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 ---------------------------
12 Why: 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
30 For more information see the p54 wiki page:
32 http://wireless.kernel.org/en/users/Drivers/p54
34 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
36 ---------------------------
41 Why: This driver is deprecated and the firmware is no longer
42 maintained. The replacement driver "carl9170" has been
43 around for a while, so the devices are still supported.
45 Who: Christian Lamparter <chunkeey@googlemail.com>
47 ---------------------------
49 What: IRQF_SAMPLE_RANDOM
50 Check: IRQF_SAMPLE_RANDOM
53 Why: Many of IRQF_SAMPLE_RANDOM users are technically bogus as entropy
54 sources in the kernel's current entropy model. To resolve this, every
55 input point to the kernel's entropy pool needs to better document the
56 type of entropy source it actually is. This will be replaced with
57 additional add_*_randomness functions in drivers/char/random.c
59 Who: Robin Getz <rgetz@blackfin.uclinux.org> & Matt Mackall <mpm@selenic.com>
61 ---------------------------
63 What: Deprecated snapshot ioctls
66 Why: The ioctls in kernel/power/user.c were marked as deprecated long time
67 ago. Now they notify users about that so that they need to replace
68 their userspace. After some more time, remove them completely.
70 Who: Jiri Slaby <jirislaby@gmail.com>
72 ---------------------------
74 What: The ieee80211_regdom module parameter
75 When: March 2010 / desktop catchup
77 Why: This was inherited by the CONFIG_WIRELESS_OLD_REGULATORY code,
78 and currently serves as an option for users to define an
79 ISO / IEC 3166 alpha2 code for the country they are currently
80 present in. Although there are userspace API replacements for this
81 through nl80211 distributions haven't yet caught up with implementing
82 decent alternatives through standard GUIs. Although available as an
83 option through iw or wpa_supplicant its just a matter of time before
84 distributions pick up good GUI options for this. The ideal solution
85 would actually consist of intelligent designs which would do this for
86 the user automatically even when travelling through different countries.
87 Until then we leave this module parameter as a compromise.
89 When userspace improves with reasonable widely-available alternatives for
90 this we will no longer need this module parameter. This entry hopes that
91 by the super-futuristically looking date of "March 2010" we will have
92 such replacements widely available.
94 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
96 ---------------------------
98 What: dev->power.power_state
100 Why: Broken design for runtime control over driver power states, confusing
101 driver-internal runtime power management with: mechanisms to support
102 system-wide sleep state transitions; event codes that distinguish
103 different phases of swsusp "sleep" transitions; and userspace policy
104 inputs. This framework was never widely used, and most attempts to
105 use it were broken. Drivers should instead be exposing domain-specific
106 interfaces either to kernel or to userspace.
107 Who: Pavel Machek <pavel@ucw.cz>
109 ---------------------------
113 Option: CONFIG_SYSCTL_SYSCALL
114 Why: The same information is available in a more convenient from
115 /proc/sys, and none of the sysctl variables appear to be
116 important performance wise.
118 Binary sysctls are a long standing source of subtle kernel
119 bugs and security issues.
121 When I looked several months ago all I could find after
122 searching several distributions were 5 user space programs and
123 glibc (which falls back to /proc/sys) using this syscall.
125 The man page for sysctl(2) documents it as unusable for user
128 sysctl(2) is not generally ABI compatible to a 32bit user
129 space application on a 64bit and a 32bit kernel.
131 For the last several months the policy has been no new binary
132 sysctls and no one has put forward an argument to use them.
134 Binary sysctls issues seem to keep happening appearing so
135 properly deprecating them (with a warning to user space) and a
136 2 year grace warning period will mean eventually we can kill
137 them and end the pain.
139 In the mean time individual binary sysctls can be dealt with
140 in a piecewise fashion.
142 Who: Eric Biederman <ebiederm@xmission.com>
144 ---------------------------
146 What: /proc/<pid>/oom_adj
148 Why: /proc/<pid>/oom_adj allows userspace to influence the oom killer's
149 badness heuristic used to determine which task to kill when the kernel
152 The badness heuristic has since been rewritten since the introduction of
153 this tunable such that its meaning is deprecated. The value was
154 implemented as a bitshift on a score generated by the badness()
155 function that did not have any precise units of measure. With the
156 rewrite, the score is given as a proportion of available memory to the
157 task allocating pages, so using a bitshift which grows the score
158 exponentially is, thus, impossible to tune with fine granularity.
160 A much more powerful interface, /proc/<pid>/oom_score_adj, was
161 introduced with the oom killer rewrite that allows users to increase or
162 decrease the badness() score linearly. This interface will replace
165 A warning will be emitted to the kernel log if an application uses this
166 deprecated interface. After it is printed once, future warnings will be
167 suppressed until the kernel is rebooted.
169 ---------------------------
171 What: CS5535/CS5536 obsolete GPIO driver
173 Files: drivers/staging/cs5535_gpio/*
174 Check: drivers/staging/cs5535_gpio/cs5535_gpio.c
175 Why: A newer driver replaces this; it is drivers/gpio/cs5535-gpio.c, and
176 integrates with the Linux GPIO subsystem. The old driver has been
177 moved to staging, and will be removed altogether around 2.6.40.
178 Please test the new driver, and ensure that the functionality you
179 need and any bugfixes from the old driver are available in the new
181 Who: Andres Salomon <dilinger@queued.net>
183 --------------------------
185 What: remove EXPORT_SYMBOL(kernel_thread)
187 Files: arch/*/kernel/*_ksyms.c
189 Why: kernel_thread is a low-level implementation detail. Drivers should
190 use the <linux/kthread.h> API instead which shields them from
191 implementation details and provides a higherlevel interface that
192 prevents bugs and code duplication
193 Who: Christoph Hellwig <hch@lst.de>
195 ---------------------------
197 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
198 (temporary transition config option provided until then)
199 The transition config option will also be removed at the same time.
201 Why: Unused symbols are both increasing the size of the kernel binary
202 and are often a sign of "wrong API"
203 Who: Arjan van de Ven <arjan@linux.intel.com>
205 ---------------------------
207 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
209 Why: The stacking of class devices makes these values misleading and
211 Class devices should not carry any of these properties, and bus
212 devices have SUBSYTEM and DRIVER as a replacement.
213 Who: Kay Sievers <kay.sievers@suse.de>
215 ---------------------------
217 What: ACPI procfs interface
219 Why: ACPI sysfs conversion should be finished by January 2008.
220 ACPI procfs interface will be removed in July 2008 so that
221 there is enough time for the user space to catch up.
222 Who: Zhang Rui <rui.zhang@intel.com>
224 ---------------------------
226 What: CONFIG_ACPI_PROCFS_POWER
228 Why: sysfs I/F for ACPI power devices, including AC and Battery,
229 has been working in upstream kenrel since 2.6.24, Sep 2007.
230 In 2.6.37, we make the sysfs I/F always built in and this option
232 Remove this option and the ACPI power procfs interface in 2.6.39.
233 Who: Zhang Rui <rui.zhang@intel.com>
235 ---------------------------
237 What: /proc/acpi/event
239 Why: /proc/acpi/event has been replaced by events via the input layer
240 and netlink since 2.6.23.
241 Who: Len Brown <len.brown@intel.com>
243 ---------------------------
245 What: i386/x86_64 bzImage symlinks
248 Why: The i386/x86_64 merge provides a symlink to the old bzImage
249 location so not yet updated user space tools, e.g. package
250 scripts, do not break.
251 Who: Thomas Gleixner <tglx@linutronix.de>
253 ---------------------------
255 What: GPIO autorequest on gpio_direction_{input,output}() in gpiolib
257 Why: All callers should use explicit gpio_request()/gpio_free().
258 The autorequest mechanism in gpiolib was provided mostly as a
259 migration aid for legacy GPIO interfaces (for SOC based GPIOs).
260 Those users have now largely migrated. Platforms implementing
261 the GPIO interfaces without using gpiolib will see no changes.
262 Who: David Brownell <dbrownell@users.sourceforge.net>
263 ---------------------------
265 What: b43 support for firmware revision < 410
266 When: The schedule was July 2008, but it was decided that we are going to keep the
267 code as long as there are no major maintanance headaches.
268 So it _could_ be removed _any_ time now, if it conflicts with something new.
269 Why: The support code for the old firmware hurts code readability/maintainability
270 and slightly hurts runtime performance. Bugfixes for the old firmware
271 are not provided by Broadcom anymore.
272 Who: Michael Buesch <mb@bu3sch.de>
274 ---------------------------
276 What: /sys/o2cb symlink
278 Why: /sys/fs/o2cb is the proper location for this information - /sys/o2cb
279 exists as a symlink for backwards compatibility for old versions of
280 ocfs2-tools. 2 years should be sufficient time to phase in new versions
281 which know to look in /sys/fs/o2cb.
282 Who: ocfs2-devel@oss.oracle.com
284 ---------------------------
286 What: Ability for non root users to shm_get hugetlb pages based on mlock
289 Why: Non root users need to be part of /proc/sys/vm/hugetlb_shm_group or
290 have CAP_IPC_LOCK to be able to allocate shm segments backed by
291 huge pages. The mlock based rlimit check to allow shm hugetlb is
292 inconsistent with mmap based allocations. Hence it is being
294 Who: Ravikiran Thirumalai <kiran@scalex86.org>
296 ---------------------------
298 What: CONFIG_THERMAL_HWMON
300 Why: This option was introduced just to allow older lm-sensors userspace
301 to keep working over the upgrade to 2.6.26. At the scheduled time of
302 removal fixed lm-sensors (2.x or 3.x) should be readily available.
303 Who: Rene Herman <rene.herman@gmail.com>
305 ---------------------------
307 What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
308 (in net/core/net-sysfs.c)
309 When: After the only user (hal) has seen a release with the patches
310 for enough time, probably some time in 2010.
311 Why: Over 1K .text/.data size reduction, data is available in other
313 Who: Johannes Berg <johannes@sipsolutions.net>
315 ---------------------------
317 What: sysfs ui for changing p4-clockmod parameters
319 Why: See commits 129f8ae9b1b5be94517da76009ea956e89104ce8 and
320 e088e4c9cdb618675874becb91b2fd581ee707e6.
321 Removal is subject to fixing any remaining bugs in ACPI which may
322 cause the thermal throttling not to happen at the right time.
323 Who: Dave Jones <davej@redhat.com>, Matthew Garrett <mjg@redhat.com>
325 -----------------------------
327 What: fakephp and associated sysfs files in /sys/bus/pci/slots/
329 Why: In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to
330 represent a machine's physical PCI slots. The change in semantics
331 had userspace implications, as the hotplug core no longer allowed
332 drivers to create multiple sysfs files per physical slot (required
333 for multi-function devices, e.g.). fakephp was seen as a developer's
334 tool only, and its interface changed. Too late, we learned that
335 there were some users of the fakephp interface.
337 In 2.6.30, the original fakephp interface was restored. At the same
338 time, the PCI core gained the ability that fakephp provided, namely
339 function-level hot-remove and hot-add.
341 Since the PCI core now provides the same functionality, exposed in:
344 /sys/bus/pci/devices/.../remove
345 /sys/bus/pci/devices/.../rescan
347 there is no functional reason to maintain fakephp as well.
349 We will keep the existing module so that 'modprobe fakephp' will
350 present the old /sys/bus/pci/slots/... interface for compatibility,
351 but users are urged to migrate their applications to the API above.
353 After a reasonable transition period, we will remove the legacy
355 Who: Alex Chiang <achiang@hp.com>
357 ---------------------------
359 What: CONFIG_RFKILL_INPUT
361 Why: Should be implemented in userspace, policy daemon.
362 Who: Johannes Berg <johannes@sipsolutions.net>
364 ----------------------------
366 What: sound-slot/service-* module aliases and related clutters in
369 Why: OSS sound_core grabs all legacy minors (0-255) of SOUND_MAJOR
370 (14) and requests modules using custom sound-slot/service-*
371 module aliases. The only benefit of doing this is allowing
372 use of custom module aliases which might as well be considered
373 a bug at this point. This preemptive claiming prevents
374 alternative OSS implementations.
376 Till the feature is removed, the kernel will be requesting
377 both sound-slot/service-* and the standard char-major-* module
378 aliases and allow turning off the pre-claiming selectively via
379 CONFIG_SOUND_OSS_CORE_PRECLAIM and soundcore.preclaim_oss
382 After the transition phase is complete, both the custom module
383 aliases and switches to disable it will go away. This removal
384 will also allow making ALSA OSS emulation independent of
385 sound_core. The dependency will be broken then too.
386 Who: Tejun Heo <tj@kernel.org>
388 ----------------------------
390 What: sysfs-class-rfkill state file
392 Files: net/rfkill/core.c
393 Why: Documented as obsolete since Feb 2010. This file is limited to 3
394 states while the rfkill drivers can have 4 states.
395 Who: anybody or Florian Mickler <florian@mickler.org>
397 ----------------------------
399 What: sysfs-class-rfkill claim file
401 Files: net/rfkill/core.c
402 Why: It is not possible to claim an rfkill driver since 2007. This is
403 Documented as obsolete since Feb 2010.
404 Who: anybody or Florian Mickler <florian@mickler.org>
406 ----------------------------
410 Files: drivers/isdn/capi/capifs.*
411 Why: udev fully replaces this special file system that only contains CAPI
412 NCCI TTY device nodes. User space (pppdcapiplugin) works without
413 noticing the difference.
414 Who: Jan Kiszka <jan.kiszka@web.de>
416 ----------------------------
418 What: KVM paravirt mmu host support
420 Why: The paravirt mmu host support is slower than non-paravirt mmu, both
421 on newer and older hardware. It is already not exposed to the guest,
422 and kept only for live migration purposes.
423 Who: Avi Kivity <avi@redhat.com>
425 ----------------------------
427 What: iwlwifi 50XX module parameters
429 Why: The "..50" modules parameters were used to configure 5000 series and
430 up devices; different set of module parameters also available for 4965
431 with same functionalities. Consolidate both set into single place
432 in drivers/net/wireless/iwlwifi/iwl-agn.c
434 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
436 ----------------------------
438 What: iwl4965 alias support
440 Why: Internal alias support has been present in module-init-tools for some
441 time, the MODULE_ALIAS("iwl4965") boilerplate aliases can be removed
444 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
446 ---------------------------
449 Files: net/netfilter/xt_NOTRACK.c
451 Why: Superseded by xt_CT
452 Who: Netfilter developer team <netfilter-devel@vger.kernel.org>
454 ----------------------------
458 Why: The flag is a NOOP as we run interrupt handlers with interrupts disabled
459 Who: Thomas Gleixner <tglx@linutronix.de>
461 ----------------------------
463 What: The acpi_sleep=s4_nonvs command line option
465 Files: arch/x86/kernel/acpi/sleep.c
466 Why: superseded by acpi_sleep=nonvs
467 Who: Rafael J. Wysocki <rjw@sisk.pl>
469 ----------------------------
471 What: PCI DMA unmap state API
473 Why: PCI DMA unmap state API (include/linux/pci-dma.h) was replaced
474 with DMA unmap state API (DMA unmap state API can be used for
476 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
478 ----------------------------
480 What: DMA_xxBIT_MASK macros
482 Why: DMA_xxBIT_MASK macros were replaced with DMA_BIT_MASK() macros.
483 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
485 ----------------------------
487 What: namespace cgroup (ns_cgroup)
489 Why: The ns_cgroup leads to some problems:
490 * cgroup creation is out-of-control
491 * cgroup name can conflict when pids are looping
492 * it is not possible to have a single process handling
493 a lot of namespaces without falling in a exponential creation time
494 * we may want to create a namespace without creating a cgroup
496 The ns_cgroup is replaced by a compatibility flag 'clone_children',
497 where a newly created cgroup will copy the parent cgroup values.
498 The userspace has to manually create a cgroup and add a task to
500 Who: Daniel Lezcano <daniel.lezcano@free.fr>
502 ----------------------------
504 What: iwlwifi disable_hw_scan module parameters
506 Why: Hareware scan is the prefer method for iwlwifi devices for
507 scanning operation. Remove software scan support for all the
510 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
512 ----------------------------
514 What: access to nfsd auth cache through sys_nfsservctl or '.' files
515 in the 'nfsd' filesystem.
517 Why: This is a legacy interface which have been replaced by a more
518 dynamic cache. Continuing to maintain this interface is an
520 Who: NeilBrown <neilb@suse.de>
522 ----------------------------
524 What: cancel_rearming_delayed_work[queue]()
527 Why: The functions have been superceded by cancel_delayed_work_sync()
528 quite some time ago. The conversion is trivial and there is no
530 Who: Tejun Heo <tj@kernel.org>
532 ----------------------------
534 What: Legacy, non-standard chassis intrusion detection interface.
536 Why: The adm9240, w83792d and w83793 hardware monitoring drivers have
537 legacy interfaces for chassis intrusion detection. A standard
538 interface has been added to each driver, so the legacy interface
540 Who: Jean Delvare <khali@linux-fr.org>
542 ----------------------------
544 What: xt_connlimit rev 0
546 Who: Jan Engelhardt <jengelh@medozas.de>
547 Files: net/netfilter/xt_connlimit.c
549 ----------------------------
551 What: noswapaccount kernel command line parameter
553 Why: The original implementation of memsw feature enabled by
554 CONFIG_CGROUP_MEM_RES_CTLR_SWAP could be disabled by the noswapaccount
555 kernel parameter (introduced in 2.6.29-rc1). Later on, this decision
556 turned out to be not ideal because we cannot have the feature compiled
557 in and disabled by default and let only interested to enable it
558 (e.g. general distribution kernels might need it). Therefore we have
559 added swapaccount[=0|1] parameter (introduced in 2.6.37) which provides
560 the both possibilities. If we remove noswapaccount we will have
561 less command line parameters with the same functionality and we
562 can also cleanup the parameter handling a bit ().
563 Who: Michal Hocko <mhocko@suse.cz>
565 ----------------------------
567 What: ipt_addrtype match include file
569 Why: superseded by xt_addrtype
570 Who: Florian Westphal <fw@strlen.de>
571 Files: include/linux/netfilter_ipv4/ipt_addrtype.h
573 ----------------------------
575 What: i2c_driver.attach_adapter
576 i2c_driver.detach_adapter
578 Why: These legacy callbacks should no longer be used as i2c-core offers
579 a variety of preferable alternative ways to instantiate I2C devices.
580 Who: Jean Delvare <khali@linux-fr.org>
582 ----------------------------