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. The suggested deprecation period is 3 releases.
7 ---------------------------
9 What: ddebug_query="query" boot cmdline param
11 Why: obsoleted by dyndbg="query" and module.dyndbg="query"
12 Who: Jim Cromie <jim.cromie@gmail.com>, Jason Baron <jbaron@redhat.com>
14 ---------------------------
16 What: CONFIG_APM_CPU_IDLE, and its ability to call APM BIOS in idle
18 Why: This optional sub-feature of APM is of dubious reliability,
19 and ancient APM laptops are likely better served by calling HLT.
20 Deleting CONFIG_APM_CPU_IDLE allows x86 to stop exporting
21 the pm_idle function pointer to modules.
22 Who: Len Brown <len.brown@intel.com>
24 ----------------------------
26 What: x86_32 "no-hlt" cmdline param
28 Why: remove a branch from idle path, simplify code used by everybody.
29 This option disabled the use of HLT in idle and machine_halt()
30 for hardware that was flakey 15-years ago. Today we have
31 "idle=poll" that removed HLT from idle, and so if such a machine
32 is still running the upstream kernel, "idle=poll" is likely sufficient.
33 Who: Len Brown <len.brown@intel.com>
35 ----------------------------
37 What: x86 "idle=mwait" cmdline param
39 Why: simplify x86 idle code
40 Who: Len Brown <len.brown@intel.com>
42 ----------------------------
47 Why: prism54 FullMAC PCI / Cardbus devices used to be supported only by the
48 prism54 wireless driver. After Intersil stopped selling these
49 devices in preference for the newer more flexible SoftMAC devices
50 a SoftMAC device driver was required and prism54 did not support
51 them. The p54pci driver now exists and has been present in the kernel for
52 a while. This driver supports both SoftMAC devices and FullMAC devices.
53 The main difference between these devices was the amount of memory which
54 could be used for the firmware. The SoftMAC devices support a smaller
55 amount of memory. Because of this the SoftMAC firmware fits into FullMAC
56 devices's memory. p54pci supports not only PCI / Cardbus but also USB
57 and SPI. Since p54pci supports all devices prism54 supports
58 you will have a conflict. I'm not quite sure how distributions are
59 handling this conflict right now. prism54 was kept around due to
60 claims users may experience issues when using the SoftMAC driver.
61 Time has passed users have not reported issues. If you use prism54
62 and for whatever reason you cannot use p54pci please let us know!
63 E-mail us at: linux-wireless@vger.kernel.org
65 For more information see the p54 wiki page:
67 http://wireless.kernel.org/en/users/Drivers/p54
69 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
71 ---------------------------
73 What: The ieee80211_regdom module parameter
74 When: March 2010 / desktop catchup
76 Why: This was inherited by the CONFIG_WIRELESS_OLD_REGULATORY code,
77 and currently serves as an option for users to define an
78 ISO / IEC 3166 alpha2 code for the country they are currently
79 present in. Although there are userspace API replacements for this
80 through nl80211 distributions haven't yet caught up with implementing
81 decent alternatives through standard GUIs. Although available as an
82 option through iw or wpa_supplicant its just a matter of time before
83 distributions pick up good GUI options for this. The ideal solution
84 would actually consist of intelligent designs which would do this for
85 the user automatically even when travelling through different countries.
86 Until then we leave this module parameter as a compromise.
88 When userspace improves with reasonable widely-available alternatives for
89 this we will no longer need this module parameter. This entry hopes that
90 by the super-futuristically looking date of "March 2010" we will have
91 such replacements widely available.
93 Who: Luis R. Rodriguez <lrodriguez@atheros.com>
95 ---------------------------
97 What: dev->power.power_state
99 Why: Broken design for runtime control over driver power states, confusing
100 driver-internal runtime power management with: mechanisms to support
101 system-wide sleep state transitions; event codes that distinguish
102 different phases of swsusp "sleep" transitions; and userspace policy
103 inputs. This framework was never widely used, and most attempts to
104 use it were broken. Drivers should instead be exposing domain-specific
105 interfaces either to kernel or to userspace.
106 Who: Pavel Machek <pavel@ucw.cz>
108 ---------------------------
110 What: /proc/<pid>/oom_adj
112 Why: /proc/<pid>/oom_adj allows userspace to influence the oom killer's
113 badness heuristic used to determine which task to kill when the kernel
116 The badness heuristic has since been rewritten since the introduction of
117 this tunable such that its meaning is deprecated. The value was
118 implemented as a bitshift on a score generated by the badness()
119 function that did not have any precise units of measure. With the
120 rewrite, the score is given as a proportion of available memory to the
121 task allocating pages, so using a bitshift which grows the score
122 exponentially is, thus, impossible to tune with fine granularity.
124 A much more powerful interface, /proc/<pid>/oom_score_adj, was
125 introduced with the oom killer rewrite that allows users to increase or
126 decrease the badness score linearly. This interface will replace
129 A warning will be emitted to the kernel log if an application uses this
130 deprecated interface. After it is printed once, future warnings will be
131 suppressed until the kernel is rebooted.
133 ---------------------------
135 What: remove EXPORT_SYMBOL(kernel_thread)
137 Files: arch/*/kernel/*_ksyms.c
139 Why: kernel_thread is a low-level implementation detail. Drivers should
140 use the <linux/kthread.h> API instead which shields them from
141 implementation details and provides a higherlevel interface that
142 prevents bugs and code duplication
143 Who: Christoph Hellwig <hch@lst.de>
145 ---------------------------
147 What: Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
148 (temporary transition config option provided until then)
149 The transition config option will also be removed at the same time.
151 Why: Unused symbols are both increasing the size of the kernel binary
152 and are often a sign of "wrong API"
153 Who: Arjan van de Ven <arjan@linux.intel.com>
155 ---------------------------
157 What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
159 Why: The stacking of class devices makes these values misleading and
161 Class devices should not carry any of these properties, and bus
162 devices have SUBSYTEM and DRIVER as a replacement.
163 Who: Kay Sievers <kay.sievers@suse.de>
165 ---------------------------
167 What: ACPI procfs interface
169 Why: ACPI sysfs conversion should be finished by January 2008.
170 ACPI procfs interface will be removed in July 2008 so that
171 there is enough time for the user space to catch up.
172 Who: Zhang Rui <rui.zhang@intel.com>
174 ---------------------------
176 What: CONFIG_ACPI_PROCFS_POWER
178 Why: sysfs I/F for ACPI power devices, including AC and Battery,
179 has been working in upstream kernel since 2.6.24, Sep 2007.
180 In 2.6.37, we make the sysfs I/F always built in and this option
182 Remove this option and the ACPI power procfs interface in 2.6.39.
183 Who: Zhang Rui <rui.zhang@intel.com>
185 ---------------------------
187 What: /proc/acpi/event
189 Why: /proc/acpi/event has been replaced by events via the input layer
190 and netlink since 2.6.23.
191 Who: Len Brown <len.brown@intel.com>
193 ---------------------------
195 What: i386/x86_64 bzImage symlinks
198 Why: The i386/x86_64 merge provides a symlink to the old bzImage
199 location so not yet updated user space tools, e.g. package
200 scripts, do not break.
201 Who: Thomas Gleixner <tglx@linutronix.de>
203 ---------------------------
205 What: GPIO autorequest on gpio_direction_{input,output}() in gpiolib
207 Why: All callers should use explicit gpio_request()/gpio_free().
208 The autorequest mechanism in gpiolib was provided mostly as a
209 migration aid for legacy GPIO interfaces (for SOC based GPIOs).
210 Those users have now largely migrated. Platforms implementing
211 the GPIO interfaces without using gpiolib will see no changes.
212 Who: David Brownell <dbrownell@users.sourceforge.net>
213 ---------------------------
215 What: b43 support for firmware revision < 410
216 When: The schedule was July 2008, but it was decided that we are going to keep the
217 code as long as there are no major maintanance headaches.
218 So it _could_ be removed _any_ time now, if it conflicts with something new.
219 Why: The support code for the old firmware hurts code readability/maintainability
220 and slightly hurts runtime performance. Bugfixes for the old firmware
221 are not provided by Broadcom anymore.
222 Who: Michael Buesch <m@bues.ch>
224 ---------------------------
226 What: Ability for non root users to shm_get hugetlb pages based on mlock
229 Why: Non root users need to be part of /proc/sys/vm/hugetlb_shm_group or
230 have CAP_IPC_LOCK to be able to allocate shm segments backed by
231 huge pages. The mlock based rlimit check to allow shm hugetlb is
232 inconsistent with mmap based allocations. Hence it is being
234 Who: Ravikiran Thirumalai <kiran@scalex86.org>
236 ---------------------------
238 What: sysfs ui for changing p4-clockmod parameters
240 Why: See commits 129f8ae9b1b5be94517da76009ea956e89104ce8 and
241 e088e4c9cdb618675874becb91b2fd581ee707e6.
242 Removal is subject to fixing any remaining bugs in ACPI which may
243 cause the thermal throttling not to happen at the right time.
244 Who: Dave Jones <davej@redhat.com>, Matthew Garrett <mjg@redhat.com>
246 -----------------------------
248 What: fakephp and associated sysfs files in /sys/bus/pci/slots/
250 Why: In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to
251 represent a machine's physical PCI slots. The change in semantics
252 had userspace implications, as the hotplug core no longer allowed
253 drivers to create multiple sysfs files per physical slot (required
254 for multi-function devices, e.g.). fakephp was seen as a developer's
255 tool only, and its interface changed. Too late, we learned that
256 there were some users of the fakephp interface.
258 In 2.6.30, the original fakephp interface was restored. At the same
259 time, the PCI core gained the ability that fakephp provided, namely
260 function-level hot-remove and hot-add.
262 Since the PCI core now provides the same functionality, exposed in:
265 /sys/bus/pci/devices/.../remove
266 /sys/bus/pci/devices/.../rescan
268 there is no functional reason to maintain fakephp as well.
270 We will keep the existing module so that 'modprobe fakephp' will
271 present the old /sys/bus/pci/slots/... interface for compatibility,
272 but users are urged to migrate their applications to the API above.
274 After a reasonable transition period, we will remove the legacy
276 Who: Alex Chiang <achiang@hp.com>
278 ---------------------------
280 What: CONFIG_RFKILL_INPUT
282 Why: Should be implemented in userspace, policy daemon.
283 Who: Johannes Berg <johannes@sipsolutions.net>
285 ----------------------------
287 What: sound-slot/service-* module aliases and related clutters in
290 Why: OSS sound_core grabs all legacy minors (0-255) of SOUND_MAJOR
291 (14) and requests modules using custom sound-slot/service-*
292 module aliases. The only benefit of doing this is allowing
293 use of custom module aliases which might as well be considered
294 a bug at this point. This preemptive claiming prevents
295 alternative OSS implementations.
297 Till the feature is removed, the kernel will be requesting
298 both sound-slot/service-* and the standard char-major-* module
299 aliases and allow turning off the pre-claiming selectively via
300 CONFIG_SOUND_OSS_CORE_PRECLAIM and soundcore.preclaim_oss
303 After the transition phase is complete, both the custom module
304 aliases and switches to disable it will go away. This removal
305 will also allow making ALSA OSS emulation independent of
306 sound_core. The dependency will be broken then too.
307 Who: Tejun Heo <tj@kernel.org>
309 ----------------------------
311 What: sysfs-class-rfkill state file
313 Files: net/rfkill/core.c
314 Why: Documented as obsolete since Feb 2010. This file is limited to 3
315 states while the rfkill drivers can have 4 states.
316 Who: anybody or Florian Mickler <florian@mickler.org>
318 ----------------------------
320 What: sysfs-class-rfkill claim file
322 Files: net/rfkill/core.c
323 Why: It is not possible to claim an rfkill driver since 2007. This is
324 Documented as obsolete since Feb 2010.
325 Who: anybody or Florian Mickler <florian@mickler.org>
327 ----------------------------
329 What: iwlwifi 50XX module parameters
331 Why: The "..50" modules parameters were used to configure 5000 series and
332 up devices; different set of module parameters also available for 4965
333 with same functionalities. Consolidate both set into single place
334 in drivers/net/wireless/iwlwifi/iwl-agn.c
336 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
338 ----------------------------
340 What: iwl4965 alias support
342 Why: Internal alias support has been present in module-init-tools for some
343 time, the MODULE_ALIAS("iwl4965") boilerplate aliases can be removed
346 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
348 ---------------------------
351 Files: net/netfilter/xt_NOTRACK.c
353 Why: Superseded by xt_CT
354 Who: Netfilter developer team <netfilter-devel@vger.kernel.org>
356 ----------------------------
360 Why: The flag is a NOOP as we run interrupt handlers with interrupts disabled
361 Who: Thomas Gleixner <tglx@linutronix.de>
363 ----------------------------
365 What: PCI DMA unmap state API
367 Why: PCI DMA unmap state API (include/linux/pci-dma.h) was replaced
368 with DMA unmap state API (DMA unmap state API can be used for
370 Who: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
372 ----------------------------
374 What: iwlwifi disable_hw_scan module parameters
376 Why: Hareware scan is the prefer method for iwlwifi devices for
377 scanning operation. Remove software scan support for all the
380 Who: Wey-Yi Guy <wey-yi.w.guy@intel.com>
382 ----------------------------
384 What: Legacy, non-standard chassis intrusion detection interface.
386 Why: The adm9240, w83792d and w83793 hardware monitoring drivers have
387 legacy interfaces for chassis intrusion detection. A standard
388 interface has been added to each driver, so the legacy interface
390 Who: Jean Delvare <khali@linux-fr.org>
392 ----------------------------
394 What: i2c_driver.attach_adapter
395 i2c_driver.detach_adapter
397 Why: These legacy callbacks should no longer be used as i2c-core offers
398 a variety of preferable alternative ways to instantiate I2C devices.
399 Who: Jean Delvare <khali@linux-fr.org>
401 ----------------------------
403 What: Opening a radio device node will no longer automatically switch the
404 tuner mode from tv to radio.
406 Why: Just opening a V4L device should not change the state of the hardware
407 like that. It's very unexpected and against the V4L spec. Instead, you
408 switch to radio mode by calling VIDIOC_S_FREQUENCY. This is the second
409 and last step of the move to consistent handling of tv and radio tuners.
410 Who: Hans Verkuil <hans.verkuil@cisco.com>
412 ----------------------------
414 What: CONFIG_CFG80211_WEXT
415 When: as soon as distributions ship new wireless tools, ie. wpa_supplicant 1.0
416 and NetworkManager/connman/etc. that are able to use nl80211
417 Why: Wireless extensions are deprecated, and userland tools are moving to
418 using nl80211. New drivers are no longer using wireless extensions,
419 and while there might still be old drivers, both new drivers and new
420 userland no longer needs them and they can't be used for an feature
421 developed in the past couple of years. As such, compatibility with
422 wireless extensions in new drivers will be removed.
423 Who: Johannes Berg <johannes@sipsolutions.net>
425 ----------------------------
427 What: g_file_storage driver
429 Why: This driver has been superseded by g_mass_storage.
430 Who: Alan Stern <stern@rowland.harvard.edu>
432 ----------------------------
434 What: threeg and interface sysfs files in /sys/devices/platform/acer-wmi
436 Why: In 3.0, we can now autodetect internal 3G device and already have
437 the threeg rfkill device. So, we plan to remove threeg sysfs support
438 for it's no longer necessary.
440 We also plan to remove interface sysfs file that exposed which ACPI-WMI
441 interface that was used by acer-wmi driver. It will replaced by
442 information log when acer-wmi initial.
443 Who: Lee, Chun-Yi <jlee@novell.com>
445 ---------------------------
447 What: /sys/devices/platform/_UDC_/udc/_UDC_/is_dualspeed file and
448 is_dualspeed line in /sys/devices/platform/ci13xxx_*/udc/device file.
450 Why: The is_dualspeed file is superseded by maximum_speed in the same
451 directory and is_dualspeed line in device file is superseded by
452 max_speed line in the same file.
454 The maximum_speed/max_speed specifies maximum speed supported by UDC.
455 To check if dualspeeed is supported, check if the value is >= 3.
456 Various possible speeds are defined in <linux/usb/ch9.h>.
457 Who: Michal Nazarewicz <mina86@mina86.com>
459 ----------------------------
461 What: The XFS nodelaylog mount option
463 Why: The delaylog mode that has been the default since 2.6.39 has proven
464 stable, and the old code is in the way of additional improvements in
466 Who: Christoph Hellwig <hch@lst.de>
468 ----------------------------
470 What: iwlagn alias support
472 Why: The iwlagn module has been renamed iwlwifi. The alias will be around
473 for backward compatibility for several cycles and then dropped.
474 Who: Don Fry <donald.h.fry@intel.com>
476 ----------------------------
478 What: pci_scan_bus_parented()
480 Why: The pci_scan_bus_parented() interface creates a new root bus. The
481 bus is created with default resources (ioport_resource and
482 iomem_resource) that are always wrong, so we rely on arch code to
483 correct them later. Callers of pci_scan_bus_parented() should
484 convert to using pci_scan_root_bus() so they can supply a list of
485 bus resources when the bus is created.
486 Who: Bjorn Helgaas <bhelgaas@google.com>
488 ----------------------------
490 What: Low Performance USB Block driver ("CONFIG_BLK_DEV_UB")
492 Why: This driver provides support for USB storage devices like "USB
493 sticks". As of now, it is deactivated in Debian, Fedora and
494 Ubuntu. All current users can switch over to usb-storage
495 (CONFIG_USB_STORAGE) which only drawback is the additional SCSI
497 Who: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
499 ----------------------------
501 What: get_robust_list syscall
503 Why: There appear to be no production users of the get_robust_list syscall,
504 and it runs the risk of leaking address locations, allowing the bypass
505 of ASLR. It was only ever intended for debugging, so it should be
507 Who: Kees Cook <keescook@chromium.org>
509 ----------------------------
511 What: Removing the pn544 raw driver.
513 Why: With the introduction of the NFC HCI and SHDL kernel layers, pn544.c
514 is being replaced by pn544_hci.c which is accessible through the netlink
515 and socket NFC APIs. Moreover, pn544.c is outdated and does not seem to
516 work properly with the latest Android stacks.
517 Having 2 drivers for the same hardware is confusing and as such we
518 should only keep the one following the kernel NFC APIs.
519 Who: Samuel Ortiz <sameo@linux.intel.com>
521 ----------------------------
523 What: setitimer accepts user NULL pointer (value)
525 Why: setitimer is not returning -EFAULT if user pointer is NULL. This
527 Who: Sasikantha Babu <sasikanth.v19@gmail.com>
529 ----------------------------
531 What: remove bogus DV presets V4L2_DV_1080I29_97, V4L2_DV_1080I30 and
534 Why: These HDTV formats do not exist and were added by a confused mind
535 (that was me, to be precise...)
536 Who: Hans Verkuil <hans.verkuil@cisco.com>
538 ----------------------------
540 What: V4L2_CID_HCENTER, V4L2_CID_VCENTER V4L2 controls
542 Why: The V4L2_CID_VCENTER, V4L2_CID_HCENTER controls have been deprecated
543 for about 4 years and they are not used by any mainline driver.
544 There are newer controls (V4L2_CID_PAN*, V4L2_CID_TILT*) that provide
545 similar functionality.
546 Who: Sylwester Nawrocki <sylvester.nawrocki@gmail.com>
548 ----------------------------
550 What: cgroup option updates via remount
552 Why: Remount currently allows changing bound subsystems and
553 release_agent. Rebinding is hardly useful as it only works
554 when the hierarchy is empty and release_agent itself should be
555 replaced with conventional fsnotify.
557 ----------------------------
559 What: xt_recent rev 0
561 Who: Pablo Neira Ayuso <pablo@netfilter.org>
562 Files: net/netfilter/xt_recent.c
564 ----------------------------
566 What: KVM debugfs statistics
568 Why: KVM tracepoints provide mostly equivalent information in a much more
571 ----------------------------
573 What: at91-mci driver ("CONFIG_MMC_AT91")
575 Why: There are two mci drivers: at91-mci and atmel-mci. The PDC support
576 was added to atmel-mci as a first step to support more chips.
577 Then at91-mci was kept only for old IP versions (on at91rm9200 and
578 at91sam9261). The support of these IP versions has just been added
579 to atmel-mci, so atmel-mci can be used for all chips.
580 Who: Ludovic Desroches <ludovic.desroches@atmel.com>
582 ----------------------------
586 Why: Unsupported/unmaintained/unused since 2.6
588 ----------------------------
590 What: V4L2 selections API target rectangle and flags unification, the
591 following definitions will be removed: V4L2_SEL_TGT_CROP_ACTIVE,
592 V4L2_SEL_TGT_COMPOSE_ACTIVE, V4L2_SUBDEV_SEL_*, V4L2_SUBDEV_SEL_FLAG_*
593 in favor of common V4L2_SEL_TGT_* and V4L2_SEL_FLAG_* definitions.
594 For more details see include/linux/v4l2-common.h.
596 Why: The regular V4L2 selections and the subdev selection API originally
597 defined distinct names for the target rectangles and flags - V4L2_SEL_*
598 and V4L2_SUBDEV_SEL_*. Although, it turned out that the meaning of these
599 target rectangles is virtually identical and the APIs were consolidated
600 to use single set of names - V4L2_SEL_*. This didn't involve any ABI
601 changes. Alias definitions were created for the original ones to avoid
602 any instabilities in the user space interface. After few cycles these
603 backward compatibility definitions will be removed.
604 Who: Sylwester Nawrocki <sylvester.nawrocki@gmail.com>
606 ----------------------------
608 What: Using V4L2_CAP_VIDEO_CAPTURE and V4L2_CAP_VIDEO_OUTPUT flags
609 to indicate a V4L2 memory-to-memory device capability
611 Why: New drivers should use new V4L2_CAP_VIDEO_M2M capability flag
612 to indicate a V4L2 video memory-to-memory (M2M) device and
613 applications can now identify a M2M video device by checking
614 for V4L2_CAP_VIDEO_M2M, with VIDIOC_QUERYCAP ioctl. Using ORed
615 V4L2_CAP_VIDEO_CAPTURE and V4L2_CAP_VIDEO_OUTPUT flags for M2M
616 devices is ambiguous and may lead, for example, to identifying
617 a M2M device as a video capture or output device.
618 Who: Sylwester Nawrocki <s.nawrocki@samsung.com>
620 ----------------------------