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: CONFIG_APM_CPU_IDLE, and its ability to call APM BIOS in idle
11 Why: This optional sub-feature of APM is of dubious reliability,
12 and ancient APM laptops are likely better served by calling HLT.
13 Deleting CONFIG_APM_CPU_IDLE allows x86 to stop exporting
14 the pm_idle function pointer to modules.
15 Who: Len Brown <len.brown@intel.com>
17 ----------------------------
19 What: x86_32 "no-hlt" cmdline param
21 Why: remove a branch from idle path, simplify code used by everybody.
22 This option disabled the use of HLT in idle and machine_halt()
23 for hardware that was flakey 15-years ago. Today we have
24 "idle=poll" that removed HLT from idle, and so if such a machine
25 is still running the upstream kernel, "idle=poll" is likely sufficient.
26 Who: Len Brown <len.brown@intel.com>
28 ----------------------------
30 What: x86 "idle=mwait" cmdline param
32 Why: simplify x86 idle code
33 Who: Len Brown <len.brown@intel.com>
35 ----------------------------
40 Why: prism54 FullMAC PCI / Cardbus devices used to be supported only by the
41 prism54 wireless driver. After Intersil stopped selling these
42 devices in preference for the newer more flexible SoftMAC devices
43 a SoftMAC device driver was required and prism54 did not support
44 them. The p54pci driver now exists and has been present in the kernel for
45 a while. This driver supports both SoftMAC devices and FullMAC devices.
46 The main difference between these devices was the amount of memory which
47 could be used for the firmware. The SoftMAC devices support a smaller
48 amount of memory. Because of this the SoftMAC firmware fits into FullMAC
49 devices's memory. p54pci supports not only PCI / Cardbus but also USB
50 and SPI. Since p54pci supports all devices prism54 supports
51 you will have a conflict. I'm not quite sure how distributions are
52 handling this conflict right now. prism54 was kept around due to
53 claims users may experience issues when using the SoftMAC driver.
54 Time has passed users have not reported issues. If you use prism54
55 and for whatever reason you cannot use p54pci please let us know!
56 E-mail us at: linux-wireless@vger.kernel.org
58 For more information see the p54 wiki page:
60 http://wireless.kernel.org/en/users/Drivers/p54
62 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
64 ---------------------------
66 What: IRQF_SAMPLE_RANDOM
67 Check: IRQF_SAMPLE_RANDOM
70 Why: Many of IRQF_SAMPLE_RANDOM users are technically bogus as entropy
71 sources in the kernel's current entropy model. To resolve this, every
72 input point to the kernel's entropy pool needs to better document the
73 type of entropy source it actually is. This will be replaced with
74 additional add_*_randomness functions in drivers/char/random.c
76 Who: Robin Getz <rgetz@blackfin.uclinux.org> & Matt Mackall <mpm@selenic.com>
78 ---------------------------
80 What: Deprecated snapshot ioctls
83 Why: The ioctls in kernel/power/user.c were marked as deprecated long time
84 ago. Now they notify users about that so that they need to replace
85 their userspace. After some more time, remove them completely.
87 Who: Jiri Slaby <jirislaby@gmail.com>
89 ---------------------------
91 What: The ieee80211_regdom module parameter
92 When: March 2010 / desktop catchup
94 Why: This was inherited by the CONFIG_WIRELESS_OLD_REGULATORY code,
95 and currently serves as an option for users to define an
96 ISO / IEC 3166 alpha2 code for the country they are currently
97 present in. Although there are userspace API replacements for this
98 through nl80211 distributions haven't yet caught up with implementing
99 decent alternatives through standard GUIs. Although available as an
100 option through iw or wpa_supplicant its just a matter of time before
101 distributions pick up good GUI options for this. The ideal solution
102 would actually consist of intelligent designs which would do this for
103 the user automatically even when travelling through different countries.
104 Until then we leave this module parameter as a compromise.
106 When userspace improves with reasonable widely-available alternatives for
107 this we will no longer need this module parameter. This entry hopes that
108 by the super-futuristically looking date of "March 2010" we will have
109 such replacements widely available.
111 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
113 ---------------------------
115 What: dev->power.power_state
117 Why: Broken design for runtime control over driver power states, confusing
118 driver-internal runtime power management with: mechanisms to support
119 system-wide sleep state transitions; event codes that distinguish
120 different phases of swsusp "sleep" transitions; and userspace policy
121 inputs. This framework was never widely used, and most attempts to
122 use it were broken. Drivers should instead be exposing domain-specific
123 interfaces either to kernel or to userspace.
124 Who: Pavel Machek <pavel@ucw.cz>
126 ---------------------------
130 Option: CONFIG_SYSCTL_SYSCALL
131 Why: The same information is available in a more convenient from
132 /proc/sys, and none of the sysctl variables appear to be
133 important performance wise.
135 Binary sysctls are a long standing source of subtle kernel
136 bugs and security issues.
138 When I looked several months ago all I could find after
139 searching several distributions were 5 user space programs and
140 glibc (which falls back to /proc/sys) using this syscall.
142 The man page for sysctl(2) documents it as unusable for user
145 sysctl(2) is not generally ABI compatible to a 32bit user
146 space application on a 64bit and a 32bit kernel.
148 For the last several months the policy has been no new binary
149 sysctls and no one has put forward an argument to use them.
151 Binary sysctls issues seem to keep happening appearing so
152 properly deprecating them (with a warning to user space) and a
153 2 year grace warning period will mean eventually we can kill
154 them and end the pain.
156 In the mean time individual binary sysctls can be dealt with
157 in a piecewise fashion.
159 Who: Eric Biederman <ebiederm@xmission.com>
161 ---------------------------
163 What: /proc/<pid>/oom_adj
165 Why: /proc/<pid>/oom_adj allows userspace to influence the oom killer's
166 badness heuristic used to determine which task to kill when the kernel
169 The badness heuristic has since been rewritten since the introduction of
170 this tunable such that its meaning is deprecated. The value was
171 implemented as a bitshift on a score generated by the badness()
172 function that did not have any precise units of measure. With the
173 rewrite, the score is given as a proportion of available memory to the
174 task allocating pages, so using a bitshift which grows the score
175 exponentially is, thus, impossible to tune with fine granularity.
177 A much more powerful interface, /proc/<pid>/oom_score_adj, was
178 introduced with the oom killer rewrite that allows users to increase or
179 decrease the badness() score linearly. This interface will replace
182 A warning will be emitted to the kernel log if an application uses this
183 deprecated interface. After it is printed once, future warnings will be
184 suppressed until the kernel is rebooted.
186 ---------------------------
188 What: CS5535/CS5536 obsolete GPIO driver
190 Files: drivers/staging/cs5535_gpio/*
191 Check: drivers/staging/cs5535_gpio/cs5535_gpio.c
192 Why: A newer driver replaces this; it is drivers/gpio/cs5535-gpio.c, and
193 integrates with the Linux GPIO subsystem. The old driver has been
194 moved to staging, and will be removed altogether around 2.6.40.
195 Please test the new driver, and ensure that the functionality you
196 need and any bugfixes from the old driver are available in the new
198 Who: Andres Salomon <dilinger@queued.net>
200 --------------------------
202 What: remove EXPORT_SYMBOL(kernel_thread)
204 Files: arch/*/kernel/*_ksyms.c
206 Why: kernel_thread is a low-level implementation detail. Drivers should
207 use the <linux/kthread.h> API instead which shields them from
208 implementation details and provides a higherlevel interface that
209 prevents bugs and code duplication
210 Who: Christoph Hellwig <hch@lst.de>
212 ---------------------------
214 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
215 (temporary transition config option provided until then)
216 The transition config option will also be removed at the same time.
218 Why: Unused symbols are both increasing the size of the kernel binary
219 and are often a sign of "wrong API"
220 Who: Arjan van de Ven <arjan@linux.intel.com>
222 ---------------------------
224 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
226 Why: The stacking of class devices makes these values misleading and
228 Class devices should not carry any of these properties, and bus
229 devices have SUBSYTEM and DRIVER as a replacement.
230 Who: Kay Sievers <kay.sievers@suse.de>
232 ---------------------------
234 What: ACPI procfs interface
236 Why: ACPI sysfs conversion should be finished by January 2008.
237 ACPI procfs interface will be removed in July 2008 so that
238 there is enough time for the user space to catch up.
239 Who: Zhang Rui <rui.zhang@intel.com>
241 ---------------------------
243 What: CONFIG_ACPI_PROCFS_POWER
245 Why: sysfs I/F for ACPI power devices, including AC and Battery,
246 has been working in upstream kernel since 2.6.24, Sep 2007.
247 In 2.6.37, we make the sysfs I/F always built in and this option
249 Remove this option and the ACPI power procfs interface in 2.6.39.
250 Who: Zhang Rui <rui.zhang@intel.com>
252 ---------------------------
254 What: /proc/acpi/event
256 Why: /proc/acpi/event has been replaced by events via the input layer
257 and netlink since 2.6.23.
258 Who: Len Brown <len.brown@intel.com>
260 ---------------------------
262 What: i386/x86_64 bzImage symlinks
265 Why: The i386/x86_64 merge provides a symlink to the old bzImage
266 location so not yet updated user space tools, e.g. package
267 scripts, do not break.
268 Who: Thomas Gleixner <tglx@linutronix.de>
270 ---------------------------
272 What: GPIO autorequest on gpio_direction_{input,output}() in gpiolib
274 Why: All callers should use explicit gpio_request()/gpio_free().
275 The autorequest mechanism in gpiolib was provided mostly as a
276 migration aid for legacy GPIO interfaces (for SOC based GPIOs).
277 Those users have now largely migrated. Platforms implementing
278 the GPIO interfaces without using gpiolib will see no changes.
279 Who: David Brownell <dbrownell@users.sourceforge.net>
280 ---------------------------
282 What: b43 support for firmware revision < 410
283 When: The schedule was July 2008, but it was decided that we are going to keep the
284 code as long as there are no major maintanance headaches.
285 So it _could_ be removed _any_ time now, if it conflicts with something new.
286 Why: The support code for the old firmware hurts code readability/maintainability
287 and slightly hurts runtime performance. Bugfixes for the old firmware
288 are not provided by Broadcom anymore.
289 Who: Michael Buesch <mb@bu3sch.de>
291 ---------------------------
293 What: Ability for non root users to shm_get hugetlb pages based on mlock
296 Why: Non root users need to be part of /proc/sys/vm/hugetlb_shm_group or
297 have CAP_IPC_LOCK to be able to allocate shm segments backed by
298 huge pages. The mlock based rlimit check to allow shm hugetlb is
299 inconsistent with mmap based allocations. Hence it is being
301 Who: Ravikiran Thirumalai <kiran@scalex86.org>
303 ---------------------------
305 What: CONFIG_THERMAL_HWMON
307 Why: This option was introduced just to allow older lm-sensors userspace
308 to keep working over the upgrade to 2.6.26. At the scheduled time of
309 removal fixed lm-sensors (2.x or 3.x) should be readily available.
310 Who: Rene Herman <rene.herman@gmail.com>
312 ---------------------------
314 What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
315 (in net/core/net-sysfs.c)
316 When: After the only user (hal) has seen a release with the patches
317 for enough time, probably some time in 2010.
318 Why: Over 1K .text/.data size reduction, data is available in other
320 Who: Johannes Berg <johannes@sipsolutions.net>
322 ---------------------------
324 What: sysfs ui for changing p4-clockmod parameters
326 Why: See commits 129f8ae9b1b5be94517da76009ea956e89104ce8 and
327 e088e4c9cdb618675874becb91b2fd581ee707e6.
328 Removal is subject to fixing any remaining bugs in ACPI which may
329 cause the thermal throttling not to happen at the right time.
330 Who: Dave Jones <davej@redhat.com>, Matthew Garrett <mjg@redhat.com>
332 -----------------------------
334 What: fakephp and associated sysfs files in /sys/bus/pci/slots/
336 Why: In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to
337 represent a machine's physical PCI slots. The change in semantics
338 had userspace implications, as the hotplug core no longer allowed
339 drivers to create multiple sysfs files per physical slot (required
340 for multi-function devices, e.g.). fakephp was seen as a developer's
341 tool only, and its interface changed. Too late, we learned that
342 there were some users of the fakephp interface.
344 In 2.6.30, the original fakephp interface was restored. At the same
345 time, the PCI core gained the ability that fakephp provided, namely
346 function-level hot-remove and hot-add.
348 Since the PCI core now provides the same functionality, exposed in:
351 /sys/bus/pci/devices/.../remove
352 /sys/bus/pci/devices/.../rescan
354 there is no functional reason to maintain fakephp as well.
356 We will keep the existing module so that 'modprobe fakephp' will
357 present the old /sys/bus/pci/slots/... interface for compatibility,
358 but users are urged to migrate their applications to the API above.
360 After a reasonable transition period, we will remove the legacy
362 Who: Alex Chiang <achiang@hp.com>
364 ---------------------------
366 What: CONFIG_RFKILL_INPUT
368 Why: Should be implemented in userspace, policy daemon.
369 Who: Johannes Berg <johannes@sipsolutions.net>
371 ----------------------------
373 What: sound-slot/service-* module aliases and related clutters in
376 Why: OSS sound_core grabs all legacy minors (0-255) of SOUND_MAJOR
377 (14) and requests modules using custom sound-slot/service-*
378 module aliases. The only benefit of doing this is allowing
379 use of custom module aliases which might as well be considered
380 a bug at this point. This preemptive claiming prevents
381 alternative OSS implementations.
383 Till the feature is removed, the kernel will be requesting
384 both sound-slot/service-* and the standard char-major-* module
385 aliases and allow turning off the pre-claiming selectively via
386 CONFIG_SOUND_OSS_CORE_PRECLAIM and soundcore.preclaim_oss
389 After the transition phase is complete, both the custom module
390 aliases and switches to disable it will go away. This removal
391 will also allow making ALSA OSS emulation independent of
392 sound_core. The dependency will be broken then too.
393 Who: Tejun Heo <tj@kernel.org>
395 ----------------------------
397 What: sysfs-class-rfkill state file
399 Files: net/rfkill/core.c
400 Why: Documented as obsolete since Feb 2010. This file is limited to 3
401 states while the rfkill drivers can have 4 states.
402 Who: anybody or Florian Mickler <florian@mickler.org>
404 ----------------------------
406 What: sysfs-class-rfkill claim file
408 Files: net/rfkill/core.c
409 Why: It is not possible to claim an rfkill driver since 2007. This is
410 Documented as obsolete since Feb 2010.
411 Who: anybody or Florian Mickler <florian@mickler.org>
413 ----------------------------
415 What: KVM paravirt mmu host support
417 Why: The paravirt mmu host support is slower than non-paravirt mmu, both
418 on newer and older hardware. It is already not exposed to the guest,
419 and kept only for live migration purposes.
420 Who: Avi Kivity <avi@redhat.com>
422 ----------------------------
424 What: iwlwifi 50XX module parameters
426 Why: The "..50" modules parameters were used to configure 5000 series and
427 up devices; different set of module parameters also available for 4965
428 with same functionalities. Consolidate both set into single place
429 in drivers/net/wireless/iwlwifi/iwl-agn.c
431 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
433 ----------------------------
435 What: iwl4965 alias support
437 Why: Internal alias support has been present in module-init-tools for some
438 time, the MODULE_ALIAS("iwl4965") boilerplate aliases can be removed
441 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
443 ---------------------------
446 Files: net/netfilter/xt_NOTRACK.c
448 Why: Superseded by xt_CT
449 Who: Netfilter developer team <netfilter-devel@vger.kernel.org>
451 ----------------------------
455 Why: The flag is a NOOP as we run interrupt handlers with interrupts disabled
456 Who: Thomas Gleixner <tglx@linutronix.de>
458 ----------------------------
460 What: PCI DMA unmap state API
462 Why: PCI DMA unmap state API (include/linux/pci-dma.h) was replaced
463 with DMA unmap state API (DMA unmap state API can be used for
465 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
467 ----------------------------
469 What: DMA_xxBIT_MASK macros
471 Why: DMA_xxBIT_MASK macros were replaced with DMA_BIT_MASK() macros.
472 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
474 ----------------------------
476 What: iwlwifi disable_hw_scan module parameters
478 Why: Hareware scan is the prefer method for iwlwifi devices for
479 scanning operation. Remove software scan support for all the
482 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
484 ----------------------------
486 What: access to nfsd auth cache through sys_nfsservctl or '.' files
487 in the 'nfsd' filesystem.
489 Why: This is a legacy interface which have been replaced by a more
490 dynamic cache. Continuing to maintain this interface is an
492 Who: NeilBrown <neilb@suse.de>
494 ----------------------------
496 What: cancel_rearming_delayed_work[queue]()
499 Why: The functions have been superceded by cancel_delayed_work_sync()
500 quite some time ago. The conversion is trivial and there is no
502 Who: Tejun Heo <tj@kernel.org>
504 ----------------------------
506 What: Legacy, non-standard chassis intrusion detection interface.
508 Why: The adm9240, w83792d and w83793 hardware monitoring drivers have
509 legacy interfaces for chassis intrusion detection. A standard
510 interface has been added to each driver, so the legacy interface
512 Who: Jean Delvare <khali@linux-fr.org>
514 ----------------------------
516 What: xt_connlimit rev 0
518 Who: Jan Engelhardt <jengelh@medozas.de>
519 Files: net/netfilter/xt_connlimit.c
521 ----------------------------
523 What: noswapaccount kernel command line parameter
525 Why: The original implementation of memsw feature enabled by
526 CONFIG_CGROUP_MEM_RES_CTLR_SWAP could be disabled by the noswapaccount
527 kernel parameter (introduced in 2.6.29-rc1). Later on, this decision
528 turned out to be not ideal because we cannot have the feature compiled
529 in and disabled by default and let only interested to enable it
530 (e.g. general distribution kernels might need it). Therefore we have
531 added swapaccount[=0|1] parameter (introduced in 2.6.37) which provides
532 the both possibilities. If we remove noswapaccount we will have
533 less command line parameters with the same functionality and we
534 can also cleanup the parameter handling a bit ().
535 Who: Michal Hocko <mhocko@suse.cz>
537 ----------------------------
539 What: ipt_addrtype match include file
541 Why: superseded by xt_addrtype
542 Who: Florian Westphal <fw@strlen.de>
543 Files: include/linux/netfilter_ipv4/ipt_addrtype.h
545 ----------------------------
547 What: i2c_driver.attach_adapter
548 i2c_driver.detach_adapter
550 Why: These legacy callbacks should no longer be used as i2c-core offers
551 a variety of preferable alternative ways to instantiate I2C devices.
552 Who: Jean Delvare <khali@linux-fr.org>
554 ----------------------------
556 What: Support for UVCIOC_CTRL_ADD in the uvcvideo driver
558 Why: The information passed to the driver by this ioctl is now queried
559 dynamically from the device.
560 Who: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
562 ----------------------------
564 What: Support for UVCIOC_CTRL_MAP_OLD in the uvcvideo driver
566 Why: Used only by applications compiled against older driver versions.
567 Superseded by UVCIOC_CTRL_MAP which supports V4L2 menu controls.
568 Who: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
570 ----------------------------
572 What: Support for UVCIOC_CTRL_GET and UVCIOC_CTRL_SET in the uvcvideo driver
574 Why: Superseded by the UVCIOC_CTRL_QUERY ioctl.
575 Who: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
577 ----------------------------
579 What: For VIDIOC_S_FREQUENCY the type field must match the device node's type.
580 If not, return -EINVAL.
582 Why: It makes no sense to switch the tuner to radio mode by calling
583 VIDIOC_S_FREQUENCY on a video node, or to switch the tuner to tv mode by
584 calling VIDIOC_S_FREQUENCY on a radio node. This is the first step of a
585 move to more consistent handling of tv and radio tuners.
586 Who: Hans Verkuil <hans.verkuil@cisco.com>
588 ----------------------------
590 What: Opening a radio device node will no longer automatically switch the
591 tuner mode from tv to radio.
593 Why: Just opening a V4L device should not change the state of the hardware
594 like that. It's very unexpected and against the V4L spec. Instead, you
595 switch to radio mode by calling VIDIOC_S_FREQUENCY. This is the second
596 and last step of the move to consistent handling of tv and radio tuners.
597 Who: Hans Verkuil <hans.verkuil@cisco.com>
599 ----------------------------