Root/Documentation/feature-removal-schedule.txt

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: PRISM54
10When: 2.6.34
11
12Why: prism54 FullMAC PCI / Cardbus devices used to be supported only by the
13    prism54 wireless driver. After Intersil stopped selling these
14    devices in preference for the newer more flexible SoftMAC devices
15    a SoftMAC device driver was required and prism54 did not support
16    them. The p54pci driver now exists and has been present in the kernel for
17    a while. This driver supports both SoftMAC devices and FullMAC devices.
18    The main difference between these devices was the amount of memory which
19    could be used for the firmware. The SoftMAC devices support a smaller
20    amount of memory. Because of this the SoftMAC firmware fits into FullMAC
21    devices's memory. p54pci supports not only PCI / Cardbus but also USB
22    and SPI. Since p54pci supports all devices prism54 supports
23    you will have a conflict. I'm not quite sure how distributions are
24    handling this conflict right now. prism54 was kept around due to
25    claims users may experience issues when using the SoftMAC driver.
26    Time has passed users have not reported issues. If you use prism54
27    and for whatever reason you cannot use p54pci please let us know!
28    E-mail us at: linux-wireless@vger.kernel.org
29
30    For more information see the p54 wiki page:
31
32    http://wireless.kernel.org/en/users/Drivers/p54
33
34Who: Luis R. Rodriguez <lrodriguez@atheros.com>
35
36---------------------------
37
38What: IRQF_SAMPLE_RANDOM
39Check: IRQF_SAMPLE_RANDOM
40When: July 2009
41
42Why: Many of IRQF_SAMPLE_RANDOM users are technically bogus as entropy
43    sources in the kernel's current entropy model. To resolve this, every
44    input point to the kernel's entropy pool needs to better document the
45    type of entropy source it actually is. This will be replaced with
46    additional add_*_randomness functions in drivers/char/random.c
47
48Who: Robin Getz <rgetz@blackfin.uclinux.org> & Matt Mackall <mpm@selenic.com>
49
50---------------------------
51
52What: Deprecated snapshot ioctls
53When: 2.6.36
54
55Why: The ioctls in kernel/power/user.c were marked as deprecated long time
56    ago. Now they notify users about that so that they need to replace
57    their userspace. After some more time, remove them completely.
58
59Who: Jiri Slaby <jirislaby@gmail.com>
60
61---------------------------
62
63What: The ieee80211_regdom module parameter
64When: March 2010 / desktop catchup
65
66Why: This was inherited by the CONFIG_WIRELESS_OLD_REGULATORY code,
67    and currently serves as an option for users to define an
68    ISO / IEC 3166 alpha2 code for the country they are currently
69    present in. Although there are userspace API replacements for this
70    through nl80211 distributions haven't yet caught up with implementing
71    decent alternatives through standard GUIs. Although available as an
72    option through iw or wpa_supplicant its just a matter of time before
73    distributions pick up good GUI options for this. The ideal solution
74    would actually consist of intelligent designs which would do this for
75    the user automatically even when travelling through different countries.
76    Until then we leave this module parameter as a compromise.
77
78    When userspace improves with reasonable widely-available alternatives for
79    this we will no longer need this module parameter. This entry hopes that
80    by the super-futuristically looking date of "March 2010" we will have
81    such replacements widely available.
82
83Who: Luis R. Rodriguez <lrodriguez@atheros.com>
84
85---------------------------
86
87What: dev->power.power_state
88When: July 2007
89Why: Broken design for runtime control over driver power states, confusing
90    driver-internal runtime power management with: mechanisms to support
91    system-wide sleep state transitions; event codes that distinguish
92    different phases of swsusp "sleep" transitions; and userspace policy
93    inputs. This framework was never widely used, and most attempts to
94    use it were broken. Drivers should instead be exposing domain-specific
95    interfaces either to kernel or to userspace.
96Who: Pavel Machek <pavel@ucw.cz>
97
98---------------------------
99
100What: Video4Linux obsolete drivers using V4L1 API
101When: kernel 2.6.39
102Files: drivers/staging/se401/* drivers/staging/usbvideo/*
103Check: drivers/staging/se401/se401.c drivers/staging/usbvideo/usbvideo.c
104Why: There are some drivers still using V4L1 API, despite all efforts we've done
105    to migrate. Those drivers are for obsolete hardware that the old maintainer
106    didn't care (or not have the hardware anymore), and that no other developer
107    could find any hardware to buy. They probably have no practical usage today,
108    and people with such old hardware could probably keep using an older version
109    of the kernel. Those drivers will be moved to staging on 2.6.38 and, if nobody
110    cares enough to port and test them with V4L2 API, they'll be removed on 2.6.39.
111Who: Mauro Carvalho Chehab <mchehab@infradead.org>
112
113---------------------------
114
115What: Video4Linux: Remove obsolete ioctl's
116When: kernel 2.6.39
117Files: include/media/videodev2.h
118Why: Some ioctl's were defined wrong on 2.6.2 and 2.6.6, using the wrong
119    type of R/W arguments. They were fixed, but the old ioctl names are
120    still there, maintained to avoid breaking binary compatibility:
121      #define VIDIOC_OVERLAY_OLD _IOWR('V', 14, int)
122      #define VIDIOC_S_PARM_OLD _IOW('V', 22, struct v4l2_streamparm)
123      #define VIDIOC_S_CTRL_OLD _IOW('V', 28, struct v4l2_control)
124      #define VIDIOC_G_AUDIO_OLD _IOWR('V', 33, struct v4l2_audio)
125      #define VIDIOC_G_AUDOUT_OLD _IOWR('V', 49, struct v4l2_audioout)
126      #define VIDIOC_CROPCAP_OLD _IOR('V', 58, struct v4l2_cropcap)
127    There's no sense on preserving those forever, as it is very doubtful
128    that someone would try to use a such old binary with a modern kernel.
129    Removing them will allow us to remove some magic done at the V4L ioctl
130    handler.
131
132Who: Mauro Carvalho Chehab <mchehab@infradead.org>
133
134---------------------------
135
136What: sys_sysctl
137When: September 2010
138Option: CONFIG_SYSCTL_SYSCALL
139Why: The same information is available in a more convenient from
140    /proc/sys, and none of the sysctl variables appear to be
141    important performance wise.
142
143    Binary sysctls are a long standing source of subtle kernel
144    bugs and security issues.
145
146    When I looked several months ago all I could find after
147    searching several distributions were 5 user space programs and
148    glibc (which falls back to /proc/sys) using this syscall.
149
150    The man page for sysctl(2) documents it as unusable for user
151    space programs.
152
153    sysctl(2) is not generally ABI compatible to a 32bit user
154    space application on a 64bit and a 32bit kernel.
155
156    For the last several months the policy has been no new binary
157    sysctls and no one has put forward an argument to use them.
158
159    Binary sysctls issues seem to keep happening appearing so
160    properly deprecating them (with a warning to user space) and a
161    2 year grace warning period will mean eventually we can kill
162    them and end the pain.
163
164    In the mean time individual binary sysctls can be dealt with
165    in a piecewise fashion.
166
167Who: Eric Biederman <ebiederm@xmission.com>
168
169---------------------------
170
171What: /proc/<pid>/oom_adj
172When: August 2012
173Why: /proc/<pid>/oom_adj allows userspace to influence the oom killer's
174    badness heuristic used to determine which task to kill when the kernel
175    is out of memory.
176
177    The badness heuristic has since been rewritten since the introduction of
178    this tunable such that its meaning is deprecated. The value was
179    implemented as a bitshift on a score generated by the badness()
180    function that did not have any precise units of measure. With the
181    rewrite, the score is given as a proportion of available memory to the
182    task allocating pages, so using a bitshift which grows the score
183    exponentially is, thus, impossible to tune with fine granularity.
184
185    A much more powerful interface, /proc/<pid>/oom_score_adj, was
186    introduced with the oom killer rewrite that allows users to increase or
187    decrease the badness() score linearly. This interface will replace
188    /proc/<pid>/oom_adj.
189
190    A warning will be emitted to the kernel log if an application uses this
191    deprecated interface. After it is printed once, future warnings will be
192    suppressed until the kernel is rebooted.
193
194---------------------------
195
196What: CS5535/CS5536 obsolete GPIO driver
197When: June 2011
198Files: drivers/staging/cs5535_gpio/*
199Check: drivers/staging/cs5535_gpio/cs5535_gpio.c
200Why: A newer driver replaces this; it is drivers/gpio/cs5535-gpio.c, and
201    integrates with the Linux GPIO subsystem. The old driver has been
202    moved to staging, and will be removed altogether around 2.6.40.
203    Please test the new driver, and ensure that the functionality you
204    need and any bugfixes from the old driver are available in the new
205    one.
206Who: Andres Salomon <dilinger@queued.net>
207
208--------------------------
209
210What: remove EXPORT_SYMBOL(kernel_thread)
211When: August 2006
212Files: arch/*/kernel/*_ksyms.c
213Check: kernel_thread
214Why: kernel_thread is a low-level implementation detail. Drivers should
215        use the <linux/kthread.h> API instead which shields them from
216    implementation details and provides a higherlevel interface that
217    prevents bugs and code duplication
218Who: Christoph Hellwig <hch@lst.de>
219
220---------------------------
221
222What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
223    (temporary transition config option provided until then)
224    The transition config option will also be removed at the same time.
225When: before 2.6.19
226Why: Unused symbols are both increasing the size of the kernel binary
227    and are often a sign of "wrong API"
228Who: Arjan van de Ven <arjan@linux.intel.com>
229
230---------------------------
231
232What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
233When: October 2008
234Why: The stacking of class devices makes these values misleading and
235    inconsistent.
236    Class devices should not carry any of these properties, and bus
237    devices have SUBSYTEM and DRIVER as a replacement.
238Who: Kay Sievers <kay.sievers@suse.de>
239
240---------------------------
241
242What: ACPI procfs interface
243When: July 2008
244Why: ACPI sysfs conversion should be finished by January 2008.
245    ACPI procfs interface will be removed in July 2008 so that
246    there is enough time for the user space to catch up.
247Who: Zhang Rui <rui.zhang@intel.com>
248
249---------------------------
250
251What: CONFIG_ACPI_PROCFS_POWER
252When: 2.6.39
253Why: sysfs I/F for ACPI power devices, including AC and Battery,
254        has been working in upstream kenrel since 2.6.24, Sep 2007.
255    In 2.6.37, we make the sysfs I/F always built in and this option
256    disabled by default.
257    Remove this option and the ACPI power procfs interface in 2.6.39.
258Who: Zhang Rui <rui.zhang@intel.com>
259
260---------------------------
261
262What: /proc/acpi/button
263When: August 2007
264Why: /proc/acpi/button has been replaced by events to the input layer
265    since 2.6.20.
266Who: Len Brown <len.brown@intel.com>
267
268---------------------------
269
270What: /proc/acpi/event
271When: February 2008
272Why: /proc/acpi/event has been replaced by events via the input layer
273    and netlink since 2.6.23.
274Who: Len Brown <len.brown@intel.com>
275
276---------------------------
277
278What: i386/x86_64 bzImage symlinks
279When: April 2010
280
281Why: The i386/x86_64 merge provides a symlink to the old bzImage
282    location so not yet updated user space tools, e.g. package
283    scripts, do not break.
284Who: Thomas Gleixner <tglx@linutronix.de>
285
286---------------------------
287
288What: GPIO autorequest on gpio_direction_{input,output}() in gpiolib
289When: February 2010
290Why: All callers should use explicit gpio_request()/gpio_free().
291    The autorequest mechanism in gpiolib was provided mostly as a
292    migration aid for legacy GPIO interfaces (for SOC based GPIOs).
293    Those users have now largely migrated. Platforms implementing
294    the GPIO interfaces without using gpiolib will see no changes.
295Who: David Brownell <dbrownell@users.sourceforge.net>
296---------------------------
297
298What: b43 support for firmware revision < 410
299When: The schedule was July 2008, but it was decided that we are going to keep the
300        code as long as there are no major maintanance headaches.
301    So it _could_ be removed _any_ time now, if it conflicts with something new.
302Why: The support code for the old firmware hurts code readability/maintainability
303    and slightly hurts runtime performance. Bugfixes for the old firmware
304    are not provided by Broadcom anymore.
305Who: Michael Buesch <mb@bu3sch.de>
306
307---------------------------
308
309What: /sys/o2cb symlink
310When: January 2010
311Why: /sys/fs/o2cb is the proper location for this information - /sys/o2cb
312    exists as a symlink for backwards compatibility for old versions of
313    ocfs2-tools. 2 years should be sufficient time to phase in new versions
314    which know to look in /sys/fs/o2cb.
315Who: ocfs2-devel@oss.oracle.com
316
317---------------------------
318
319What: Ability for non root users to shm_get hugetlb pages based on mlock
320    resource limits
321When: 2.6.31
322Why: Non root users need to be part of /proc/sys/vm/hugetlb_shm_group or
323    have CAP_IPC_LOCK to be able to allocate shm segments backed by
324    huge pages. The mlock based rlimit check to allow shm hugetlb is
325    inconsistent with mmap based allocations. Hence it is being
326    deprecated.
327Who: Ravikiran Thirumalai <kiran@scalex86.org>
328
329---------------------------
330
331What: CONFIG_THERMAL_HWMON
332When: January 2009
333Why: This option was introduced just to allow older lm-sensors userspace
334    to keep working over the upgrade to 2.6.26. At the scheduled time of
335    removal fixed lm-sensors (2.x or 3.x) should be readily available.
336Who: Rene Herman <rene.herman@gmail.com>
337
338---------------------------
339
340What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
341    (in net/core/net-sysfs.c)
342When: After the only user (hal) has seen a release with the patches
343    for enough time, probably some time in 2010.
344Why: Over 1K .text/.data size reduction, data is available in other
345    ways (ioctls)
346Who: Johannes Berg <johannes@sipsolutions.net>
347
348---------------------------
349
350What: sysfs ui for changing p4-clockmod parameters
351When: September 2009
352Why: See commits 129f8ae9b1b5be94517da76009ea956e89104ce8 and
353    e088e4c9cdb618675874becb91b2fd581ee707e6.
354    Removal is subject to fixing any remaining bugs in ACPI which may
355    cause the thermal throttling not to happen at the right time.
356Who: Dave Jones <davej@redhat.com>, Matthew Garrett <mjg@redhat.com>
357
358-----------------------------
359
360What: fakephp and associated sysfs files in /sys/bus/pci/slots/
361When: 2011
362Why: In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to
363    represent a machine's physical PCI slots. The change in semantics
364    had userspace implications, as the hotplug core no longer allowed
365    drivers to create multiple sysfs files per physical slot (required
366    for multi-function devices, e.g.). fakephp was seen as a developer's
367    tool only, and its interface changed. Too late, we learned that
368    there were some users of the fakephp interface.
369
370    In 2.6.30, the original fakephp interface was restored. At the same
371    time, the PCI core gained the ability that fakephp provided, namely
372    function-level hot-remove and hot-add.
373
374    Since the PCI core now provides the same functionality, exposed in:
375
376        /sys/bus/pci/rescan
377        /sys/bus/pci/devices/.../remove
378        /sys/bus/pci/devices/.../rescan
379
380    there is no functional reason to maintain fakephp as well.
381
382    We will keep the existing module so that 'modprobe fakephp' will
383    present the old /sys/bus/pci/slots/... interface for compatibility,
384    but users are urged to migrate their applications to the API above.
385
386    After a reasonable transition period, we will remove the legacy
387    fakephp interface.
388Who: Alex Chiang <achiang@hp.com>
389
390---------------------------
391
392What: CONFIG_RFKILL_INPUT
393When: 2.6.33
394Why: Should be implemented in userspace, policy daemon.
395Who: Johannes Berg <johannes@sipsolutions.net>
396
397----------------------------
398
399What: sound-slot/service-* module aliases and related clutters in
400    sound/sound_core.c
401When: August 2010
402Why: OSS sound_core grabs all legacy minors (0-255) of SOUND_MAJOR
403    (14) and requests modules using custom sound-slot/service-*
404    module aliases. The only benefit of doing this is allowing
405    use of custom module aliases which might as well be considered
406    a bug at this point. This preemptive claiming prevents
407    alternative OSS implementations.
408
409    Till the feature is removed, the kernel will be requesting
410    both sound-slot/service-* and the standard char-major-* module
411    aliases and allow turning off the pre-claiming selectively via
412    CONFIG_SOUND_OSS_CORE_PRECLAIM and soundcore.preclaim_oss
413    kernel parameter.
414
415    After the transition phase is complete, both the custom module
416    aliases and switches to disable it will go away. This removal
417    will also allow making ALSA OSS emulation independent of
418    sound_core. The dependency will be broken then too.
419Who: Tejun Heo <tj@kernel.org>
420
421----------------------------
422
423What: Support for lcd_switch and display_get in asus-laptop driver
424When: March 2010
425Why: These two features use non-standard interfaces. There are the
426    only features that really need multiple path to guess what's
427    the right method name on a specific laptop.
428
429    Removing them will allow to remove a lot of code an significantly
430    clean the drivers.
431
432    This will affect the backlight code which won't be able to know
433    if the backlight is on or off. The platform display file will also be
434    write only (like the one in eeepc-laptop).
435
436    This should'nt affect a lot of user because they usually know
437    when their display is on or off.
438
439Who: Corentin Chary <corentin.chary@gmail.com>
440
441----------------------------
442
443What: sysfs-class-rfkill state file
444When: Feb 2014
445Files: net/rfkill/core.c
446Why: Documented as obsolete since Feb 2010. This file is limited to 3
447    states while the rfkill drivers can have 4 states.
448Who: anybody or Florian Mickler <florian@mickler.org>
449
450----------------------------
451
452What: sysfs-class-rfkill claim file
453When: Feb 2012
454Files: net/rfkill/core.c
455Why: It is not possible to claim an rfkill driver since 2007. This is
456    Documented as obsolete since Feb 2010.
457Who: anybody or Florian Mickler <florian@mickler.org>
458
459----------------------------
460
461What: capifs
462When: February 2011
463Files: drivers/isdn/capi/capifs.*
464Why: udev fully replaces this special file system that only contains CAPI
465    NCCI TTY device nodes. User space (pppdcapiplugin) works without
466    noticing the difference.
467Who: Jan Kiszka <jan.kiszka@web.de>
468
469----------------------------
470
471What: KVM paravirt mmu host support
472When: January 2011
473Why: The paravirt mmu host support is slower than non-paravirt mmu, both
474    on newer and older hardware. It is already not exposed to the guest,
475    and kept only for live migration purposes.
476Who: Avi Kivity <avi@redhat.com>
477
478----------------------------
479
480What: iwlwifi 50XX module parameters
481When: 2.6.40
482Why: The "..50" modules parameters were used to configure 5000 series and
483    up devices; different set of module parameters also available for 4965
484    with same functionalities. Consolidate both set into single place
485    in drivers/net/wireless/iwlwifi/iwl-agn.c
486
487Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
488
489----------------------------
490
491What: iwl4965 alias support
492When: 2.6.40
493Why: Internal alias support has been present in module-init-tools for some
494    time, the MODULE_ALIAS("iwl4965") boilerplate aliases can be removed
495    with no impact.
496
497Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
498
499---------------------------
500
501What: xt_NOTRACK
502Files: net/netfilter/xt_NOTRACK.c
503When: April 2011
504Why: Superseded by xt_CT
505Who: Netfilter developer team <netfilter-devel@vger.kernel.org>
506
507----------------------------
508
509What: IRQF_DISABLED
510When: 2.6.36
511Why: The flag is a NOOP as we run interrupt handlers with interrupts disabled
512Who: Thomas Gleixner <tglx@linutronix.de>
513
514----------------------------
515
516What: The acpi_sleep=s4_nonvs command line option
517When: 2.6.37
518Files: arch/x86/kernel/acpi/sleep.c
519Why: superseded by acpi_sleep=nonvs
520Who: Rafael J. Wysocki <rjw@sisk.pl>
521
522----------------------------
523
524What: PCI DMA unmap state API
525When: August 2012
526Why: PCI DMA unmap state API (include/linux/pci-dma.h) was replaced
527    with DMA unmap state API (DMA unmap state API can be used for
528    any bus).
529Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
530
531----------------------------
532
533What: DMA_xxBIT_MASK macros
534When: Jun 2011
535Why: DMA_xxBIT_MASK macros were replaced with DMA_BIT_MASK() macros.
536Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
537
538----------------------------
539
540What: namespace cgroup (ns_cgroup)
541When: 2.6.38
542Why: The ns_cgroup leads to some problems:
543    * cgroup creation is out-of-control
544    * cgroup name can conflict when pids are looping
545    * it is not possible to have a single process handling
546    a lot of namespaces without falling in a exponential creation time
547    * we may want to create a namespace without creating a cgroup
548
549    The ns_cgroup is replaced by a compatibility flag 'clone_children',
550    where a newly created cgroup will copy the parent cgroup values.
551    The userspace has to manually create a cgroup and add a task to
552    the 'tasks' file.
553Who: Daniel Lezcano <daniel.lezcano@free.fr>
554
555----------------------------
556
557What: iwlwifi disable_hw_scan module parameters
558When: 2.6.40
559Why: Hareware scan is the prefer method for iwlwifi devices for
560    scanning operation. Remove software scan support for all the
561    iwlwifi devices.
562
563Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
564
565----------------------------
566
567What: access to nfsd auth cache through sys_nfsservctl or '.' files
568        in the 'nfsd' filesystem.
569When: 2.6.40
570Why: This is a legacy interface which have been replaced by a more
571        dynamic cache. Continuing to maintain this interface is an
572        unnecessary burden.
573Who: NeilBrown <neilb@suse.de>
574
575----------------------------
576
577What: i2c_adapter.id
578When: June 2011
579Why: This field is deprecated. I2C device drivers shouldn't change their
580    behavior based on the underlying I2C adapter. Instead, the I2C
581    adapter driver should instantiate the I2C devices and provide the
582    needed platform-specific information.
583Who: Jean Delvare <khali@linux-fr.org>
584
585----------------------------
586
587What: cancel_rearming_delayed_work[queue]()
588When: 2.6.39
589
590Why: The functions have been superceded by cancel_delayed_work_sync()
591    quite some time ago. The conversion is trivial and there is no
592    in-kernel user left.
593Who: Tejun Heo <tj@kernel.org>
594
595----------------------------
596
597What: Legacy, non-standard chassis intrusion detection interface.
598When: June 2011
599Why: The adm9240, w83792d and w83793 hardware monitoring drivers have
600    legacy interfaces for chassis intrusion detection. A standard
601    interface has been added to each driver, so the legacy interface
602    can be removed.
603Who: Jean Delvare <khali@linux-fr.org>
604
605----------------------------
606
607What: noswapaccount kernel command line parameter
608When: 2.6.40
609Why: The original implementation of memsw feature enabled by
610    CONFIG_CGROUP_MEM_RES_CTLR_SWAP could be disabled by the noswapaccount
611    kernel parameter (introduced in 2.6.29-rc1). Later on, this decision
612    turned out to be not ideal because we cannot have the feature compiled
613    in and disabled by default and let only interested to enable it
614    (e.g. general distribution kernels might need it). Therefore we have
615    added swapaccount[=0|1] parameter (introduced in 2.6.37) which provides
616    the both possibilities. If we remove noswapaccount we will have
617    less command line parameters with the same functionality and we
618    can also cleanup the parameter handling a bit ().
619Who: Michal Hocko <mhocko@suse.cz>
620
621----------------------------
622

Archive Download this file



interactive