Documentation/feature-removal-schedule.txt

   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.
   6
   7 ---------------------------
   8
   9 What:   CONFIG_APM_CPU_IDLE, and its ability to call APM BIOS in idle
  10 When:   2012
  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>
  16
  17 ----------------------------
  18
  19 What:   x86_32 "no-hlt" cmdline param
  20 When:   2012
  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>
  27
  28 ----------------------------
  29
  30 What:   x86 "idle=mwait" cmdline param
  31 When:   2012
  32 Why:    simplify x86 idle code
  33 Who:    Len Brown <len.brown@intel.com>
  34
  35 ----------------------------
  36
  37 What:   PRISM54
  38 When:   2.6.34
  39
  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
  57
  58         For more information see the p54 wiki page:
  59
  60         http://wireless.kernel.org/en/users/Drivers/p54
  61
  62 Who:    Luis R. Rodriguez <lrodriguez@atheros.com>
  63
  64 ---------------------------
  65
  66 What:   IRQF_SAMPLE_RANDOM
  67 Check:  IRQF_SAMPLE_RANDOM
  68 When:   July 2009
  69
  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
  75
  76 Who:    Robin Getz <rgetz@blackfin.uclinux.org> & Matt Mackall <mpm@selenic.com>
  77
  78 ---------------------------
  79
  80 What:   The ieee80211_regdom module parameter
  81 When:   March 2010 / desktop catchup
  82
  83 Why:    This was inherited by the CONFIG_WIRELESS_OLD_REGULATORY code,
  84         and currently serves as an option for users to define an
  85         ISO / IEC 3166 alpha2 code for the country they are currently
  86         present in. Although there are userspace API replacements for this
  87         through nl80211 distributions haven't yet caught up with implementing
  88         decent alternatives through standard GUIs. Although available as an
  89         option through iw or wpa_supplicant its just a matter of time before
  90         distributions pick up good GUI options for this. The ideal solution
  91         would actually consist of intelligent designs which would do this for
  92         the user automatically even when travelling through different countries.
  93         Until then we leave this module parameter as a compromise.
  94
  95         When userspace improves with reasonable widely-available alternatives for
  96         this we will no longer need this module parameter. This entry hopes that
  97         by the super-futuristically looking date of "March 2010" we will have
  98         such replacements widely available.
  99
 100 Who:    Luis R. Rodriguez <lrodriguez@atheros.com>
 101
 102 ---------------------------
 103
 104 What:   dev->power.power_state
 105 When:   July 2007
 106 Why:    Broken design for runtime control over driver power states, confusing
 107         driver-internal runtime power management with:  mechanisms to support
 108         system-wide sleep state transitions; event codes that distinguish
 109         different phases of swsusp "sleep" transitions; and userspace policy
 110         inputs.  This framework was never widely used, and most attempts to
 111         use it were broken.  Drivers should instead be exposing domain-specific
 112         interfaces either to kernel or to userspace.
 113 Who:    Pavel Machek <pavel@ucw.cz>
 114
 115 ---------------------------
 116
 117 What:   /proc/<pid>/oom_adj
 118 When:   August 2012
 119 Why:    /proc/<pid>/oom_adj allows userspace to influence the oom killer's
 120         badness heuristic used to determine which task to kill when the kernel
 121         is out of memory.
 122
 123         The badness heuristic has since been rewritten since the introduction of
 124         this tunable such that its meaning is deprecated.  The value was
 125         implemented as a bitshift on a score generated by the badness()
 126         function that did not have any precise units of measure.  With the
 127         rewrite, the score is given as a proportion of available memory to the
 128         task allocating pages, so using a bitshift which grows the score
 129         exponentially is, thus, impossible to tune with fine granularity.
 130
 131         A much more powerful interface, /proc/<pid>/oom_score_adj, was
 132         introduced with the oom killer rewrite that allows users to increase or
 133         decrease the badness score linearly.  This interface will replace
 134         /proc/<pid>/oom_adj.
 135
 136         A warning will be emitted to the kernel log if an application uses this
 137         deprecated interface.  After it is printed once, future warnings will be
 138         suppressed until the kernel is rebooted.
 139
 140 ---------------------------
 141
 142 What:   remove EXPORT_SYMBOL(kernel_thread)
 143 When:   August 2006
 144 Files:  arch/*/kernel/*_ksyms.c
 145 Check:  kernel_thread
 146 Why:    kernel_thread is a low-level implementation detail.  Drivers should
 147         use the <linux/kthread.h> API instead which shields them from
 148         implementation details and provides a higherlevel interface that
 149         prevents bugs and code duplication
 150 Who:    Christoph Hellwig <hch@lst.de>
 151
 152 ---------------------------
 153
 154 What:   Unused EXPORT_SYMBOL/EXPORT_SYMBOL_GPL exports
 155         (temporary transition config option provided until then)
 156         The transition config option will also be removed at the same time.
 157 When:   before 2.6.19
 158 Why:    Unused symbols are both increasing the size of the kernel binary
 159         and are often a sign of "wrong API"
 160 Who:    Arjan van de Ven <arjan@linux.intel.com>
 161
 162 ---------------------------
 163
 164 What:   PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
 165 When:   October 2008
 166 Why:    The stacking of class devices makes these values misleading and
 167         inconsistent.
 168         Class devices should not carry any of these properties, and bus
 169         devices have SUBSYTEM and DRIVER as a replacement.
 170 Who:    Kay Sievers <kay.sievers@suse.de>
 171
 172 ---------------------------
 173
 174 What:   ACPI procfs interface
 175 When:   July 2008
 176 Why:    ACPI sysfs conversion should be finished by January 2008.
 177         ACPI procfs interface will be removed in July 2008 so that
 178         there is enough time for the user space to catch up.
 179 Who:    Zhang Rui <rui.zhang@intel.com>
 180
 181 ---------------------------
 182
 183 What:   CONFIG_ACPI_PROCFS_POWER
 184 When:   2.6.39
 185 Why:    sysfs I/F for ACPI power devices, including AC and Battery,
 186         has been working in upstream kernel since 2.6.24, Sep 2007.
 187         In 2.6.37, we make the sysfs I/F always built in and this option
 188         disabled by default.
 189         Remove this option and the ACPI power procfs interface in 2.6.39.
 190 Who:    Zhang Rui <rui.zhang@intel.com>
 191
 192 ---------------------------
 193
 194 What:   /proc/acpi/event
 195 When:   February 2008
 196 Why:    /proc/acpi/event has been replaced by events via the input layer
 197         and netlink since 2.6.23.
 198 Who:    Len Brown <len.brown@intel.com>
 199
 200 ---------------------------
 201
 202 What:   i386/x86_64 bzImage symlinks
 203 When:   April 2010
 204
 205 Why:    The i386/x86_64 merge provides a symlink to the old bzImage
 206         location so not yet updated user space tools, e.g. package
 207         scripts, do not break.
 208 Who:    Thomas Gleixner <tglx@linutronix.de>
 209
 210 ---------------------------
 211
 212 What:   GPIO autorequest on gpio_direction_{input,output}() in gpiolib
 213 When:   February 2010
 214 Why:    All callers should use explicit gpio_request()/gpio_free().
 215         The autorequest mechanism in gpiolib was provided mostly as a
 216         migration aid for legacy GPIO interfaces (for SOC based GPIOs).
 217         Those users have now largely migrated.  Platforms implementing
 218         the GPIO interfaces without using gpiolib will see no changes.
 219 Who:    David Brownell <dbrownell@users.sourceforge.net>
 220 ---------------------------
 221
 222 What:   b43 support for firmware revision < 410
 223 When:   The schedule was July 2008, but it was decided that we are going to keep the
 224         code as long as there are no major maintanance headaches.
 225         So it _could_ be removed _any_ time now, if it conflicts with something new.
 226 Why:    The support code for the old firmware hurts code readability/maintainability
 227         and slightly hurts runtime performance. Bugfixes for the old firmware
 228         are not provided by Broadcom anymore.
 229 Who:    Michael Buesch <m@bues.ch>
 230
 231 ---------------------------
 232
 233 What:   Ability for non root users to shm_get hugetlb pages based on mlock
 234         resource limits
 235 When:   2.6.31
 236 Why:    Non root users need to be part of /proc/sys/vm/hugetlb_shm_group or
 237         have CAP_IPC_LOCK to be able to allocate shm segments backed by
 238         huge pages.  The mlock based rlimit check to allow shm hugetlb is
 239         inconsistent with mmap based allocations.  Hence it is being
 240         deprecated.
 241 Who:    Ravikiran Thirumalai <kiran@scalex86.org>
 242
 243 ---------------------------
 244
 245 What:   Code that is now under CONFIG_WIRELESS_EXT_SYSFS
 246         (in net/core/net-sysfs.c)
 247 When:   3.5
 248 Why:    Over 1K .text/.data size reduction, data is available in other
 249         ways (ioctls)
 250 Who:    Johannes Berg <johannes@sipsolutions.net>
 251
 252 ---------------------------
 253
 254 What:   sysfs ui for changing p4-clockmod parameters
 255 When:   September 2009
 256 Why:    See commits 129f8ae9b1b5be94517da76009ea956e89104ce8 and
 257         e088e4c9cdb618675874becb91b2fd581ee707e6.
 258         Removal is subject to fixing any remaining bugs in ACPI which may
 259         cause the thermal throttling not to happen at the right time.
 260 Who:    Dave Jones <davej@redhat.com>, Matthew Garrett <mjg@redhat.com>
 261
 262 -----------------------------
 263
 264 What:   fakephp and associated sysfs files in /sys/bus/pci/slots/
 265 When:   2011
 266 Why:    In 2.6.27, the semantics of /sys/bus/pci/slots was redefined to
 267         represent a machine's physical PCI slots. The change in semantics
 268         had userspace implications, as the hotplug core no longer allowed
 269         drivers to create multiple sysfs files per physical slot (required
 270         for multi-function devices, e.g.). fakephp was seen as a developer's
 271         tool only, and its interface changed. Too late, we learned that
 272         there were some users of the fakephp interface.
 273
 274         In 2.6.30, the original fakephp interface was restored. At the same
 275         time, the PCI core gained the ability that fakephp provided, namely
 276         function-level hot-remove and hot-add.
 277
 278         Since the PCI core now provides the same functionality, exposed in:
 279
 280                 /sys/bus/pci/rescan
 281                 /sys/bus/pci/devices/.../remove
 282                 /sys/bus/pci/devices/.../rescan
 283
 284         there is no functional reason to maintain fakephp as well.
 285
 286         We will keep the existing module so that 'modprobe fakephp' will
 287         present the old /sys/bus/pci/slots/... interface for compatibility,
 288         but users are urged to migrate their applications to the API above.
 289
 290         After a reasonable transition period, we will remove the legacy
 291         fakephp interface.
 292 Who:    Alex Chiang <achiang@hp.com>
 293
 294 ---------------------------
 295
 296 What:   CONFIG_RFKILL_INPUT
 297 When:   2.6.33
 298 Why:    Should be implemented in userspace, policy daemon.
 299 Who:    Johannes Berg <johannes@sipsolutions.net>
 300
 301 ----------------------------
 302
 303 What:   sound-slot/service-* module aliases and related clutters in
 304         sound/sound_core.c
 305 When:   August 2010
 306 Why:    OSS sound_core grabs all legacy minors (0-255) of SOUND_MAJOR
 307         (14) and requests modules using custom sound-slot/service-*
 308         module aliases.  The only benefit of doing this is allowing
 309         use of custom module aliases which might as well be considered
 310         a bug at this point.  This preemptive claiming prevents
 311         alternative OSS implementations.
 312
 313         Till the feature is removed, the kernel will be requesting
 314         both sound-slot/service-* and the standard char-major-* module
 315         aliases and allow turning off the pre-claiming selectively via
 316         CONFIG_SOUND_OSS_CORE_PRECLAIM and soundcore.preclaim_oss
 317         kernel parameter.
 318
 319         After the transition phase is complete, both the custom module
 320         aliases and switches to disable it will go away.  This removal
 321         will also allow making ALSA OSS emulation independent of
 322         sound_core.  The dependency will be broken then too.
 323 Who:    Tejun Heo <tj@kernel.org>
 324
 325 ----------------------------
 326
 327 What:   sysfs-class-rfkill state file
 328 When:   Feb 2014
 329 Files:  net/rfkill/core.c
 330 Why:    Documented as obsolete since Feb 2010. This file is limited to 3
 331         states while the rfkill drivers can have 4 states.
 332 Who:    anybody or Florian Mickler <florian@mickler.org>
 333
 334 ----------------------------
 335
 336 What:   sysfs-class-rfkill claim file
 337 When:   Feb 2012
 338 Files:  net/rfkill/core.c
 339 Why:    It is not possible to claim an rfkill driver since 2007. This is
 340         Documented as obsolete since Feb 2010.
 341 Who:    anybody or Florian Mickler <florian@mickler.org>
 342
 343 ----------------------------
 344
 345 What:   iwlwifi 50XX module parameters
 346 When:   3.0
 347 Why:    The "..50" modules parameters were used to configure 5000 series and
 348         up devices; different set of module parameters also available for 4965
 349         with same functionalities. Consolidate both set into single place
 350         in drivers/net/wireless/iwlwifi/iwl-agn.c
 351
 352 Who:    Wey-Yi Guy <wey-yi.w.guy@intel.com>
 353
 354 ----------------------------
 355
 356 What:   iwl4965 alias support
 357 When:   3.0
 358 Why:    Internal alias support has been present in module-init-tools for some
 359         time, the MODULE_ALIAS("iwl4965") boilerplate aliases can be removed
 360         with no impact.
 361
 362 Who:    Wey-Yi Guy <wey-yi.w.guy@intel.com>
 363
 364 ---------------------------
 365
 366 What:   xt_NOTRACK
 367 Files:  net/netfilter/xt_NOTRACK.c
 368 When:   April 2011
 369 Why:    Superseded by xt_CT
 370 Who:    Netfilter developer team <netfilter-devel@vger.kernel.org>
 371
 372 ----------------------------
 373
 374 What:   IRQF_DISABLED
 375 When:   2.6.36
 376 Why:    The flag is a NOOP as we run interrupt handlers with interrupts disabled
 377 Who:    Thomas Gleixner <tglx@linutronix.de>
 378
 379 ----------------------------
 380
 381 What:   PCI DMA unmap state API
 382 When:   August 2012
 383 Why:    PCI DMA unmap state API (include/linux/pci-dma.h) was replaced
 384         with DMA unmap state API (DMA unmap state API can be used for
 385         any bus).
 386 Who:    FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
 387
 388 ----------------------------
 389
 390 What:   iwlwifi disable_hw_scan module parameters
 391 When:   3.0
 392 Why:    Hareware scan is the prefer method for iwlwifi devices for
 393         scanning operation. Remove software scan support for all the
 394         iwlwifi devices.
 395
 396 Who:    Wey-Yi Guy <wey-yi.w.guy@intel.com>
 397
 398 ----------------------------
 399
 400 What:   Legacy, non-standard chassis intrusion detection interface.
 401 When:   June 2011
 402 Why:    The adm9240, w83792d and w83793 hardware monitoring drivers have
 403         legacy interfaces for chassis intrusion detection. A standard
 404         interface has been added to each driver, so the legacy interface
 405         can be removed.
 406 Who:    Jean Delvare <khali@linux-fr.org>
 407
 408 ----------------------------
 409
 410 What:   xt_connlimit rev 0
 411 When:   2012
 412 Who:    Jan Engelhardt <jengelh@medozas.de>
 413 Files:  net/netfilter/xt_connlimit.c
 414
 415 ----------------------------
 416
 417 What:   ipt_addrtype match include file
 418 When:   2012
 419 Why:    superseded by xt_addrtype
 420 Who:    Florian Westphal <fw@strlen.de>
 421 Files:  include/linux/netfilter_ipv4/ipt_addrtype.h
 422
 423 ----------------------------
 424
 425 What:   i2c_driver.attach_adapter
 426         i2c_driver.detach_adapter
 427 When:   September 2011
 428 Why:    These legacy callbacks should no longer be used as i2c-core offers
 429         a variety of preferable alternative ways to instantiate I2C devices.
 430 Who:    Jean Delvare <khali@linux-fr.org>
 431
 432 ----------------------------
 433
 434 What:   Opening a radio device node will no longer automatically switch the
 435         tuner mode from tv to radio.
 436 When:   3.3
 437 Why:    Just opening a V4L device should not change the state of the hardware
 438         like that. It's very unexpected and against the V4L spec. Instead, you
 439         switch to radio mode by calling VIDIOC_S_FREQUENCY. This is the second
 440         and last step of the move to consistent handling of tv and radio tuners.
 441 Who:    Hans Verkuil <hans.verkuil@cisco.com>
 442
 443 ----------------------------
 444
 445 What:   g_file_storage driver
 446 When:   3.8
 447 Why:    This driver has been superseded by g_mass_storage.
 448 Who:    Alan Stern <stern@rowland.harvard.edu>
 449
 450 ----------------------------
 451
 452 What:   threeg and interface sysfs files in /sys/devices/platform/acer-wmi
 453 When:   2012
 454 Why:    In 3.0, we can now autodetect internal 3G device and already have
 455         the threeg rfkill device. So, we plan to remove threeg sysfs support
 456         for it's no longer necessary.
 457
 458         We also plan to remove interface sysfs file that exposed which ACPI-WMI
 459         interface that was used by acer-wmi driver. It will replaced by
 460         information log when acer-wmi initial.
 461 Who:    Lee, Chun-Yi <jlee@novell.com>
 462
 463 ---------------------------
 464
 465 What:   /sys/devices/platform/_UDC_/udc/_UDC_/is_dualspeed file and
 466         is_dualspeed line in /sys/devices/platform/ci13xxx_*/udc/device file.
 467 When:   3.8
 468 Why:    The is_dualspeed file is superseded by maximum_speed in the same
 469         directory and is_dualspeed line in device file is superseded by
 470         max_speed line in the same file.
 471
 472         The maximum_speed/max_speed specifies maximum speed supported by UDC.
 473         To check if dualspeeed is supported, check if the value is >= 3.
 474         Various possible speeds are defined in <linux/usb/ch9.h>.
 475 Who:    Michal Nazarewicz <mina86@mina86.com>
 476
 477 ----------------------------
 478
 479 What:   The XFS nodelaylog mount option
 480 When:   3.3
 481 Why:    The delaylog mode that has been the default since 2.6.39 has proven
 482         stable, and the old code is in the way of additional improvements in
 483         the log code.
 484 Who:    Christoph Hellwig <hch@lst.de>
 485
 486 ----------------------------
 487
 488 What:   iwlagn alias support
 489 When:   3.5
 490 Why:    The iwlagn module has been renamed iwlwifi.  The alias will be around
 491         for backward compatibility for several cycles and then dropped.
 492 Who:    Don Fry <donald.h.fry@intel.com>
 493
 494 ----------------------------
 495
 496 What:   pci_scan_bus_parented()
 497 When:   3.5
 498 Why:    The pci_scan_bus_parented() interface creates a new root bus.  The
 499         bus is created with default resources (ioport_resource and
 500         iomem_resource) that are always wrong, so we rely on arch code to
 501         correct them later.  Callers of pci_scan_bus_parented() should
 502         convert to using pci_scan_root_bus() so they can supply a list of
 503         bus resources when the bus is created.
 504 Who:    Bjorn Helgaas <bhelgaas@google.com>
 505
 506 ----------------------------
 507
 508 What:   Low Performance USB Block driver ("CONFIG_BLK_DEV_UB")
 509 When:   3.6
 510 Why:    This driver provides support for USB storage devices like "USB
 511         sticks". As of now, it is deactivated in Debian, Fedora and
 512         Ubuntu. All current users can switch over to usb-storage
 513         (CONFIG_USB_STORAGE) which only drawback is the additional SCSI
 514         stack.
 515 Who:    Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
 516
 517 ----------------------------
 518
 519 What:   kmap_atomic(page, km_type)
 520 When:   3.5
 521 Why:    The old kmap_atomic() with two arguments is deprecated, we only
 522         keep it for backward compatibility for few cycles and then drop it.
 523 Who:    Cong Wang <amwang@redhat.com>
 524
 525 ----------------------------
 526
 527 What:   get_robust_list syscall
 528 When:   2013
 529 Why:    There appear to be no production users of the get_robust_list syscall,
 530         and it runs the risk of leaking address locations, allowing the bypass
 531         of ASLR. It was only ever intended for debugging, so it should be
 532         removed.
 533 Who:    Kees Cook <keescook@chromium.org>
 534
 535 ----------------------------
 536
 537 What:   setitimer accepts user NULL pointer (value)
 538 When:   3.6
 539 Why:    setitimer is not returning -EFAULT if user pointer is NULL. This
 540         violates the spec.
 541 Who:    Sasikantha Babu <sasikanth.v19@gmail.com>