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: x86 floppy disable_hlt
11 Why: ancient workaround of dubious utility clutters the
12 code used by everybody else.
13 Who: Len Brown <len.brown@intel.com>
15 ---------------------------
17 What: CONFIG_APM_CPU_IDLE, and its ability to call APM BIOS in idle
19 Why: This optional sub-feature of APM is of dubious reliability,
20 and ancient APM laptops are likely better served by calling HLT.
21 Deleting CONFIG_APM_CPU_IDLE allows x86 to stop exporting
22 the pm_idle function pointer to modules.
23 Who: Len Brown <len.brown@intel.com>
25 ----------------------------
30 Why: prism54 FullMAC PCI / Cardbus devices used to be supported only by the
31 prism54 wireless driver. After Intersil stopped selling these
32 devices in preference for the newer more flexible SoftMAC devices
33 a SoftMAC device driver was required and prism54 did not support
34 them. The p54pci driver now exists and has been present in the kernel for
35 a while. This driver supports both SoftMAC devices and FullMAC devices.
36 The main difference between these devices was the amount of memory which
37 could be used for the firmware. The SoftMAC devices support a smaller
38 amount of memory. Because of this the SoftMAC firmware fits into FullMAC
39 devices's memory. p54pci supports not only PCI / Cardbus but also USB
40 and SPI. Since p54pci supports all devices prism54 supports
41 you will have a conflict. I'm not quite sure how distributions are
42 handling this conflict right now. prism54 was kept around due to
43 claims users may experience issues when using the SoftMAC driver.
44 Time has passed users have not reported issues. If you use prism54
45 and for whatever reason you cannot use p54pci please let us know!
46 E-mail us at: linux-wireless@vger.kernel.org
48 For more information see the p54 wiki page:
50 http://wireless.kernel.org/en/users/Drivers/p54
52 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
54 ---------------------------
56 What: IRQF_SAMPLE_RANDOM
57 Check: IRQF_SAMPLE_RANDOM
60 Why: Many of IRQF_SAMPLE_RANDOM users are technically bogus as entropy
61 sources in the kernel's current entropy model. To resolve this, every
62 input point to the kernel's entropy pool needs to better document the
63 type of entropy source it actually is. This will be replaced with
64 additional add_*_randomness functions in drivers/char/random.c
66 Who: Robin Getz <rgetz@blackfin.uclinux.org> & Matt Mackall <mpm@selenic.com>
68 ---------------------------
70 What: Deprecated snapshot ioctls
73 Why: The ioctls in kernel/power/user.c were marked as deprecated long time
74 ago. Now they notify users about that so that they need to replace
75 their userspace. After some more time, remove them completely.
77 Who: Jiri Slaby <jirislaby@gmail.com>
79 ---------------------------
81 What: The ieee80211_regdom module parameter
82 When: March 2010 / desktop catchup
84 Why: This was inherited by the CONFIG_WIRELESS_OLD_REGULATORY code,
85 and currently serves as an option for users to define an
86 ISO / IEC 3166 alpha2 code for the country they are currently
87 present in. Although there are userspace API replacements for this
88 through nl80211 distributions haven't yet caught up with implementing
89 decent alternatives through standard GUIs. Although available as an
90 option through iw or wpa_supplicant its just a matter of time before
91 distributions pick up good GUI options for this. The ideal solution
92 would actually consist of intelligent designs which would do this for
93 the user automatically even when travelling through different countries.
94 Until then we leave this module parameter as a compromise.
96 When userspace improves with reasonable widely-available alternatives for
97 this we will no longer need this module parameter. This entry hopes that
98 by the super-futuristically looking date of "March 2010" we will have
99 such replacements widely available.
101 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
103 ---------------------------
105 What: dev->power.power_state
107 Why: Broken design for runtime control over driver power states, confusing
108 driver-internal runtime power management with: mechanisms to support
109 system-wide sleep state transitions; event codes that distinguish
110 different phases of swsusp "sleep" transitions; and userspace policy
111 inputs. This framework was never widely used, and most attempts to
112 use it were broken. Drivers should instead be exposing domain-specific
113 interfaces either to kernel or to userspace.
114 Who: Pavel Machek <pavel@ucw.cz>
116 ---------------------------
118 What: Video4Linux obsolete drivers using V4L1 API
120 Files: drivers/staging/se401/* drivers/staging/usbvideo/*
121 Check: drivers/staging/se401/se401.c drivers/staging/usbvideo/usbvideo.c
122 Why: There are some drivers still using V4L1 API, despite all efforts we've done
123 to migrate. Those drivers are for obsolete hardware that the old maintainer
124 didn't care (or not have the hardware anymore), and that no other developer
125 could find any hardware to buy. They probably have no practical usage today,
126 and people with such old hardware could probably keep using an older version
127 of the kernel. Those drivers will be moved to staging on 2.6.38 and, if nobody
128 cares enough to port and test them with V4L2 API, they'll be removed on 2.6.39.
129 Who: Mauro Carvalho Chehab <mchehab@infradead.org>
131 ---------------------------
133 What: Video4Linux: Remove obsolete ioctl's
135 Files: include/media/videodev2.h
136 Why: Some ioctl's were defined wrong on 2.6.2 and 2.6.6, using the wrong
137 type of R/W arguments. They were fixed, but the old ioctl names are
138 still there, maintained to avoid breaking binary compatibility:
139 #define VIDIOC_OVERLAY_OLD _IOWR('V', 14, int)
140 #define VIDIOC_S_PARM_OLD _IOW('V', 22, struct v4l2_streamparm)
141 #define VIDIOC_S_CTRL_OLD _IOW('V', 28, struct v4l2_control)
142 #define VIDIOC_G_AUDIO_OLD _IOWR('V', 33, struct v4l2_audio)
143 #define VIDIOC_G_AUDOUT_OLD _IOWR('V', 49, struct v4l2_audioout)
144 #define VIDIOC_CROPCAP_OLD _IOR('V', 58, struct v4l2_cropcap)
145 There's no sense on preserving those forever, as it is very doubtful
146 that someone would try to use a such old binary with a modern kernel.
147 Removing them will allow us to remove some magic done at the V4L ioctl
150 Who: Mauro Carvalho Chehab <mchehab@infradead.org>
152 ---------------------------
156 Option: CONFIG_SYSCTL_SYSCALL
157 Why: The same information is available in a more convenient from
158 /proc/sys, and none of the sysctl variables appear to be
159 important performance wise.
161 Binary sysctls are a long standing source of subtle kernel
162 bugs and security issues.
164 When I looked several months ago all I could find after
165 searching several distributions were 5 user space programs and
166 glibc (which falls back to /proc/sys) using this syscall.
168 The man page for sysctl(2) documents it as unusable for user
171 sysctl(2) is not generally ABI compatible to a 32bit user
172 space application on a 64bit and a 32bit kernel.
174 For the last several months the policy has been no new binary
175 sysctls and no one has put forward an argument to use them.
177 Binary sysctls issues seem to keep happening appearing so
178 properly deprecating them (with a warning to user space) and a
179 2 year grace warning period will mean eventually we can kill
180 them and end the pain.
182 In the mean time individual binary sysctls can be dealt with
183 in a piecewise fashion.
185 Who: Eric Biederman <ebiederm@xmission.com>
187 ---------------------------
189 What: /proc/<pid>/oom_adj
191 Why: /proc/<pid>/oom_adj allows userspace to influence the oom killer's
192 badness heuristic used to determine which task to kill when the kernel
195 The badness heuristic has since been rewritten since the introduction of
196 this tunable such that its meaning is deprecated. The value was
197 implemented as a bitshift on a score generated by the badness()
198 function that did not have any precise units of measure. With the
199 rewrite, the score is given as a proportion of available memory to the
200 task allocating pages, so using a bitshift which grows the score
201 exponentially is, thus, impossible to tune with fine granularity.
203 A much more powerful interface, /proc/<pid>/oom_score_adj, was
204 introduced with the oom killer rewrite that allows users to increase or
205 decrease the badness() score linearly. This interface will replace
208 A warning will be emitted to the kernel log if an application uses this
209 deprecated interface. After it is printed once, future warnings will be
210 suppressed until the kernel is rebooted.
212 ---------------------------
214 What: CS5535/CS5536 obsolete GPIO driver
216 Files: drivers/staging/cs5535_gpio/*
217 Check: drivers/staging/cs5535_gpio/cs5535_gpio.c
218 Why: A newer driver replaces this; it is drivers/gpio/cs5535-gpio.c, and
219 integrates with the Linux GPIO subsystem. The old driver has been
220 moved to staging, and will be removed altogether around 2.6.40.
221 Please test the new driver, and ensure that the functionality you
222 need and any bugfixes from the old driver are available in the new
224 Who: Andres Salomon <dilinger@queued.net>
226 --------------------------
228 What: remove EXPORT_SYMBOL(kernel_thread)
230 Files: arch/*/kernel/*_ksyms.c
232 Why: kernel_thread is a low-level implementation detail. Drivers should
233 use the <linux/kthread.h> API instead which shields them from
234 implementation details and provides a higherlevel interface that
235 prevents bugs and code duplication
236 Who: Christoph Hellwig <hch@lst.de>
238 ---------------------------
240 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
241 (temporary transition config option provided until then)
242 The transition config option will also be removed at the same time.
244 Why: Unused symbols are both increasing the size of the kernel binary
245 and are often a sign of "wrong API"
246 Who: Arjan van de Ven <arjan@linux.intel.com>
248 ---------------------------
250 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
252 Why: The stacking of class devices makes these values misleading and
254 Class devices should not carry any of these properties, and bus
255 devices have SUBSYTEM and DRIVER as a replacement.
256 Who: Kay Sievers <kay.sievers@suse.de>
258 ---------------------------
260 What: ACPI procfs interface
262 Why: ACPI sysfs conversion should be finished by January 2008.
263 ACPI procfs interface will be removed in July 2008 so that
264 there is enough time for the user space to catch up.
265 Who: Zhang Rui <rui.zhang@intel.com>
267 ---------------------------
269 What: CONFIG_ACPI_PROCFS_POWER
271 Why: sysfs I/F for ACPI power devices, including AC and Battery,
272 has been working in upstream kenrel since 2.6.24, Sep 2007.
273 In 2.6.37, we make the sysfs I/F always built in and this option
275 Remove this option and the ACPI power procfs interface in 2.6.39.
276 Who: Zhang Rui <rui.zhang@intel.com>
278 ---------------------------
280 What: /proc/acpi/button
282 Why: /proc/acpi/button has been replaced by events to the input layer
284 Who: Len Brown <len.brown@intel.com>
286 ---------------------------
288 What: /proc/acpi/event
290 Why: /proc/acpi/event has been replaced by events via the input layer
291 and netlink since 2.6.23.
292 Who: Len Brown <len.brown@intel.com>
294 ---------------------------
296 What: i386/x86_64 bzImage symlinks
299 Why: The i386/x86_64 merge provides a symlink to the old bzImage
300 location so not yet updated user space tools, e.g. package
301 scripts, do not break.
302 Who: Thomas Gleixner <tglx@linutronix.de>
304 ---------------------------
306 What: GPIO autorequest on gpio_direction_{input,output}() in gpiolib
308 Why: All callers should use explicit gpio_request()/gpio_free().
309 The autorequest mechanism in gpiolib was provided mostly as a
310 migration aid for legacy GPIO interfaces (for SOC based GPIOs).
311 Those users have now largely migrated. Platforms implementing
312 the GPIO interfaces without using gpiolib will see no changes.
313 Who: David Brownell <dbrownell@users.sourceforge.net>
314 ---------------------------
316 What: b43 support for firmware revision < 410
317 When: The schedule was July 2008, but it was decided that we are going to keep the
318 code as long as there are no major maintanance headaches.
319 So it _could_ be removed _any_ time now, if it conflicts with something new.
320 Why: The support code for the old firmware hurts code readability/maintainability
321 and slightly hurts runtime performance. Bugfixes for the old firmware
322 are not provided by Broadcom anymore.
323 Who: Michael Buesch <mb@bu3sch.de>
325 ---------------------------
327 What: /sys/o2cb symlink
329 Why: /sys/fs/o2cb is the proper location for this information - /sys/o2cb
330 exists as a symlink for backwards compatibility for old versions of
331 ocfs2-tools. 2 years should be sufficient time to phase in new versions
332 which know to look in /sys/fs/o2cb.
333 Who: ocfs2-devel@oss.oracle.com
335 ---------------------------
337 What: Ability for non root users to shm_get hugetlb pages based on mlock
340 Why: Non root users need to be part of /proc/sys/vm/hugetlb_shm_group or
341 have CAP_IPC_LOCK to be able to allocate shm segments backed by
342 huge pages. The mlock based rlimit check to allow shm hugetlb is
343 inconsistent with mmap based allocations. Hence it is being
345 Who: Ravikiran Thirumalai <kiran@scalex86.org>
347 ---------------------------
349 What: CONFIG_THERMAL_HWMON
351 Why: This option was introduced just to allow older lm-sensors userspace
352 to keep working over the upgrade to 2.6.26. At the scheduled time of
353 removal fixed lm-sensors (2.x or 3.x) should be readily available.
354 Who: Rene Herman <rene.herman@gmail.com>
356 ---------------------------
358 What: Code that is now under CONFIG_WIRELESS_EXT_SYSFS
359 (in net/core/net-sysfs.c)
360 When: After the only user (hal) has seen a release with the patches
361 for enough time, probably some time in 2010.
362 Why: Over 1K .text/.data size reduction, data is available in other
364 Who: Johannes Berg <johannes@sipsolutions.net>
366 ---------------------------
368 What: sysfs ui for changing p4-clockmod parameters
370 Why: See commits 129f8ae9b1b5be94517da76009ea956e89104ce8 and
371 e088e4c9cdb618675874becb91b2fd581ee707e6.
372 Removal is subject to fixing any remaining bugs in ACPI which may
373 cause the thermal throttling not to happen at the right time.
374 Who: Dave Jones <davej@redhat.com>, Matthew Garrett <mjg@redhat.com>
376 -----------------------------
378 What: fakephp and associated sysfs files in /sys/bus/pci/slots/
380 Why: In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to
381 represent a machine's physical PCI slots. The change in semantics
382 had userspace implications, as the hotplug core no longer allowed
383 drivers to create multiple sysfs files per physical slot (required
384 for multi-function devices, e.g.). fakephp was seen as a developer's
385 tool only, and its interface changed. Too late, we learned that
386 there were some users of the fakephp interface.
388 In 2.6.30, the original fakephp interface was restored. At the same
389 time, the PCI core gained the ability that fakephp provided, namely
390 function-level hot-remove and hot-add.
392 Since the PCI core now provides the same functionality, exposed in:
395 /sys/bus/pci/devices/.../remove
396 /sys/bus/pci/devices/.../rescan
398 there is no functional reason to maintain fakephp as well.
400 We will keep the existing module so that 'modprobe fakephp' will
401 present the old /sys/bus/pci/slots/... interface for compatibility,
402 but users are urged to migrate their applications to the API above.
404 After a reasonable transition period, we will remove the legacy
406 Who: Alex Chiang <achiang@hp.com>
408 ---------------------------
410 What: CONFIG_RFKILL_INPUT
412 Why: Should be implemented in userspace, policy daemon.
413 Who: Johannes Berg <johannes@sipsolutions.net>
415 ----------------------------
417 What: sound-slot/service-* module aliases and related clutters in
420 Why: OSS sound_core grabs all legacy minors (0-255) of SOUND_MAJOR
421 (14) and requests modules using custom sound-slot/service-*
422 module aliases. The only benefit of doing this is allowing
423 use of custom module aliases which might as well be considered
424 a bug at this point. This preemptive claiming prevents
425 alternative OSS implementations.
427 Till the feature is removed, the kernel will be requesting
428 both sound-slot/service-* and the standard char-major-* module
429 aliases and allow turning off the pre-claiming selectively via
430 CONFIG_SOUND_OSS_CORE_PRECLAIM and soundcore.preclaim_oss
433 After the transition phase is complete, both the custom module
434 aliases and switches to disable it will go away. This removal
435 will also allow making ALSA OSS emulation independent of
436 sound_core. The dependency will be broken then too.
437 Who: Tejun Heo <tj@kernel.org>
439 ----------------------------
441 What: Support for lcd_switch and display_get in asus-laptop driver
443 Why: These two features use non-standard interfaces. There are the
444 only features that really need multiple path to guess what's
445 the right method name on a specific laptop.
447 Removing them will allow to remove a lot of code an significantly
450 This will affect the backlight code which won't be able to know
451 if the backlight is on or off. The platform display file will also be
452 write only (like the one in eeepc-laptop).
454 This should'nt affect a lot of user because they usually know
455 when their display is on or off.
457 Who: Corentin Chary <corentin.chary@gmail.com>
459 ----------------------------
461 What: sysfs-class-rfkill state file
463 Files: net/rfkill/core.c
464 Why: Documented as obsolete since Feb 2010. This file is limited to 3
465 states while the rfkill drivers can have 4 states.
466 Who: anybody or Florian Mickler <florian@mickler.org>
468 ----------------------------
470 What: sysfs-class-rfkill claim file
472 Files: net/rfkill/core.c
473 Why: It is not possible to claim an rfkill driver since 2007. This is
474 Documented as obsolete since Feb 2010.
475 Who: anybody or Florian Mickler <florian@mickler.org>
477 ----------------------------
481 Files: drivers/isdn/capi/capifs.*
482 Why: udev fully replaces this special file system that only contains CAPI
483 NCCI TTY device nodes. User space (pppdcapiplugin) works without
484 noticing the difference.
485 Who: Jan Kiszka <jan.kiszka@web.de>
487 ----------------------------
489 What: KVM paravirt mmu host support
491 Why: The paravirt mmu host support is slower than non-paravirt mmu, both
492 on newer and older hardware. It is already not exposed to the guest,
493 and kept only for live migration purposes.
494 Who: Avi Kivity <avi@redhat.com>
496 ----------------------------
498 What: iwlwifi 50XX module parameters
500 Why: The "..50" modules parameters were used to configure 5000 series and
501 up devices; different set of module parameters also available for 4965
502 with same functionalities. Consolidate both set into single place
503 in drivers/net/wireless/iwlwifi/iwl-agn.c
505 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
507 ----------------------------
509 What: iwl4965 alias support
511 Why: Internal alias support has been present in module-init-tools for some
512 time, the MODULE_ALIAS("iwl4965") boilerplate aliases can be removed
515 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
517 ---------------------------
520 Files: net/netfilter/xt_NOTRACK.c
522 Why: Superseded by xt_CT
523 Who: Netfilter developer team <netfilter-devel@vger.kernel.org>
525 ----------------------------
529 Why: The flag is a NOOP as we run interrupt handlers with interrupts disabled
530 Who: Thomas Gleixner <tglx@linutronix.de>
532 ----------------------------
534 What: The acpi_sleep=s4_nonvs command line option
536 Files: arch/x86/kernel/acpi/sleep.c
537 Why: superseded by acpi_sleep=nonvs
538 Who: Rafael J. Wysocki <rjw@sisk.pl>
540 ----------------------------
542 What: PCI DMA unmap state API
544 Why: PCI DMA unmap state API (include/linux/pci-dma.h) was replaced
545 with DMA unmap state API (DMA unmap state API can be used for
547 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
549 ----------------------------
551 What: DMA_xxBIT_MASK macros
553 Why: DMA_xxBIT_MASK macros were replaced with DMA_BIT_MASK() macros.
554 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
556 ----------------------------
558 What: namespace cgroup (ns_cgroup)
560 Why: The ns_cgroup leads to some problems:
561 * cgroup creation is out-of-control
562 * cgroup name can conflict when pids are looping
563 * it is not possible to have a single process handling
564 a lot of namespaces without falling in a exponential creation time
565 * we may want to create a namespace without creating a cgroup
567 The ns_cgroup is replaced by a compatibility flag 'clone_children',
568 where a newly created cgroup will copy the parent cgroup values.
569 The userspace has to manually create a cgroup and add a task to
571 Who: Daniel Lezcano <daniel.lezcano@free.fr>
573 ----------------------------
575 What: iwlwifi disable_hw_scan module parameters
577 Why: Hareware scan is the prefer method for iwlwifi devices for
578 scanning operation. Remove software scan support for all the
581 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
583 ----------------------------
585 What: access to nfsd auth cache through sys_nfsservctl or '.' files
586 in the 'nfsd' filesystem.
588 Why: This is a legacy interface which have been replaced by a more
589 dynamic cache. Continuing to maintain this interface is an
591 Who: NeilBrown <neilb@suse.de>
593 ----------------------------
597 Why: This field is deprecated. I2C device drivers shouldn't change their
598 behavior based on the underlying I2C adapter. Instead, the I2C
599 adapter driver should instantiate the I2C devices and provide the
600 needed platform-specific information.
601 Who: Jean Delvare <khali@linux-fr.org>
603 ----------------------------
605 What: cancel_rearming_delayed_work[queue]()
608 Why: The functions have been superceded by cancel_delayed_work_sync()
609 quite some time ago. The conversion is trivial and there is no
611 Who: Tejun Heo <tj@kernel.org>
613 ----------------------------
615 What: Legacy, non-standard chassis intrusion detection interface.
617 Why: The adm9240, w83792d and w83793 hardware monitoring drivers have
618 legacy interfaces for chassis intrusion detection. A standard
619 interface has been added to each driver, so the legacy interface
621 Who: Jean Delvare <khali@linux-fr.org>
623 ----------------------------
625 What: noswapaccount kernel command line parameter
627 Why: The original implementation of memsw feature enabled by
628 CONFIG_CGROUP_MEM_RES_CTLR_SWAP could be disabled by the noswapaccount
629 kernel parameter (introduced in 2.6.29-rc1). Later on, this decision
630 turned out to be not ideal because we cannot have the feature compiled
631 in and disabled by default and let only interested to enable it
632 (e.g. general distribution kernels might need it). Therefore we have
633 added swapaccount[=0|1] parameter (introduced in 2.6.37) which provides
634 the both possibilities. If we remove noswapaccount we will have
635 less command line parameters with the same functionality and we
636 can also cleanup the parameter handling a bit ().
637 Who: Michal Hocko <mhocko@suse.cz>
639 ----------------------------