors
250 It also adds allocation overhead even if not enabled.
251 It is not clear if anyone is still using it.
252Who: Stephen Hemminger <shemminger@osdl.org>
253
254---------------------------
255
256
257What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
258When: Oktober 2008
259Why: The stacking of class devices makes these values misleading and
260 inconsistent.
261 Class devices should not carry any of these properties, and bus
262 devices have SUBSYTEM and DRIVER as a replacement.
263Who: Kay Sievers <kay.sievers@suse.de>
264
265---------------------------
266
267What: i2c-isa
268When: December 2006
269Why: i2c-isa is a non-sense and doesn't fit in the device driver
270 model. Drivers relying on it are better implemented as platform
271 drivers.
272Who: Jean Delvare <khali@linux-fr.org>
273
274---------------------------
275
276What: ftape
277When: 2.6.20
278Why: Orphaned for ages. SMP bugs long unfixed. Few users left
279 in the world.
280Who: Jeff Garzik <jeff@garzik.org>
281
282---------------------------