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: /sys/devices/.../power/state
10 dev->power.power_state
11 dpm_runtime_{suspend,resume)()
13 Why: 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.
20 Who: Pavel Machek <pavel@suse.cz>
22 ---------------------------
24 What: RAW driver (CONFIG_RAW_DRIVER)
26 Why: declared obsolete since kernel 2.6.3
27 O_DIRECT can be used instead
28 Who: Adrian Bunk <bunk@stusta.de>
30 ---------------------------
32 What: raw1394: requests of type RAW1394_REQ_ISO_SEND, RAW1394_REQ_ISO_LISTEN
34 Why: Deprecated in favour of the more efficient and robust rawiso interface.
35 Affected are applications which use the deprecated part of libraw1394
36 (raw1394_iso_write, raw1394_start_iso_write, raw1394_start_iso_rcv,
37 raw1394_stop_iso_rcv) or bypass libraw1394.
38 Who: Dan Dennedy <dan@dennedy.org>, Stefan Richter <stefanr@s5r6.in-berlin.de>
40 ---------------------------
42 What: dv1394 driver (CONFIG_IEEE1394_DV1394)
44 Why: Replaced by raw1394 + userspace libraries, notably libiec61883. This
45 shift of application support has been indicated on www.linux1394.org
46 and developers' mailinglists for quite some time. Major applications
47 have been converted, with the exception of ffmpeg and hence xine.
48 Piped output of dvgrab2 is a partial equivalent to dv1394.
49 Who: Dan Dennedy <dan@dennedy.org>, Stefan Richter <stefanr@s5r6.in-berlin.de>
51 ---------------------------
53 What: Video4Linux API 1 ioctls and video_decoder.h from Video devices.
55 Why: V4L1 AP1 was replaced by V4L2 API. during migration from 2.4 to 2.6
56 series. The old API have lots of drawbacks and don't provide enough
57 means to work with all video and audio standards. The newer API is
58 already available on the main drivers and should be used instead.
59 Newer drivers should use v4l_compat_translate_ioctl function to handle
60 old calls, replacing to newer ones.
61 Decoder iocts are using internally to allow video drivers to
62 communicate with video decoders. This should also be improved to allow
63 V4L2 calls being translated into compatible internal ioctls.
64 Who: Mauro Carvalho Chehab <mchehab@brturbo.com.br>
66 ---------------------------
68 What: PCMCIA control ioctl (needed for pcmcia-cs [cardmgr, cardctl])
70 Files: drivers/pcmcia/: pcmcia_ioctl.c
71 Why: 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/
81 Who: Dominik Brodowski <linux@brodo.de>
83 ---------------------------
85 What: remove EXPORT_SYMBOL(kernel_thread)
87 Files: arch/*/kernel/*_ksyms.c
88 Why: kernel_thread is a low-level implementation detail. Drivers should
89 use the <linux/kthread.h> API instead which shields them from
90 implementation details and provides a higherlevel interface that
91 prevents bugs and code duplication
92 Who: Christoph Hellwig <hch@lst.de>
94 ---------------------------
96 What: CONFIG_FORCED_INLINING
98 Why: Config option is there to see if gcc is good enough. (in january
99 2006). If it is, the behavior should just be the default. If it's not,
100 the option should just go away entirely.
101 Who: Arjan van de Ven
103 ---------------------------
105 What: eepro100 network driver
107 Why: replaced by the e100 driver
108 Who: Adrian Bunk <bunk@stusta.de>
110 ---------------------------
112 What: drivers depending on OSS_OBSOLETE_DRIVER
113 When: options in 2.6.20, code in 2.6.22
114 Why: OSS drivers with ALSA replacements
115 Who: Adrian Bunk <bunk@stusta.de>
117 ---------------------------
119 What: pci_module_init(driver)
121 Why: Is replaced by pci_register_driver(pci_driver).
122 Who: Richard Knutsson <ricknu-0@student.ltu.se> and Greg Kroah-Hartman <gregkh@suse.de>
124 ---------------------------
126 What: Usage of invalid timevals in setitimer
128 Why: POSIX requires to validate timevals in the setitimer call. This
129 was never done by Linux. The invalid (e.g. negative timevals) were
130 silently converted to more or less random timeouts and intervals.
131 Until the removal a per boot limited number of warnings is printed
132 and the timevals are sanitized.
134 Who: Thomas Gleixner <tglx@linutronix.de>
136 ---------------------------
138 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
139 (temporary transition config option provided until then)
140 The transition config option will also be removed at the same time.
142 Why: Unused symbols are both increasing the size of the kernel binary
143 and are often a sign of "wrong API"
144 Who: Arjan van de Ven <arjan@linux.intel.com>
146 ---------------------------
148 What: mount/umount uevents
150 Why: These events are not correct, and do not properly let userspace know
151 when a file system has been mounted or unmounted. Userspace should
152 poll the /proc/mounts file instead to detect this properly.
153 Who: Greg Kroah-Hartman <gregkh@suse.de>
155 ---------------------------
157 What: USB driver API moves to EXPORT_SYMBOL_GPL
159 Files: include/linux/usb.h, drivers/usb/core/driver.c
160 Why: The USB subsystem has changed a lot over time, and it has been
161 possible to create userspace USB drivers using usbfs/libusb/gadgetfs
162 that operate as fast as the USB bus allows. Because of this, the USB
163 subsystem will not be allowing closed source kernel drivers to
164 register with it, after this grace period is over. If anyone needs
165 any help in converting their closed source drivers over to use the
166 userspace filesystems, please contact the
167 linux-usb-devel@lists.sourceforge.net mailing list, and the developers
168 there will be glad to help you out.
169 Who: Greg Kroah-Hartman <gregkh@suse.de>
171 ---------------------------
173 What: find_trylock_page
175 Why: The interface no longer has any callers left in the kernel. It
176 is an odd interface (compared with other find_*_page functions), in
177 that it does not take a refcount to the page, only the page lock.
178 It should be replaced with find_get_page or find_lock_page if possible.
179 This feature removal can be reevaluated if users of the interface
180 cannot cleanly use something else.
181 Who: Nick Piggin <npiggin@suse.de>
183 ---------------------------
185 What: Interrupt only SA_* flags
187 Why: The interrupt related SA_* flags are replaced by IRQF_* to move them
188 out of the signal namespace.
190 Who: Thomas Gleixner <tglx@linutronix.de>
192 ---------------------------
194 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
196 Why: The stacking of class devices makes these values misleading and
198 Class devices should not carry any of these properties, and bus
199 devices have SUBSYTEM and DRIVER as a replacement.
200 Who: Kay Sievers <kay.sievers@suse.de>
202 ---------------------------
206 Why: i2c-isa is a non-sense and doesn't fit in the device driver
207 model. Drivers relying on it are better implemented as platform
209 Who: Jean Delvare <khali@linux-fr.org>
211 ---------------------------
213 What: i2c_adapter.dev
216 Why: Superfluous, given i2c_adapter.class_dev:
217 * The "dev" was a stand-in for the physical device node that legacy
218 drivers would not have; but now it's almost always present. Any
219 remaining legacy drivers must upgrade (they now trigger warnings).
220 * The "list" duplicates class device children.
221 The delay in removing this is so upgraded lm_sensors and libsensors
222 can get deployed. (Removal causes minor changes in the sysfs layout,
223 notably the location of the adapter type name and parenting the i2c
224 client hardware directly from their controller.)
225 Who: Jean Delvare <khali@linux-fr.org>,
226 David Brownell <dbrownell@users.sourceforge.net>
228 ---------------------------
230 What: IPv4 only connection tracking/NAT/helpers
232 Why: The new layer 3 independant connection tracking replaces the old
233 IPv4 only version. After some stabilization of the new code the
234 old one will be removed.
235 Who: Patrick McHardy <kaber@trash.net>
237 ---------------------------
239 What: ACPI hooks (X86_SPEEDSTEP_CENTRINO_ACPI) in speedstep-centrino driver
241 Why: Speedstep-centrino driver with ACPI hooks and acpi-cpufreq driver are
242 functionally very much similar. They talk to ACPI in same way. Only
243 difference between them is the way they do frequency transitions.
244 One uses MSRs and the other one uses IO ports. Functionaliy of
245 speedstep_centrino with ACPI hooks is now merged into acpi-cpufreq.
246 That means one common driver will support all Intel Enhanced Speedstep
247 capable CPUs. That means less confusion over name of
248 speedstep-centrino driver (with that driver supposed to be used on
249 non-centrino platforms). That means less duplication of code and
250 less maintenance effort and no possibility of these two drivers
252 Current users of speedstep_centrino with ACPI hooks are requested to
253 switch over to acpi-cpufreq driver. speedstep-centrino will continue
254 to work using older non-ACPI static table based scheme even after this
257 Who: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
259 ---------------------------
261 What: ACPI hotkey driver (CONFIG_ACPI_HOTKEY)
263 Why: hotkey.c was an attempt to consolidate multiple drivers that use
264 ACPI to implement hotkeys. However, hotkeys are not documented
265 in the ACPI specification, so the drivers used undocumented
266 vendor-specific hooks and turned out to be more different than
269 Further, the keys and the features supplied by each platform
270 are different, so there will always be a need for
271 platform-specific drivers.
273 So the new plan is to delete hotkey.c and instead, work on the
274 platform specific drivers to try to make them look the same
275 to the user when they supply the same features.
277 hotkey.c has always depended on CONFIG_EXPERIMENTAL
279 Who: Len Brown <len.brown@intel.com>
281 ---------------------------
283 What: /sys/firmware/acpi/namespace
285 Why: The ACPI namespace is effectively the symbol list for
286 the BIOS. The device names are completely arbitrary
287 and have no place being exposed to user-space.
289 For those interested in the BIOS ACPI namespace,
290 the BIOS can be extracted and disassembled with acpidump
291 and iasl as documented in the pmtools package here:
292 http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/utils
294 Who: Len Brown <len.brown@intel.com>
296 ---------------------------
298 What: /proc/acpi/button
300 Why: /proc/acpi/button has been replaced by events to the input layer
302 Who: Len Brown <len.brown@intel.com>
304 ---------------------------
306 What: JFFS (version 1)
308 Why: Unmaintained for years, superceded by JFFS2 for years.
309 Who: Jeff Garzik <jeff@garzik.org>
311 ---------------------------