]> git.proxmox.com Git - mirror_ubuntu-artful-kernel.git/blob - Documentation/kernel-parameters.txt
945bc0de7f0750acefae3f89695c9b1b510b18be
[mirror_ubuntu-artful-kernel.git] / Documentation / kernel-parameters.txt
1 Kernel Parameters
2 ~~~~~~~~~~~~~~~~~
3
4 The following is a consolidated list of the kernel parameters as
5 implemented by the __setup(), core_param() and module_param() macros
6 and sorted into English Dictionary order (defined as ignoring all
7 punctuation and sorting digits before letters in a case insensitive
8 manner), and with descriptions where known.
9
10 The kernel parses parameters from the kernel command line up to "--";
11 if it doesn't recognize a parameter and it doesn't contain a '.', the
12 parameter gets passed to init: parameters with '=' go into init's
13 environment, others are passed as command line arguments to init.
14 Everything after "--" is passed as an argument to init.
15
16 Module parameters can be specified in two ways: via the kernel command
17 line with a module name prefix, or via modprobe, e.g.:
18
19 (kernel command line) usbcore.blinkenlights=1
20 (modprobe command line) modprobe usbcore blinkenlights=1
21
22 Parameters for modules which are built into the kernel need to be
23 specified on the kernel command line. modprobe looks through the
24 kernel command line (/proc/cmdline) and collects module parameters
25 when it loads a module, so the kernel command line can be used for
26 loadable modules too.
27
28 Hyphens (dashes) and underscores are equivalent in parameter names, so
29 log_buf_len=1M print-fatal-signals=1
30 can also be entered as
31 log-buf-len=1M print_fatal_signals=1
32
33 Double-quotes can be used to protect spaces in values, e.g.:
34 param="spaces in here"
35
36 This document may not be entirely up to date and comprehensive. The command
37 "modinfo -p ${modulename}" shows a current list of all parameters of a loadable
38 module. Loadable modules, after being loaded into the running kernel, also
39 reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
40 parameters may be changed at runtime by the command
41 "echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
42
43 The parameters listed below are only valid if certain kernel build options were
44 enabled and if respective hardware is present. The text in square brackets at
45 the beginning of each description states the restrictions within which a
46 parameter is applicable:
47
48 ACPI ACPI support is enabled.
49 AGP AGP (Accelerated Graphics Port) is enabled.
50 ALSA ALSA sound support is enabled.
51 APIC APIC support is enabled.
52 APM Advanced Power Management support is enabled.
53 ARM ARM architecture is enabled.
54 AVR32 AVR32 architecture is enabled.
55 AX25 Appropriate AX.25 support is enabled.
56 BLACKFIN Blackfin architecture is enabled.
57 CLK Common clock infrastructure is enabled.
58 CMA Contiguous Memory Area support is enabled.
59 DRM Direct Rendering Management support is enabled.
60 DYNAMIC_DEBUG Build in debug messages and enable them at runtime
61 EDD BIOS Enhanced Disk Drive Services (EDD) is enabled
62 EFI EFI Partitioning (GPT) is enabled
63 EIDE EIDE/ATAPI support is enabled.
64 EVM Extended Verification Module
65 FB The frame buffer device is enabled.
66 FTRACE Function tracing enabled.
67 GCOV GCOV profiling is enabled.
68 HW Appropriate hardware is enabled.
69 IA-64 IA-64 architecture is enabled.
70 IMA Integrity measurement architecture is enabled.
71 IOSCHED More than one I/O scheduler is enabled.
72 IP_PNP IP DHCP, BOOTP, or RARP is enabled.
73 IPV6 IPv6 support is enabled.
74 ISAPNP ISA PnP code is enabled.
75 ISDN Appropriate ISDN support is enabled.
76 JOY Appropriate joystick support is enabled.
77 KGDB Kernel debugger support is enabled.
78 KVM Kernel Virtual Machine support is enabled.
79 LIBATA Libata driver is enabled
80 LP Printer support is enabled.
81 LOOP Loopback device support is enabled.
82 M68k M68k architecture is enabled.
83 These options have more detailed description inside of
84 Documentation/m68k/kernel-options.txt.
85 MDA MDA console support is enabled.
86 MIPS MIPS architecture is enabled.
87 MOUSE Appropriate mouse support is enabled.
88 MSI Message Signaled Interrupts (PCI).
89 MTD MTD (Memory Technology Device) support is enabled.
90 NET Appropriate network support is enabled.
91 NUMA NUMA support is enabled.
92 NFS Appropriate NFS support is enabled.
93 OSS OSS sound support is enabled.
94 PV_OPS A paravirtualized kernel is enabled.
95 PARIDE The ParIDE (parallel port IDE) subsystem is enabled.
96 PARISC The PA-RISC architecture is enabled.
97 PCI PCI bus support is enabled.
98 PCIE PCI Express support is enabled.
99 PCMCIA The PCMCIA subsystem is enabled.
100 PNP Plug & Play support is enabled.
101 PPC PowerPC architecture is enabled.
102 PPT Parallel port support is enabled.
103 PS2 Appropriate PS/2 support is enabled.
104 RAM RAM disk support is enabled.
105 S390 S390 architecture is enabled.
106 SCSI Appropriate SCSI support is enabled.
107 A lot of drivers have their options described inside
108 the Documentation/scsi/ sub-directory.
109 SECURITY Different security models are enabled.
110 SELINUX SELinux support is enabled.
111 APPARMOR AppArmor support is enabled.
112 SERIAL Serial support is enabled.
113 SH SuperH architecture is enabled.
114 SMP The kernel is an SMP kernel.
115 SPARC Sparc architecture is enabled.
116 SWSUSP Software suspend (hibernation) is enabled.
117 SUSPEND System suspend states are enabled.
118 TPM TPM drivers are enabled.
119 TS Appropriate touchscreen support is enabled.
120 UMS USB Mass Storage support is enabled.
121 USB USB support is enabled.
122 USBHID USB Human Interface Device support is enabled.
123 V4L Video For Linux support is enabled.
124 VMMIO Driver for memory mapped virtio devices is enabled.
125 VGA The VGA console has been enabled.
126 VT Virtual terminal support is enabled.
127 WDT Watchdog support is enabled.
128 XT IBM PC/XT MFM hard disk support is enabled.
129 X86-32 X86-32, aka i386 architecture is enabled.
130 X86-64 X86-64 architecture is enabled.
131 More X86-64 boot options can be found in
132 Documentation/x86/x86_64/boot-options.txt .
133 X86 Either 32-bit or 64-bit x86 (same as X86-32+X86-64)
134 XEN Xen support is enabled
135
136 In addition, the following text indicates that the option:
137
138 BUGS= Relates to possible processor bugs on the said processor.
139 KNL Is a kernel start-up parameter.
140 BOOT Is a boot loader parameter.
141
142 Parameters denoted with BOOT are actually interpreted by the boot
143 loader, and have no meaning to the kernel directly.
144 Do not modify the syntax of boot loader parameters without extreme
145 need or coordination with <Documentation/x86/boot.txt>.
146
147 There are also arch-specific kernel-parameters not documented here.
148 See for example <Documentation/x86/x86_64/boot-options.txt>.
149
150 Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
151 a trailing = on the name of any parameter states that that parameter will
152 be entered as an environment variable, whereas its absence indicates that
153 it will appear as a kernel argument readable via /proc/cmdline by programs
154 running once the system is up.
155
156 The number of kernel parameters is not limited, but the length of the
157 complete command line (parameters including spaces etc.) is limited to
158 a fixed number of characters. This limit depends on the architecture
159 and is between 256 and 4096 characters. It is defined in the file
160 ./include/asm/setup.h as COMMAND_LINE_SIZE.
161
162 Finally, the [KMG] suffix is commonly described after a number of kernel
163 parameter values. These 'K', 'M', and 'G' letters represent the _binary_
164 multipliers 'Kilo', 'Mega', and 'Giga', equalling 2^10, 2^20, and 2^30
165 bytes respectively. Such letter suffixes can also be entirely omitted.
166
167
168 acpi= [HW,ACPI,X86,ARM64]
169 Advanced Configuration and Power Interface
170 Format: { force | off | strict | noirq | rsdt }
171 force -- enable ACPI if default was off
172 off -- disable ACPI if default was on
173 noirq -- do not use ACPI for IRQ routing
174 strict -- Be less tolerant of platforms that are not
175 strictly ACPI specification compliant.
176 rsdt -- prefer RSDT over (default) XSDT
177 copy_dsdt -- copy DSDT to memory
178 For ARM64, ONLY "acpi=off" or "acpi=force" are available
179
180 See also Documentation/power/runtime_pm.txt, pci=noacpi
181
182 acpi_apic_instance= [ACPI, IOAPIC]
183 Format: <int>
184 2: use 2nd APIC table, if available
185 1,0: use 1st APIC table
186 default: 0
187
188 acpi_backlight= [HW,ACPI]
189 acpi_backlight=vendor
190 acpi_backlight=video
191 If set to vendor, prefer vendor specific driver
192 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
193 of the ACPI video.ko driver.
194
195 acpica_no_return_repair [HW, ACPI]
196 Disable AML predefined validation mechanism
197 This mechanism can repair the evaluation result to make
198 the return objects more ACPI specification compliant.
199 This option is useful for developers to identify the
200 root cause of an AML interpreter issue when the issue
201 has something to do with the repair mechanism.
202
203 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
204 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
205 Format: <int>
206 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
207 debug output. Bits in debug_layer correspond to a
208 _COMPONENT in an ACPI source file, e.g.,
209 #define _COMPONENT ACPI_PCI_COMPONENT
210 Bits in debug_level correspond to a level in
211 ACPI_DEBUG_PRINT statements, e.g.,
212 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
213 The debug_level mask defaults to "info". See
214 Documentation/acpi/debug.txt for more information about
215 debug layers and levels.
216
217 Enable processor driver info messages:
218 acpi.debug_layer=0x20000000
219 Enable PCI/PCI interrupt routing info messages:
220 acpi.debug_layer=0x400000
221 Enable AML "Debug" output, i.e., stores to the Debug
222 object while interpreting AML:
223 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
224 Enable all messages related to ACPI hardware:
225 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
226
227 Some values produce so much output that the system is
228 unusable. The "log_buf_len" parameter may be useful
229 if you need to capture more output.
230
231 acpi_enforce_resources= [ACPI]
232 { strict | lax | no }
233 Check for resource conflicts between native drivers
234 and ACPI OperationRegions (SystemIO and SystemMemory
235 only). IO ports and memory declared in ACPI might be
236 used by the ACPI subsystem in arbitrary AML code and
237 can interfere with legacy drivers.
238 strict (default): access to resources claimed by ACPI
239 is denied; legacy drivers trying to access reserved
240 resources will fail to bind to device using them.
241 lax: access to resources claimed by ACPI is allowed;
242 legacy drivers trying to access reserved resources
243 will bind successfully but a warning message is logged.
244 no: ACPI OperationRegions are not marked as reserved,
245 no further checks are performed.
246
247 acpi_force_table_verification [HW,ACPI]
248 Enable table checksum verification during early stage.
249 By default, this is disabled due to x86 early mapping
250 size limitation.
251
252 acpi_irq_balance [HW,ACPI]
253 ACPI will balance active IRQs
254 default in APIC mode
255
256 acpi_irq_nobalance [HW,ACPI]
257 ACPI will not move active IRQs (default)
258 default in PIC mode
259
260 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
261 Format: <irq>,<irq>...
262
263 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
264 use by PCI
265 Format: <irq>,<irq>...
266
267 acpi_no_auto_serialize [HW,ACPI]
268 Disable auto-serialization of AML methods
269 AML control methods that contain the opcodes to create
270 named objects will be marked as "Serialized" by the
271 auto-serialization feature.
272 This feature is enabled by default.
273 This option allows to turn off the feature.
274
275 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
276 kernels.
277
278 acpi_no_static_ssdt [HW,ACPI]
279 Disable installation of static SSDTs at early boot time
280 By default, SSDTs contained in the RSDT/XSDT will be
281 installed automatically and they will appear under
282 /sys/firmware/acpi/tables.
283 This option turns off this feature.
284 Note that specifying this option does not affect
285 dynamic table installation which will install SSDT
286 tables to /sys/firmware/acpi/tables/dynamic.
287
288 acpi_rsdp= [ACPI,EFI,KEXEC]
289 Pass the RSDP address to the kernel, mostly used
290 on machines running EFI runtime service to boot the
291 second kernel for kdump.
292
293 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
294 Format: To spoof as Windows 98: ="Microsoft Windows"
295
296 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
297 of 2 which is mandated by ACPI 6) as the supported ACPI
298 specification revision (when using this switch, it may
299 be necessary to carry out a cold reboot _twice_ in a
300 row to make it take effect on the platform firmware).
301
302 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
303 acpi_osi="string1" # add string1
304 acpi_osi="!string2" # remove string2
305 acpi_osi=!* # remove all strings
306 acpi_osi=! # disable all built-in OS vendor
307 strings
308 acpi_osi= # disable all strings
309
310 'acpi_osi=!' can be used in combination with single or
311 multiple 'acpi_osi="string1"' to support specific OS
312 vendor string(s). Note that such command can only
313 affect the default state of the OS vendor strings, thus
314 it cannot affect the default state of the feature group
315 strings and the current state of the OS vendor strings,
316 specifying it multiple times through kernel command line
317 is meaningless. This command is useful when one do not
318 care about the state of the feature group strings which
319 should be controlled by the OSPM.
320 Examples:
321 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
322 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
323 can make '_OSI("Windows 2000")' TRUE.
324
325 'acpi_osi=' cannot be used in combination with other
326 'acpi_osi=' command lines, the _OSI method will not
327 exist in the ACPI namespace. NOTE that such command can
328 only affect the _OSI support state, thus specifying it
329 multiple times through kernel command line is also
330 meaningless.
331 Examples:
332 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
333 FALSE.
334
335 'acpi_osi=!*' can be used in combination with single or
336 multiple 'acpi_osi="string1"' to support specific
337 string(s). Note that such command can affect the
338 current state of both the OS vendor strings and the
339 feature group strings, thus specifying it multiple times
340 through kernel command line is meaningful. But it may
341 still not able to affect the final state of a string if
342 there are quirks related to this string. This command
343 is useful when one want to control the state of the
344 feature group strings to debug BIOS issues related to
345 the OSPM features.
346 Examples:
347 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
348 '_OSI("Module Device")' FALSE.
349 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
350 '_OSI("Module Device")' TRUE.
351 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
352 equivalent to
353 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
354 and
355 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
356 they all will make '_OSI("Windows 2000")' TRUE.
357
358 acpi_pm_good [X86]
359 Override the pmtimer bug detection: force the kernel
360 to assume that this machine's pmtimer latches its value
361 and always returns good values.
362
363 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
364 Format: { level | edge | high | low }
365
366 acpi_skip_timer_override [HW,ACPI]
367 Recognize and ignore IRQ0/pin2 Interrupt Override.
368 For broken nForce2 BIOS resulting in XT-PIC timer.
369
370 acpi_sleep= [HW,ACPI] Sleep options
371 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
372 old_ordering, nonvs, sci_force_enable }
373 See Documentation/power/video.txt for information on
374 s3_bios and s3_mode.
375 s3_beep is for debugging; it makes the PC's speaker beep
376 as soon as the kernel's real-mode entry point is called.
377 s4_nohwsig prevents ACPI hardware signature from being
378 used during resume from hibernation.
379 old_ordering causes the ACPI 1.0 ordering of the _PTS
380 control method, with respect to putting devices into
381 low power states, to be enforced (the ACPI 2.0 ordering
382 of _PTS is used by default).
383 nonvs prevents the kernel from saving/restoring the
384 ACPI NVS memory during suspend/hibernation and resume.
385 sci_force_enable causes the kernel to set SCI_EN directly
386 on resume from S1/S3 (which is against the ACPI spec,
387 but some broken systems don't work without it).
388
389 acpi_use_timer_override [HW,ACPI]
390 Use timer override. For some broken Nvidia NF5 boards
391 that require a timer override, but don't have HPET
392
393 add_efi_memmap [EFI; X86] Include EFI memory map in
394 kernel's map of available physical RAM.
395
396 agp= [AGP]
397 { off | try_unsupported }
398 off: disable AGP support
399 try_unsupported: try to drive unsupported chipsets
400 (may crash computer or cause data corruption)
401
402 ALSA [HW,ALSA]
403 See Documentation/sound/alsa/alsa-parameters.txt
404
405 alignment= [KNL,ARM]
406 Allow the default userspace alignment fault handler
407 behaviour to be specified. Bit 0 enables warnings,
408 bit 1 enables fixups, and bit 2 sends a segfault.
409
410 align_va_addr= [X86-64]
411 Align virtual addresses by clearing slice [14:12] when
412 allocating a VMA at process creation time. This option
413 gives you up to 3% performance improvement on AMD F15h
414 machines (where it is enabled by default) for a
415 CPU-intensive style benchmark, and it can vary highly in
416 a microbenchmark depending on workload and compiler.
417
418 32: only for 32-bit processes
419 64: only for 64-bit processes
420 on: enable for both 32- and 64-bit processes
421 off: disable for both 32- and 64-bit processes
422
423 alloc_snapshot [FTRACE]
424 Allocate the ftrace snapshot buffer on boot up when the
425 main buffer is allocated. This is handy if debugging
426 and you need to use tracing_snapshot() on boot up, and
427 do not want to use tracing_snapshot_alloc() as it needs
428 to be done where GFP_KERNEL allocations are allowed.
429
430 amd_iommu= [HW,X86-64]
431 Pass parameters to the AMD IOMMU driver in the system.
432 Possible values are:
433 fullflush - enable flushing of IO/TLB entries when
434 they are unmapped. Otherwise they are
435 flushed before they will be reused, which
436 is a lot of faster
437 off - do not initialize any AMD IOMMU found in
438 the system
439 force_isolation - Force device isolation for all
440 devices. The IOMMU driver is not
441 allowed anymore to lift isolation
442 requirements as needed. This option
443 does not override iommu=pt
444
445 amd_iommu_dump= [HW,X86-64]
446 Enable AMD IOMMU driver option to dump the ACPI table
447 for AMD IOMMU. With this option enabled, AMD IOMMU
448 driver will print ACPI tables for AMD IOMMU during
449 IOMMU initialization.
450
451 amijoy.map= [HW,JOY] Amiga joystick support
452 Map of devices attached to JOY0DAT and JOY1DAT
453 Format: <a>,<b>
454 See also Documentation/input/joystick.txt
455
456 analog.map= [HW,JOY] Analog joystick and gamepad support
457 Specifies type or capabilities of an analog joystick
458 connected to one of 16 gameports
459 Format: <type1>,<type2>,..<type16>
460
461 apc= [HW,SPARC]
462 Power management functions (SPARCstation-4/5 + deriv.)
463 Format: noidle
464 Disable APC CPU standby support. SPARCstation-Fox does
465 not play well with APC CPU idle - disable it if you have
466 APC and your system crashes randomly.
467
468 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
469 Change the output verbosity whilst booting
470 Format: { quiet (default) | verbose | debug }
471 Change the amount of debugging information output
472 when initialising the APIC and IO-APIC components.
473
474 autoconf= [IPV6]
475 See Documentation/networking/ipv6.txt.
476
477 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
478 Limit apic dumping. The parameter defines the maximal
479 number of local apics being dumped. Also it is possible
480 to set it to "all" by meaning -- no limit here.
481 Format: { 1 (default) | 2 | ... | all }.
482 The parameter valid if only apic=debug or
483 apic=verbose is specified.
484 Example: apic=debug show_lapic=all
485
486 apm= [APM] Advanced Power Management
487 See header of arch/x86/kernel/apm_32.c.
488
489 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
490 Format: <io>,<irq>,<nodeID>
491
492 ataflop= [HW,M68k]
493
494 atarimouse= [HW,MOUSE] Atari Mouse
495
496 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
497 EzKey and similar keyboards
498
499 atkbd.reset= [HW] Reset keyboard during initialization
500
501 atkbd.set= [HW] Select keyboard code set
502 Format: <int> (2 = AT (default), 3 = PS/2)
503
504 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
505 keyboards
506
507 atkbd.softraw= [HW] Choose between synthetic and real raw mode
508 Format: <bool> (0 = real, 1 = synthetic (default))
509
510 atkbd.softrepeat= [HW]
511 Use software keyboard repeat
512
513 audit= [KNL] Enable the audit sub-system
514 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
515 0 - kernel audit is disabled and can not be enabled
516 until the next reboot
517 unset - kernel audit is initialized but disabled and
518 will be fully enabled by the userspace auditd.
519 1 - kernel audit is initialized and partially enabled,
520 storing at most audit_backlog_limit messages in
521 RAM until it is fully enabled by the userspace
522 auditd.
523 Default: unset
524
525 audit_backlog_limit= [KNL] Set the audit queue size limit.
526 Format: <int> (must be >=0)
527 Default: 64
528
529 baycom_epp= [HW,AX25]
530 Format: <io>,<mode>
531
532 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
533 Format: <io>,<mode>
534 See header of drivers/net/hamradio/baycom_par.c.
535
536 baycom_ser_fdx= [HW,AX25]
537 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
538 Format: <io>,<irq>,<mode>[,<baud>]
539 See header of drivers/net/hamradio/baycom_ser_fdx.c.
540
541 baycom_ser_hdx= [HW,AX25]
542 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
543 Format: <io>,<irq>,<mode>
544 See header of drivers/net/hamradio/baycom_ser_hdx.c.
545
546 blkdevparts= Manual partition parsing of block device(s) for
547 embedded devices based on command line input.
548 See Documentation/block/cmdline-partition.txt
549
550 boot_delay= Milliseconds to delay each printk during boot.
551 Values larger than 10 seconds (10000) are changed to
552 no delay (0).
553 Format: integer
554
555 bootmem_debug [KNL] Enable bootmem allocator debug messages.
556
557 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
558 bttv.radio= Most important insmod options are available as
559 kernel args too.
560 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
561 bttv.tuner=
562
563 bulk_remove=off [PPC] This parameter disables the use of the pSeries
564 firmware feature for flushing multiple hpte entries
565 at a time.
566
567 c101= [NET] Moxa C101 synchronous serial card
568
569 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
570 Sometimes CPU hardware bugs make them report the cache
571 size incorrectly. The kernel will attempt work arounds
572 to fix known problems, but for some CPUs it is not
573 possible to determine what the correct size should be.
574 This option provides an override for these situations.
575
576 ca_keys= [KEYS] This parameter identifies a specific key(s) on
577 the system trusted keyring to be used for certificate
578 trust validation.
579 format: { id:<keyid> | builtin }
580
581 cca= [MIPS] Override the kernel pages' cache coherency
582 algorithm. Accepted values range from 0 to 7
583 inclusive. See arch/mips/include/asm/pgtable-bits.h
584 for platform specific values (SB1, Loongson3 and
585 others).
586
587 ccw_timeout_log [S390]
588 See Documentation/s390/CommonIO for details.
589
590 cgroup_disable= [KNL] Disable a particular controller
591 Format: {name of the controller(s) to disable}
592 The effects of cgroup_disable=foo are:
593 - foo isn't auto-mounted if you mount all cgroups in
594 a single hierarchy
595 - foo isn't visible as an individually mountable
596 subsystem
597 {Currently only "memory" controller deal with this and
598 cut the overhead, others just disable the usage. So
599 only cgroup_disable=memory is actually worthy}
600
601 checkreqprot [SELINUX] Set initial checkreqprot flag value.
602 Format: { "0" | "1" }
603 See security/selinux/Kconfig help text.
604 0 -- check protection applied by kernel (includes
605 any implied execute protection).
606 1 -- check protection requested by application.
607 Default value is set via a kernel config option.
608 Value can be changed at runtime via
609 /selinux/checkreqprot.
610
611 cio_ignore= [S390]
612 See Documentation/s390/CommonIO for details.
613 clk_ignore_unused
614 [CLK]
615 Prevents the clock framework from automatically gating
616 clocks that have not been explicitly enabled by a Linux
617 device driver but are enabled in hardware at reset or
618 by the bootloader/firmware. Note that this does not
619 force such clocks to be always-on nor does it reserve
620 those clocks in any way. This parameter is useful for
621 debug and development, but should not be needed on a
622 platform with proper driver support. For more
623 information, see Documentation/clk.txt.
624
625 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
626 [Deprecated]
627 Forces specified clocksource (if available) to be used
628 when calculating gettimeofday(). If specified
629 clocksource is not available, it defaults to PIT.
630 Format: { pit | tsc | cyclone | pmtmr }
631
632 clocksource= Override the default clocksource
633 Format: <string>
634 Override the default clocksource and use the clocksource
635 with the name specified.
636 Some clocksource names to choose from, depending on
637 the platform:
638 [all] jiffies (this is the base, fallback clocksource)
639 [ACPI] acpi_pm
640 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
641 pxa_timer,timer3,32k_counter,timer0_1
642 [AVR32] avr32
643 [X86-32] pit,hpet,tsc;
644 scx200_hrt on Geode; cyclone on IBM x440
645 [MIPS] MIPS
646 [PARISC] cr16
647 [S390] tod
648 [SH] SuperH
649 [SPARC64] tick
650 [X86-64] hpet,tsc
651
652 clearcpuid=BITNUM [X86]
653 Disable CPUID feature X for the kernel. See
654 arch/x86/include/asm/cpufeature.h for the valid bit
655 numbers. Note the Linux specific bits are not necessarily
656 stable over kernel options, but the vendor specific
657 ones should be.
658 Also note that user programs calling CPUID directly
659 or using the feature without checking anything
660 will still see it. This just prevents it from
661 being used by the kernel or shown in /proc/cpuinfo.
662 Also note the kernel might malfunction if you disable
663 some critical bits.
664
665 cma=nn[MG]@[start[MG][-end[MG]]]
666 [ARM,X86,KNL]
667 Sets the size of kernel global memory area for
668 contiguous memory allocations and optionally the
669 placement constraint by the physical address range of
670 memory allocations. A value of 0 disables CMA
671 altogether. For more information, see
672 include/linux/dma-contiguous.h
673
674 cmo_free_hint= [PPC] Format: { yes | no }
675 Specify whether pages are marked as being inactive
676 when they are freed. This is used in CMO environments
677 to determine OS memory pressure for page stealing by
678 a hypervisor.
679 Default: yes
680
681 coherent_pool=nn[KMG] [ARM,KNL]
682 Sets the size of memory pool for coherent, atomic dma
683 allocations, by default set to 256K.
684
685 code_bytes [X86] How many bytes of object code to print
686 in an oops report.
687 Range: 0 - 8192
688 Default: 64
689
690 com20020= [HW,NET] ARCnet - COM20020 chipset
691 Format:
692 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
693
694 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
695 Format: <io>[,<irq>]
696
697 com90xx= [HW,NET]
698 ARCnet - COM90xx chipset (memory-mapped buffers)
699 Format: <io>[,<irq>[,<memstart>]]
700
701 condev= [HW,S390] console device
702 conmode=
703
704 console= [KNL] Output console device and options.
705
706 tty<n> Use the virtual console device <n>.
707
708 ttyS<n>[,options]
709 ttyUSB0[,options]
710 Use the specified serial port. The options are of
711 the form "bbbbpnf", where "bbbb" is the baud rate,
712 "p" is parity ("n", "o", or "e"), "n" is number of
713 bits, and "f" is flow control ("r" for RTS or
714 omit it). Default is "9600n8".
715
716 See Documentation/serial-console.txt for more
717 information. See
718 Documentation/networking/netconsole.txt for an
719 alternative.
720
721 uart[8250],io,<addr>[,options]
722 uart[8250],mmio,<addr>[,options]
723 uart[8250],mmio32,<addr>[,options]
724 uart[8250],0x<addr>[,options]
725 Start an early, polled-mode console on the 8250/16550
726 UART at the specified I/O port or MMIO address,
727 switching to the matching ttyS device later.
728 MMIO inter-register address stride is either 8-bit
729 (mmio) or 32-bit (mmio32).
730 If none of [io|mmio|mmio32], <addr> is assumed to be
731 equivalent to 'mmio'. 'options' are specified in the
732 same format described for ttyS above; if unspecified,
733 the h/w is not re-initialized.
734
735 hvc<n> Use the hypervisor console device <n>. This is for
736 both Xen and PowerPC hypervisors.
737
738 If the device connected to the port is not a TTY but a braille
739 device, prepend "brl," before the device type, for instance
740 console=brl,ttyS0
741 For now, only VisioBraille is supported.
742
743 consoleblank= [KNL] The console blank (screen saver) timeout in
744 seconds. Defaults to 10*60 = 10mins. A value of 0
745 disables the blank timer.
746
747 coredump_filter=
748 [KNL] Change the default value for
749 /proc/<pid>/coredump_filter.
750 See also Documentation/filesystems/proc.txt.
751
752 cpuidle.off=1 [CPU_IDLE]
753 disable the cpuidle sub-system
754
755 cpu_init_udelay=N
756 [X86] Delay for N microsec between assert and de-assert
757 of APIC INIT to start processors. This delay occurs
758 on every CPU online, such as boot, and resume from suspend.
759 Default: 10000
760
761 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
762 Format:
763 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
764
765 crashkernel=size[KMG][@offset[KMG]]
766 [KNL] Using kexec, Linux can switch to a 'crash kernel'
767 upon panic. This parameter reserves the physical
768 memory region [offset, offset + size] for that kernel
769 image. If '@offset' is omitted, then a suitable offset
770 is selected automatically. Check
771 Documentation/kdump/kdump.txt for further details.
772
773 crashkernel=range1:size1[,range2:size2,...][@offset]
774 [KNL] Same as above, but depends on the memory
775 in the running system. The syntax of range is
776 start-[end] where start and end are both
777 a memory unit (amount[KMG]). See also
778 Documentation/kdump/kdump.txt for an example.
779
780 crashkernel=size[KMG],high
781 [KNL, x86_64] range could be above 4G. Allow kernel
782 to allocate physical memory region from top, so could
783 be above 4G if system have more than 4G ram installed.
784 Otherwise memory region will be allocated below 4G, if
785 available.
786 It will be ignored if crashkernel=X is specified.
787 crashkernel=size[KMG],low
788 [KNL, x86_64] range under 4G. When crashkernel=X,high
789 is passed, kernel could allocate physical memory region
790 above 4G, that cause second kernel crash on system
791 that require some amount of low memory, e.g. swiotlb
792 requires at least 64M+32K low memory, also enough extra
793 low memory is needed to make sure DMA buffers for 32-bit
794 devices won't run out. Kernel would try to allocate at
795 at least 256M below 4G automatically.
796 This one let user to specify own low range under 4G
797 for second kernel instead.
798 0: to disable low allocation.
799 It will be ignored when crashkernel=X,high is not used
800 or memory reserved is below 4G.
801
802 cs89x0_dma= [HW,NET]
803 Format: <dma>
804
805 cs89x0_media= [HW,NET]
806 Format: { rj45 | aui | bnc }
807
808 dasd= [HW,NET]
809 See header of drivers/s390/block/dasd_devmap.c.
810
811 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
812 (one device per port)
813 Format: <port#>,<type>
814 See also Documentation/input/joystick-parport.txt
815
816 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
817 time. See Documentation/dynamic-debug-howto.txt for
818 details. Deprecated, see dyndbg.
819
820 debug [KNL] Enable kernel debugging (events log level).
821
822 debug_locks_verbose=
823 [KNL] verbose self-tests
824 Format=<0|1>
825 Print debugging info while doing the locking API
826 self-tests.
827 We default to 0 (no extra messages), setting it to
828 1 will print _a lot_ more information - normally
829 only useful to kernel developers.
830
831 debug_objects [KNL] Enable object debugging
832
833 no_debug_objects
834 [KNL] Disable object debugging
835
836 debug_guardpage_minorder=
837 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
838 parameter allows control of the order of pages that will
839 be intentionally kept free (and hence protected) by the
840 buddy allocator. Bigger value increase the probability
841 of catching random memory corruption, but reduce the
842 amount of memory for normal system use. The maximum
843 possible value is MAX_ORDER/2. Setting this parameter
844 to 1 or 2 should be enough to identify most random
845 memory corruption problems caused by bugs in kernel or
846 driver code when a CPU writes to (or reads from) a
847 random memory location. Note that there exists a class
848 of memory corruptions problems caused by buggy H/W or
849 F/W or by drivers badly programing DMA (basically when
850 memory is written at bus level and the CPU MMU is
851 bypassed) which are not detectable by
852 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
853 tracking down these problems.
854
855 debug_pagealloc=
856 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
857 parameter enables the feature at boot time. In
858 default, it is disabled. We can avoid allocating huge
859 chunk of memory for debug pagealloc if we don't enable
860 it at boot time and the system will work mostly same
861 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
862 on: enable the feature
863
864 debugpat [X86] Enable PAT debugging
865
866 decnet.addr= [HW,NET]
867 Format: <area>[,<node>]
868 See also Documentation/networking/decnet.txt.
869
870 default_hugepagesz=
871 [same as hugepagesz=] The size of the default
872 HugeTLB page size. This is the size represented by
873 the legacy /proc/ hugepages APIs, used for SHM, and
874 default size when mounting hugetlbfs filesystems.
875 Defaults to the default architecture's huge page size
876 if not specified.
877
878 dhash_entries= [KNL]
879 Set number of hash buckets for dentry cache.
880
881 disable= [IPV6]
882 See Documentation/networking/ipv6.txt.
883
884 disable_cpu_apicid= [X86,APIC,SMP]
885 Format: <int>
886 The number of initial APIC ID for the
887 corresponding CPU to be disabled at boot,
888 mostly used for the kdump 2nd kernel to
889 disable BSP to wake up multiple CPUs without
890 causing system reset or hang due to sending
891 INIT from AP to BSP.
892
893 disable_ddw [PPC/PSERIES]
894 Disable Dynamic DMA Window support. Use this if
895 to workaround buggy firmware.
896
897 disable_ipv6= [IPV6]
898 See Documentation/networking/ipv6.txt.
899
900 disable_mtrr_cleanup [X86]
901 The kernel tries to adjust MTRR layout from continuous
902 to discrete, to make X server driver able to add WB
903 entry later. This parameter disables that.
904
905 disable_mtrr_trim [X86, Intel and AMD only]
906 By default the kernel will trim any uncacheable
907 memory out of your available memory pool based on
908 MTRR settings. This parameter disables that behavior,
909 possibly causing your machine to run very slowly.
910
911 disable_timer_pin_1 [X86]
912 Disable PIN 1 of APIC timer
913 Can be useful to work around chipset bugs.
914
915 dis_ucode_ldr [X86] Disable the microcode loader.
916
917 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
918 this option disables the debugging code at boot.
919
920 dma_debug_entries=<number>
921 This option allows to tune the number of preallocated
922 entries for DMA-API debugging code. One entry is
923 required per DMA-API allocation. Use this if the
924 DMA-API debugging code disables itself because the
925 architectural default is too low.
926
927 dma_debug_driver=<driver_name>
928 With this option the DMA-API debugging driver
929 filter feature can be enabled at boot time. Just
930 pass the driver to filter for as the parameter.
931 The filter can be disabled or changed to another
932 driver later using sysfs.
933
934 drm_kms_helper.edid_firmware=[<connector>:]<file>
935 Broken monitors, graphic adapters and KVMs may
936 send no or incorrect EDID data sets. This parameter
937 allows to specify an EDID data set in the
938 /lib/firmware directory that is used instead.
939 Generic built-in EDID data sets are used, if one of
940 edid/1024x768.bin, edid/1280x1024.bin,
941 edid/1680x1050.bin, or edid/1920x1080.bin is given
942 and no file with the same name exists. Details and
943 instructions how to build your own EDID data are
944 available in Documentation/EDID/HOWTO.txt. An EDID
945 data set will only be used for a particular connector,
946 if its name and a colon are prepended to the EDID
947 name.
948
949 dscc4.setup= [NET]
950
951 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
952 module.dyndbg[="val"]
953 Enable debug messages at boot time. See
954 Documentation/dynamic-debug-howto.txt for details.
955
956 nompx [X86] Disables Intel Memory Protection Extensions.
957 See Documentation/x86/intel_mpx.txt for more
958 information about the feature.
959
960 eagerfpu= [X86]
961 on enable eager fpu restore
962 off disable eager fpu restore
963 auto selects the default scheme, which automatically
964 enables eagerfpu restore for xsaveopt.
965
966 module.async_probe [KNL]
967 Enable asynchronous probe on this module.
968
969 early_ioremap_debug [KNL]
970 Enable debug messages in early_ioremap support. This
971 is useful for tracking down temporary early mappings
972 which are not unmapped.
973
974 earlycon= [KNL] Output early console device and options.
975
976 cdns,<addr>
977 Start an early, polled-mode console on a cadence serial
978 port at the specified address. The cadence serial port
979 must already be setup and configured. Options are not
980 yet supported.
981
982 uart[8250],io,<addr>[,options]
983 uart[8250],mmio,<addr>[,options]
984 uart[8250],mmio32,<addr>[,options]
985 uart[8250],mmio32be,<addr>[,options]
986 uart[8250],0x<addr>[,options]
987 Start an early, polled-mode console on the 8250/16550
988 UART at the specified I/O port or MMIO address.
989 MMIO inter-register address stride is either 8-bit
990 (mmio) or 32-bit (mmio32 or mmio32be).
991 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
992 to be equivalent to 'mmio'. 'options' are specified
993 in the same format described for "console=ttyS<n>"; if
994 unspecified, the h/w is not initialized.
995
996 pl011,<addr>
997 Start an early, polled-mode console on a pl011 serial
998 port at the specified address. The pl011 serial port
999 must already be setup and configured. Options are not
1000 yet supported.
1001
1002 msm_serial,<addr>
1003 Start an early, polled-mode console on an msm serial
1004 port at the specified address. The serial port
1005 must already be setup and configured. Options are not
1006 yet supported.
1007
1008 msm_serial_dm,<addr>
1009 Start an early, polled-mode console on an msm serial
1010 dm port at the specified address. The serial port
1011 must already be setup and configured. Options are not
1012 yet supported.
1013
1014 smh Use ARM semihosting calls for early console.
1015
1016 s3c2410,<addr>
1017 s3c2412,<addr>
1018 s3c2440,<addr>
1019 s3c6400,<addr>
1020 s5pv210,<addr>
1021 exynos4210,<addr>
1022 Use early console provided by serial driver available
1023 on Samsung SoCs, requires selecting proper type and
1024 a correct base address of the selected UART port. The
1025 serial port must already be setup and configured.
1026 Options are not yet supported.
1027
1028 earlyprintk= [X86,SH,BLACKFIN,ARM,M68k]
1029 earlyprintk=vga
1030 earlyprintk=efi
1031 earlyprintk=xen
1032 earlyprintk=serial[,ttySn[,baudrate]]
1033 earlyprintk=serial[,0x...[,baudrate]]
1034 earlyprintk=ttySn[,baudrate]
1035 earlyprintk=dbgp[debugController#]
1036 earlyprintk=pciserial,bus:device.function[,baudrate]
1037
1038 earlyprintk is useful when the kernel crashes before
1039 the normal console is initialized. It is not enabled by
1040 default because it has some cosmetic problems.
1041
1042 Append ",keep" to not disable it when the real console
1043 takes over.
1044
1045 Only one of vga, efi, serial, or usb debug port can
1046 be used at a time.
1047
1048 Currently only ttyS0 and ttyS1 may be specified by
1049 name. Other I/O ports may be explicitly specified
1050 on some architectures (x86 and arm at least) by
1051 replacing ttySn with an I/O port address, like this:
1052 earlyprintk=serial,0x1008,115200
1053 You can find the port for a given device in
1054 /proc/tty/driver/serial:
1055 2: uart:ST16650V2 port:00001008 irq:18 ...
1056
1057 Interaction with the standard serial driver is not
1058 very good.
1059
1060 The VGA and EFI output is eventually overwritten by
1061 the real console.
1062
1063 The xen output can only be used by Xen PV guests.
1064
1065 edac_report= [HW,EDAC] Control how to report EDAC event
1066 Format: {"on" | "off" | "force"}
1067 on: enable EDAC to report H/W event. May be overridden
1068 by other higher priority error reporting module.
1069 off: disable H/W event reporting through EDAC.
1070 force: enforce the use of EDAC to report H/W event.
1071 default: on.
1072
1073 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1074 ekgdboc=kbd
1075
1076 This is designed to be used in conjunction with
1077 the boot argument: earlyprintk=vga
1078
1079 edd= [EDD]
1080 Format: {"off" | "on" | "skip[mbr]"}
1081
1082 efi= [EFI]
1083 Format: { "old_map", "nochunk", "noruntime", "debug" }
1084 old_map [X86-64]: switch to the old ioremap-based EFI
1085 runtime services mapping. 32-bit still uses this one by
1086 default.
1087 nochunk: disable reading files in "chunks" in the EFI
1088 boot stub, as chunking can cause problems with some
1089 firmware implementations.
1090 noruntime : disable EFI runtime services support
1091 debug: enable misc debug output
1092
1093 efi_no_storage_paranoia [EFI; X86]
1094 Using this parameter you can use more than 50% of
1095 your efi variable storage. Use this parameter only if
1096 you are really sure that your UEFI does sane gc and
1097 fulfills the spec otherwise your board may brick.
1098
1099 eisa_irq_edge= [PARISC,HW]
1100 See header of drivers/parisc/eisa.c.
1101
1102 elanfreq= [X86-32]
1103 See comment before function elanfreq_setup() in
1104 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1105
1106 elevator= [IOSCHED]
1107 Format: {"cfq" | "deadline" | "noop"}
1108 See Documentation/block/cfq-iosched.txt and
1109 Documentation/block/deadline-iosched.txt for details.
1110
1111 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1112 Specifies physical address of start of kernel core
1113 image elf header and optionally the size. Generally
1114 kexec loader will pass this option to capture kernel.
1115 See Documentation/kdump/kdump.txt for details.
1116
1117 enable_mtrr_cleanup [X86]
1118 The kernel tries to adjust MTRR layout from continuous
1119 to discrete, to make X server driver able to add WB
1120 entry later. This parameter enables that.
1121
1122 enable_timer_pin_1 [X86]
1123 Enable PIN 1 of APIC timer
1124 Can be useful to work around chipset bugs
1125 (in particular on some ATI chipsets).
1126 The kernel tries to set a reasonable default.
1127
1128 enforcing [SELINUX] Set initial enforcing status.
1129 Format: {"0" | "1"}
1130 See security/selinux/Kconfig help text.
1131 0 -- permissive (log only, no denials).
1132 1 -- enforcing (deny and log).
1133 Default value is 0.
1134 Value can be changed at runtime via /selinux/enforce.
1135
1136 erst_disable [ACPI]
1137 Disable Error Record Serialization Table (ERST)
1138 support.
1139
1140 ether= [HW,NET] Ethernet cards parameters
1141 This option is obsoleted by the "netdev=" option, which
1142 has equivalent usage. See its documentation for details.
1143
1144 evm= [EVM]
1145 Format: { "fix" }
1146 Permit 'security.evm' to be updated regardless of
1147 current integrity status.
1148
1149 failslab=
1150 fail_page_alloc=
1151 fail_make_request=[KNL]
1152 General fault injection mechanism.
1153 Format: <interval>,<probability>,<space>,<times>
1154 See also Documentation/fault-injection/.
1155
1156 floppy= [HW]
1157 See Documentation/blockdev/floppy.txt.
1158
1159 force_pal_cache_flush
1160 [IA-64] Avoid check_sal_cache_flush which may hang on
1161 buggy SAL_CACHE_FLUSH implementations. Using this
1162 parameter will force ia64_sal_cache_flush to call
1163 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1164
1165 forcepae [X86-32]
1166 Forcefully enable Physical Address Extension (PAE).
1167 Many Pentium M systems disable PAE but may have a
1168 functionally usable PAE implementation.
1169 Warning: use of this parameter will taint the kernel
1170 and may cause unknown problems.
1171
1172 ftrace=[tracer]
1173 [FTRACE] will set and start the specified tracer
1174 as early as possible in order to facilitate early
1175 boot debugging.
1176
1177 ftrace_dump_on_oops[=orig_cpu]
1178 [FTRACE] will dump the trace buffers on oops.
1179 If no parameter is passed, ftrace will dump
1180 buffers of all CPUs, but if you pass orig_cpu, it will
1181 dump only the buffer of the CPU that triggered the
1182 oops.
1183
1184 ftrace_filter=[function-list]
1185 [FTRACE] Limit the functions traced by the function
1186 tracer at boot up. function-list is a comma separated
1187 list of functions. This list can be changed at run
1188 time by the set_ftrace_filter file in the debugfs
1189 tracing directory.
1190
1191 ftrace_notrace=[function-list]
1192 [FTRACE] Do not trace the functions specified in
1193 function-list. This list can be changed at run time
1194 by the set_ftrace_notrace file in the debugfs
1195 tracing directory.
1196
1197 ftrace_graph_filter=[function-list]
1198 [FTRACE] Limit the top level callers functions traced
1199 by the function graph tracer at boot up.
1200 function-list is a comma separated list of functions
1201 that can be changed at run time by the
1202 set_graph_function file in the debugfs tracing directory.
1203
1204 ftrace_graph_notrace=[function-list]
1205 [FTRACE] Do not trace from the functions specified in
1206 function-list. This list is a comma separated list of
1207 functions that can be changed at run time by the
1208 set_graph_notrace file in the debugfs tracing directory.
1209
1210 gamecon.map[2|3]=
1211 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1212 support via parallel port (up to 5 devices per port)
1213 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1214 See also Documentation/input/joystick-parport.txt
1215
1216 gamma= [HW,DRM]
1217
1218 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1219 Format: off | on
1220 default: on
1221
1222 gcov_persist= [GCOV] When non-zero (default), profiling data for
1223 kernel modules is saved and remains accessible via
1224 debugfs, even when the module is unloaded/reloaded.
1225 When zero, profiling data is discarded and associated
1226 debugfs files are removed at module unload time.
1227
1228 gpt [EFI] Forces disk with valid GPT signature but
1229 invalid Protective MBR to be treated as GPT. If the
1230 primary GPT is corrupted, it enables the backup/alternate
1231 GPT to be used instead.
1232
1233 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1234 the "Enable 0" bit of the configuration register.
1235 Format: 0 | 1
1236 Default: 0
1237 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1238 the "Enable 0" bit of the configuration register.
1239 Format: 0 | 1
1240 Default: 0
1241 grcan.select= [HW] Select which physical interface to use.
1242 Format: 0 | 1
1243 Default: 0
1244 grcan.txsize= [HW] Sets the size of the tx buffer.
1245 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1246 Default: 1024
1247 grcan.rxsize= [HW] Sets the size of the rx buffer.
1248 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1249 Default: 1024
1250
1251 hashdist= [KNL,NUMA] Large hashes allocated during boot
1252 are distributed across NUMA nodes. Defaults on
1253 for 64-bit NUMA, off otherwise.
1254 Format: 0 | 1 (for off | on)
1255
1256 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1257
1258 hd= [EIDE] (E)IDE hard drive subsystem geometry
1259 Format: <cyl>,<head>,<sect>
1260
1261 hest_disable [ACPI]
1262 Disable Hardware Error Source Table (HEST) support;
1263 corresponding firmware-first mode error processing
1264 logic will be disabled.
1265
1266 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1267 size of <nn>. This works even on boxes that have no
1268 highmem otherwise. This also works to reduce highmem
1269 size on bigger boxes.
1270
1271 highres= [KNL] Enable/disable high resolution timer mode.
1272 Valid parameters: "on", "off"
1273 Default: "on"
1274
1275 hisax= [HW,ISDN]
1276 See Documentation/isdn/README.HiSax.
1277
1278 hlt [BUGS=ARM,SH]
1279
1280 hpet= [X86-32,HPET] option to control HPET usage
1281 Format: { enable (default) | disable | force |
1282 verbose }
1283 disable: disable HPET and use PIT instead
1284 force: allow force enabled of undocumented chips (ICH4,
1285 VIA, nVidia)
1286 verbose: show contents of HPET registers during setup
1287
1288 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1289 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1290
1291 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1292 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1293 On x86-64 and powerpc, this option can be specified
1294 multiple times interleaved with hugepages= to reserve
1295 huge pages of different sizes. Valid pages sizes on
1296 x86-64 are 2M (when the CPU supports "pse") and 1G
1297 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1298
1299 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1300 terminal devices. Valid values: 0..8
1301 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1302 If specified, z/VM IUCV HVC accepts connections
1303 from listed z/VM user IDs only.
1304
1305 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1306 hardware thread id mappings.
1307 Format: <cpu>:<hwthread>
1308
1309 keep_bootcon [KNL]
1310 Do not unregister boot console at start. This is only
1311 useful for debugging when something happens in the window
1312 between unregistering the boot console and initializing
1313 the real console.
1314
1315 i2c_bus= [HW] Override the default board specific I2C bus speed
1316 or register an additional I2C bus that is not
1317 registered from board initialization code.
1318 Format:
1319 <bus_id>,<clkrate>
1320
1321 i8042.debug [HW] Toggle i8042 debug mode
1322 i8042.unmask_kbd_data
1323 [HW] Enable printing of interrupt data from the KBD port
1324 (disabled by default, and as a pre-condition
1325 requires that i8042.debug=1 be enabled)
1326 i8042.direct [HW] Put keyboard port into non-translated mode
1327 i8042.dumbkbd [HW] Pretend that controller can only read data from
1328 keyboard and cannot control its state
1329 (Don't attempt to blink the leds)
1330 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1331 i8042.nokbd [HW] Don't check/create keyboard port
1332 i8042.noloop [HW] Disable the AUX Loopback command while probing
1333 for the AUX port
1334 i8042.nomux [HW] Don't check presence of an active multiplexing
1335 controller
1336 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1337 controllers
1338 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1339 i8042.reset [HW] Reset the controller during init and cleanup
1340 i8042.unlock [HW] Unlock (ignore) the keylock
1341 i8042.kbdreset [HW] Reset device connected to KBD port
1342
1343 i810= [HW,DRM]
1344
1345 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1346 indicates that the driver is running on unsupported
1347 hardware.
1348 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1349 does not match list of supported models.
1350 i8k.power_status
1351 [HW] Report power status in /proc/i8k
1352 (disabled by default)
1353 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1354 capability is set.
1355
1356 i915.invert_brightness=
1357 [DRM] Invert the sense of the variable that is used to
1358 set the brightness of the panel backlight. Normally a
1359 brightness value of 0 indicates backlight switched off,
1360 and the maximum of the brightness value sets the backlight
1361 to maximum brightness. If this parameter is set to 0
1362 (default) and the machine requires it, or this parameter
1363 is set to 1, a brightness value of 0 sets the backlight
1364 to maximum brightness, and the maximum of the brightness
1365 value switches the backlight off.
1366 -1 -- never invert brightness
1367 0 -- machine default
1368 1 -- force brightness inversion
1369
1370 icn= [HW,ISDN]
1371 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1372
1373 ide-core.nodma= [HW] (E)IDE subsystem
1374 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1375 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1376 .cdrom .chs .ignore_cable are additional options
1377 See Documentation/ide/ide.txt.
1378
1379 ide-generic.probe-mask= [HW] (E)IDE subsystem
1380 Format: <int>
1381 Probe mask for legacy ISA IDE ports. Depending on
1382 platform up to 6 ports are supported, enabled by
1383 setting corresponding bits in the mask to 1. The
1384 default value is 0x0, which has a special meaning.
1385 On systems that have PCI, it triggers scanning the
1386 PCI bus for the first and the second port, which
1387 are then probed. On systems without PCI the value
1388 of 0x0 enables probing the two first ports as if it
1389 was 0x3.
1390
1391 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1392 Claim all unknown PCI IDE storage controllers.
1393
1394 idle= [X86]
1395 Format: idle=poll, idle=halt, idle=nomwait
1396 Poll forces a polling idle loop that can slightly
1397 improve the performance of waking up a idle CPU, but
1398 will use a lot of power and make the system run hot.
1399 Not recommended.
1400 idle=halt: Halt is forced to be used for CPU idle.
1401 In such case C2/C3 won't be used again.
1402 idle=nomwait: Disable mwait for CPU C-states
1403
1404 ignore_loglevel [KNL]
1405 Ignore loglevel setting - this will print /all/
1406 kernel messages to the console. Useful for debugging.
1407 We also add it as printk module parameter, so users
1408 could change it dynamically, usually by
1409 /sys/module/printk/parameters/ignore_loglevel.
1410
1411 ihash_entries= [KNL]
1412 Set number of hash buckets for inode cache.
1413
1414 ima_appraise= [IMA] appraise integrity measurements
1415 Format: { "off" | "enforce" | "fix" | "log" }
1416 default: "enforce"
1417
1418 ima_appraise_tcb [IMA]
1419 The builtin appraise policy appraises all files
1420 owned by uid=0.
1421
1422 ima_hash= [IMA]
1423 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1424 | sha512 | ... }
1425 default: "sha1"
1426
1427 The list of supported hash algorithms is defined
1428 in crypto/hash_info.h.
1429
1430 ima_policy= [IMA]
1431 The builtin measurement policy to load during IMA
1432 setup. Specyfing "tcb" as the value, measures all
1433 programs exec'd, files mmap'd for exec, and all files
1434 opened with the read mode bit set by either the
1435 effective uid (euid=0) or uid=0.
1436 Format: "tcb"
1437
1438 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1439 Load a policy which meets the needs of the Trusted
1440 Computing Base. This means IMA will measure all
1441 programs exec'd, files mmap'd for exec, and all files
1442 opened for read by uid=0.
1443
1444 ima_template= [IMA]
1445 Select one of defined IMA measurements template formats.
1446 Formats: { "ima" | "ima-ng" | "ima-sig" }
1447 Default: "ima-ng"
1448
1449 ima_template_fmt=
1450 [IMA] Define a custom template format.
1451 Format: { "field1|...|fieldN" }
1452
1453 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1454 Format: <min_file_size>
1455 Set the minimal file size for using asynchronous hash.
1456 If left unspecified, ahash usage is disabled.
1457
1458 ahash performance varies for different data sizes on
1459 different crypto accelerators. This option can be used
1460 to achieve the best performance for a particular HW.
1461
1462 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1463 Format: <bufsize>
1464 Set hashing buffer size. Default: 4k.
1465
1466 ahash performance varies for different chunk sizes on
1467 different crypto accelerators. This option can be used
1468 to achieve best performance for particular HW.
1469
1470 init= [KNL]
1471 Format: <full_path>
1472 Run specified binary instead of /sbin/init as init
1473 process.
1474
1475 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1476 for working out where the kernel is dying during
1477 startup.
1478
1479 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1480 initcall functions. Useful for debugging built-in
1481 modules and initcalls.
1482
1483 initrd= [BOOT] Specify the location of the initial ramdisk
1484
1485 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1486 Format: <irq>
1487
1488 int_pln_enable [x86] Enable power limit notification interrupt
1489
1490 integrity_audit=[IMA]
1491 Format: { "0" | "1" }
1492 0 -- basic integrity auditing messages. (Default)
1493 1 -- additional integrity auditing messages.
1494
1495 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1496 on
1497 Enable intel iommu driver.
1498 off
1499 Disable intel iommu driver.
1500 igfx_off [Default Off]
1501 By default, gfx is mapped as normal device. If a gfx
1502 device has a dedicated DMAR unit, the DMAR unit is
1503 bypassed by not enabling DMAR with this option. In
1504 this case, gfx device will use physical address for
1505 DMA.
1506 forcedac [x86_64]
1507 With this option iommu will not optimize to look
1508 for io virtual address below 32-bit forcing dual
1509 address cycle on pci bus for cards supporting greater
1510 than 32-bit addressing. The default is to look
1511 for translation below 32-bit and if not available
1512 then look in the higher range.
1513 strict [Default Off]
1514 With this option on every unmap_single operation will
1515 result in a hardware IOTLB flush operation as opposed
1516 to batching them for performance.
1517 sp_off [Default Off]
1518 By default, super page will be supported if Intel IOMMU
1519 has the capability. With this option, super page will
1520 not be supported.
1521 ecs_off [Default Off]
1522 By default, extended context tables will be supported if
1523 the hardware advertises that it has support both for the
1524 extended tables themselves, and also PASID support. With
1525 this option set, extended tables will not be used even
1526 on hardware which claims to support them.
1527
1528 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1529 0 disables intel_idle and fall back on acpi_idle.
1530 1 to 6 specify maximum depth of C-state.
1531
1532 intel_pstate= [X86]
1533 disable
1534 Do not enable intel_pstate as the default
1535 scaling driver for the supported processors
1536 force
1537 Enable intel_pstate on systems that prohibit it by default
1538 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1539 instead of acpi-cpufreq may disable platform features, such
1540 as thermal controls and power capping, that rely on ACPI
1541 P-States information being indicated to OSPM and therefore
1542 should be used with caution. This option does not work with
1543 processors that aren't supported by the intel_pstate driver
1544 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1545 no_hwp
1546 Do not enable hardware P state control (HWP)
1547 if available.
1548 hwp_only
1549 Only load intel_pstate on systems which support
1550 hardware P state control (HWP) if available.
1551
1552 intremap= [X86-64, Intel-IOMMU]
1553 on enable Interrupt Remapping (default)
1554 off disable Interrupt Remapping
1555 nosid disable Source ID checking
1556 no_x2apic_optout
1557 BIOS x2APIC opt-out request will be ignored
1558
1559 iomem= Disable strict checking of access to MMIO memory
1560 strict regions from userspace.
1561 relaxed
1562
1563 iommu= [x86]
1564 off
1565 force
1566 noforce
1567 biomerge
1568 panic
1569 nopanic
1570 merge
1571 nomerge
1572 forcesac
1573 soft
1574 pt [x86, IA-64]
1575 nobypass [PPC/POWERNV]
1576 Disable IOMMU bypass, using IOMMU for PCI devices.
1577
1578
1579 io7= [HW] IO7 for Marvel based alpha systems
1580 See comment before marvel_specify_io7 in
1581 arch/alpha/kernel/core_marvel.c.
1582
1583 io_delay= [X86] I/O delay method
1584 0x80
1585 Standard port 0x80 based delay
1586 0xed
1587 Alternate port 0xed based delay (needed on some systems)
1588 udelay
1589 Simple two microseconds delay
1590 none
1591 No delay
1592
1593 ip= [IP_PNP]
1594 See Documentation/filesystems/nfs/nfsroot.txt.
1595
1596 irqfixup [HW]
1597 When an interrupt is not handled search all handlers
1598 for it. Intended to get systems with badly broken
1599 firmware running.
1600
1601 irqpoll [HW]
1602 When an interrupt is not handled search all handlers
1603 for it. Also check all handlers each timer
1604 interrupt. Intended to get systems with badly broken
1605 firmware running.
1606
1607 isapnp= [ISAPNP]
1608 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1609
1610 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1611 Format:
1612 <cpu number>,...,<cpu number>
1613 or
1614 <cpu number>-<cpu number>
1615 (must be a positive range in ascending order)
1616 or a mixture
1617 <cpu number>,...,<cpu number>-<cpu number>
1618
1619 This option can be used to specify one or more CPUs
1620 to isolate from the general SMP balancing and scheduling
1621 algorithms. You can move a process onto or off an
1622 "isolated" CPU via the CPU affinity syscalls or cpuset.
1623 <cpu number> begins at 0 and the maximum value is
1624 "number of CPUs in system - 1".
1625
1626 This option is the preferred way to isolate CPUs. The
1627 alternative -- manually setting the CPU mask of all
1628 tasks in the system -- can cause problems and
1629 suboptimal load balancer performance.
1630
1631 iucv= [HW,NET]
1632
1633 ivrs_ioapic [HW,X86_64]
1634 Provide an override to the IOAPIC-ID<->DEVICE-ID
1635 mapping provided in the IVRS ACPI table. For
1636 example, to map IOAPIC-ID decimal 10 to
1637 PCI device 00:14.0 write the parameter as:
1638 ivrs_ioapic[10]=00:14.0
1639
1640 ivrs_hpet [HW,X86_64]
1641 Provide an override to the HPET-ID<->DEVICE-ID
1642 mapping provided in the IVRS ACPI table. For
1643 example, to map HPET-ID decimal 0 to
1644 PCI device 00:14.0 write the parameter as:
1645 ivrs_hpet[0]=00:14.0
1646
1647 js= [HW,JOY] Analog joystick
1648 See Documentation/input/joystick.txt.
1649
1650 kaslr/nokaslr [X86]
1651 Enable/disable kernel and module base offset ASLR
1652 (Address Space Layout Randomization) if built into
1653 the kernel. When CONFIG_HIBERNATION is selected,
1654 kASLR is disabled by default. When kASLR is enabled,
1655 hibernation will be disabled.
1656
1657 keepinitrd [HW,ARM]
1658
1659 kernelcore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
1660 specifies the amount of memory usable by the kernel
1661 for non-movable allocations. The requested amount is
1662 spread evenly throughout all nodes in the system. The
1663 remaining memory in each node is used for Movable
1664 pages. In the event, a node is too small to have both
1665 kernelcore and Movable pages, kernelcore pages will
1666 take priority and other nodes will have a larger number
1667 of Movable pages. The Movable zone is used for the
1668 allocation of pages that may be reclaimed or moved
1669 by the page migration subsystem. This means that
1670 HugeTLB pages may not be allocated from this zone.
1671 Note that allocations like PTEs-from-HighMem still
1672 use the HighMem zone if it exists, and the Normal
1673 zone if it does not.
1674
1675 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1676 Format: <Controller#>[,poll interval]
1677 The controller # is the number of the ehci usb debug
1678 port as it is probed via PCI. The poll interval is
1679 optional and is the number seconds in between
1680 each poll cycle to the debug port in case you need
1681 the functionality for interrupting the kernel with
1682 gdb or control-c on the dbgp connection. When
1683 not using this parameter you use sysrq-g to break into
1684 the kernel debugger.
1685
1686 kgdboc= [KGDB,HW] kgdb over consoles.
1687 Requires a tty driver that supports console polling,
1688 or a supported polling keyboard driver (non-usb).
1689 Serial only format: <serial_device>[,baud]
1690 keyboard only format: kbd
1691 keyboard and serial format: kbd,<serial_device>[,baud]
1692 Optional Kernel mode setting:
1693 kms, kbd format: kms,kbd
1694 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1695
1696 kgdbwait [KGDB] Stop kernel execution and enter the
1697 kernel debugger at the earliest opportunity.
1698
1699 kmac= [MIPS] korina ethernet MAC address.
1700 Configure the RouterBoard 532 series on-chip
1701 Ethernet adapter MAC address.
1702
1703 kmemleak= [KNL] Boot-time kmemleak enable/disable
1704 Valid arguments: on, off
1705 Default: on
1706 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1707 the default is off.
1708
1709 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1710 Valid arguments: 0, 1, 2
1711 kmemcheck=0 (disabled)
1712 kmemcheck=1 (enabled)
1713 kmemcheck=2 (one-shot mode)
1714 Default: 2 (one-shot mode)
1715
1716 kstack=N [X86] Print N words from the kernel stack
1717 in oops dumps.
1718
1719 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1720 Default is 0 (don't ignore, but inject #GP)
1721
1722 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1723 KVM MMU at runtime.
1724 Default is 0 (off)
1725
1726 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1727 Default is 1 (enabled)
1728
1729 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1730 for all guests.
1731 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1732
1733 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1734 (virtualized MMU) support on capable Intel chips.
1735 Default is 1 (enabled)
1736
1737 kvm-intel.emulate_invalid_guest_state=
1738 [KVM,Intel] Enable emulation of invalid guest states
1739 Default is 0 (disabled)
1740
1741 kvm-intel.flexpriority=
1742 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1743 Default is 1 (enabled)
1744
1745 kvm-intel.nested=
1746 [KVM,Intel] Enable VMX nesting (nVMX).
1747 Default is 0 (disabled)
1748
1749 kvm-intel.unrestricted_guest=
1750 [KVM,Intel] Disable unrestricted guest feature
1751 (virtualized real and unpaged mode) on capable
1752 Intel chips. Default is 1 (enabled)
1753
1754 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1755 feature (tagged TLBs) on capable Intel chips.
1756 Default is 1 (enabled)
1757
1758 l2cr= [PPC]
1759
1760 l3cr= [PPC]
1761
1762 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1763 disabled it.
1764
1765 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1766 value for LAPIC timer one-shot implementation. Default
1767 back to the programmable timer unit in the LAPIC.
1768
1769 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1770 in C2 power state.
1771
1772 libata.dma= [LIBATA] DMA control
1773 libata.dma=0 Disable all PATA and SATA DMA
1774 libata.dma=1 PATA and SATA Disk DMA only
1775 libata.dma=2 ATAPI (CDROM) DMA only
1776 libata.dma=4 Compact Flash DMA only
1777 Combinations also work, so libata.dma=3 enables DMA
1778 for disks and CDROMs, but not CFs.
1779
1780 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1781 libata.ignore_hpa=0 keep BIOS limits (default)
1782 libata.ignore_hpa=1 ignore limits, using full disk
1783
1784 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1785 when set.
1786 Format: <int>
1787
1788 libata.force= [LIBATA] Force configurations. The format is comma
1789 separated list of "[ID:]VAL" where ID is
1790 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
1791 matching port, link or device. Basically, it matches
1792 the ATA ID string printed on console by libata. If
1793 the whole ID part is omitted, the last PORT and DEVICE
1794 values are used. If ID hasn't been specified yet, the
1795 configuration applies to all ports, links and devices.
1796
1797 If only DEVICE is omitted, the parameter applies to
1798 the port and all links and devices behind it. DEVICE
1799 number of 0 either selects the first device or the
1800 first fan-out link behind PMP device. It does not
1801 select the host link. DEVICE number of 15 selects the
1802 host link and device attached to it.
1803
1804 The VAL specifies the configuration to force. As long
1805 as there's no ambiguity shortcut notation is allowed.
1806 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1807 The following configurations can be forced.
1808
1809 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1810 Any ID with matching PORT is used.
1811
1812 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1813
1814 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1815 udma[/][16,25,33,44,66,100,133] notation is also
1816 allowed.
1817
1818 * [no]ncq: Turn on or off NCQ.
1819
1820 * [no]ncqtrim: Turn off queued DSM TRIM.
1821
1822 * nohrst, nosrst, norst: suppress hard, soft
1823 and both resets.
1824
1825 * rstonce: only attempt one reset during
1826 hot-unplug link recovery
1827
1828 * dump_id: dump IDENTIFY data.
1829
1830 * atapi_dmadir: Enable ATAPI DMADIR bridge support
1831
1832 * disable: Disable this device.
1833
1834 If there are multiple matching configurations changing
1835 the same attribute, the last one is used.
1836
1837 memblock=debug [KNL] Enable memblock debug messages.
1838
1839 load_ramdisk= [RAM] List of ramdisks to load from floppy
1840 See Documentation/blockdev/ramdisk.txt.
1841
1842 lockd.nlm_grace_period=P [NFS] Assign grace period.
1843 Format: <integer>
1844
1845 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1846 Format: <integer>
1847
1848 lockd.nlm_timeout=T [NFS] Assign timeout value.
1849 Format: <integer>
1850
1851 lockd.nlm_udpport=M [NFS] Assign UDP port.
1852 Format: <integer>
1853
1854 locktorture.nreaders_stress= [KNL]
1855 Set the number of locking read-acquisition kthreads.
1856 Defaults to being automatically set based on the
1857 number of online CPUs.
1858
1859 locktorture.nwriters_stress= [KNL]
1860 Set the number of locking write-acquisition kthreads.
1861
1862 locktorture.onoff_holdoff= [KNL]
1863 Set time (s) after boot for CPU-hotplug testing.
1864
1865 locktorture.onoff_interval= [KNL]
1866 Set time (s) between CPU-hotplug operations, or
1867 zero to disable CPU-hotplug testing.
1868
1869 locktorture.shuffle_interval= [KNL]
1870 Set task-shuffle interval (jiffies). Shuffling
1871 tasks allows some CPUs to go into dyntick-idle
1872 mode during the locktorture test.
1873
1874 locktorture.shutdown_secs= [KNL]
1875 Set time (s) after boot system shutdown. This
1876 is useful for hands-off automated testing.
1877
1878 locktorture.stat_interval= [KNL]
1879 Time (s) between statistics printk()s.
1880
1881 locktorture.stutter= [KNL]
1882 Time (s) to stutter testing, for example,
1883 specifying five seconds causes the test to run for
1884 five seconds, wait for five seconds, and so on.
1885 This tests the locking primitive's ability to
1886 transition abruptly to and from idle.
1887
1888 locktorture.torture_runnable= [BOOT]
1889 Start locktorture running at boot time.
1890
1891 locktorture.torture_type= [KNL]
1892 Specify the locking implementation to test.
1893
1894 locktorture.verbose= [KNL]
1895 Enable additional printk() statements.
1896
1897 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
1898 Format: <irq>
1899
1900 loglevel= All Kernel Messages with a loglevel smaller than the
1901 console loglevel will be printed to the console. It can
1902 also be changed with klogd or other programs. The
1903 loglevels are defined as follows:
1904
1905 0 (KERN_EMERG) system is unusable
1906 1 (KERN_ALERT) action must be taken immediately
1907 2 (KERN_CRIT) critical conditions
1908 3 (KERN_ERR) error conditions
1909 4 (KERN_WARNING) warning conditions
1910 5 (KERN_NOTICE) normal but significant condition
1911 6 (KERN_INFO) informational
1912 7 (KERN_DEBUG) debug-level messages
1913
1914 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
1915 in bytes. n must be a power of two and greater
1916 than the minimal size. The minimal size is defined
1917 by LOG_BUF_SHIFT kernel config parameter. There is
1918 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
1919 that allows to increase the default size depending on
1920 the number of CPUs. See init/Kconfig for more details.
1921
1922 logo.nologo [FB] Disables display of the built-in Linux logo.
1923 This may be used to provide more screen space for
1924 kernel log messages and is useful when debugging
1925 kernel boot problems.
1926
1927 lp=0 [LP] Specify parallel ports to use, e.g,
1928 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
1929 lp=reset first parallel port). 'lp=0' disables the
1930 lp=auto printer driver. 'lp=reset' (which can be
1931 specified in addition to the ports) causes
1932 attached printers to be reset. Using
1933 lp=port1,port2,... specifies the parallel ports
1934 to associate lp devices with, starting with
1935 lp0. A port specification may be 'none' to skip
1936 that lp device, or a parport name such as
1937 'parport0'. Specifying 'lp=auto' instead of a
1938 port specification list means that device IDs
1939 from each port should be examined, to see if
1940 an IEEE 1284-compliant printer is attached; if
1941 so, the driver will manage that printer.
1942 See also header of drivers/char/lp.c.
1943
1944 lpj=n [KNL]
1945 Sets loops_per_jiffy to given constant, thus avoiding
1946 time-consuming boot-time autodetection (up to 250 ms per
1947 CPU). 0 enables autodetection (default). To determine
1948 the correct value for your kernel, boot with normal
1949 autodetection and see what value is printed. Note that
1950 on SMP systems the preset will be applied to all CPUs,
1951 which is likely to cause problems if your CPUs need
1952 significantly divergent settings. An incorrect value
1953 will cause delays in the kernel to be wrong, leading to
1954 unpredictable I/O errors and other breakage. Although
1955 unlikely, in the extreme case this might damage your
1956 hardware.
1957
1958 ltpc= [NET]
1959 Format: <io>,<irq>,<dma>
1960
1961 machvec= [IA-64] Force the use of a particular machine-vector
1962 (machvec) in a generic kernel.
1963 Example: machvec=hpzx1_swiotlb
1964
1965 machtype= [Loongson] Share the same kernel image file between different
1966 yeeloong laptop.
1967 Example: machtype=lemote-yeeloong-2f-7inch
1968
1969 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
1970 than or equal to this physical address is ignored.
1971
1972 maxcpus= [SMP] Maximum number of processors that an SMP kernel
1973 should make use of. maxcpus=n : n >= 0 limits the
1974 kernel to using 'n' processors. n=0 is a special case,
1975 it is equivalent to "nosmp", which also disables
1976 the IO APIC.
1977
1978 max_loop= [LOOP] The number of loop block devices that get
1979 (loop.max_loop) unconditionally pre-created at init time. The default
1980 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
1981 of statically allocating a predefined number, loop
1982 devices can be requested on-demand with the
1983 /dev/loop-control interface.
1984
1985 mce [X86-32] Machine Check Exception
1986
1987 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
1988
1989 md= [HW] RAID subsystems devices and level
1990 See Documentation/md.txt.
1991
1992 mdacon= [MDA]
1993 Format: <first>,<last>
1994 Specifies range of consoles to be captured by the MDA.
1995
1996 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
1997 Amount of memory to be used when the kernel is not able
1998 to see the whole system memory or for test.
1999 [X86] Work as limiting max address. Use together
2000 with memmap= to avoid physical address space collisions.
2001 Without memmap= PCI devices could be placed at addresses
2002 belonging to unused RAM.
2003
2004 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2005 memory.
2006
2007 memchunk=nn[KMG]
2008 [KNL,SH] Allow user to override the default size for
2009 per-device physically contiguous DMA buffers.
2010
2011 memmap=exactmap [KNL,X86] Enable setting of an exact
2012 E820 memory map, as specified by the user.
2013 Such memmap=exactmap lines can be constructed based on
2014 BIOS output or other requirements. See the memmap=nn@ss
2015 option description.
2016
2017 memmap=nn[KMG]@ss[KMG]
2018 [KNL] Force usage of a specific region of memory.
2019 Region of memory to be used is from ss to ss+nn.
2020
2021 memmap=nn[KMG]#ss[KMG]
2022 [KNL,ACPI] Mark specific memory as ACPI data.
2023 Region of memory to be marked is from ss to ss+nn.
2024
2025 memmap=nn[KMG]$ss[KMG]
2026 [KNL,ACPI] Mark specific memory as reserved.
2027 Region of memory to be reserved is from ss to ss+nn.
2028 Example: Exclude memory from 0x18690000-0x1869ffff
2029 memmap=64K$0x18690000
2030 or
2031 memmap=0x10000$0x18690000
2032
2033 memmap=nn[KMG]!ss[KMG]
2034 [KNL,X86] Mark specific memory as protected.
2035 Region of memory to be used, from ss to ss+nn.
2036 The memory region may be marked as e820 type 12 (0xc)
2037 and is NVDIMM or ADR memory.
2038
2039 memory_corruption_check=0/1 [X86]
2040 Some BIOSes seem to corrupt the first 64k of
2041 memory when doing things like suspend/resume.
2042 Setting this option will scan the memory
2043 looking for corruption. Enabling this will
2044 both detect corruption and prevent the kernel
2045 from using the memory being corrupted.
2046 However, its intended as a diagnostic tool; if
2047 repeatable BIOS-originated corruption always
2048 affects the same memory, you can use memmap=
2049 to prevent the kernel from using that memory.
2050
2051 memory_corruption_check_size=size [X86]
2052 By default it checks for corruption in the low
2053 64k, making this memory unavailable for normal
2054 use. Use this parameter to scan for
2055 corruption in more or less memory.
2056
2057 memory_corruption_check_period=seconds [X86]
2058 By default it checks for corruption every 60
2059 seconds. Use this parameter to check at some
2060 other rate. 0 disables periodic checking.
2061
2062 memtest= [KNL,X86,ARM] Enable memtest
2063 Format: <integer>
2064 default : 0 <disable>
2065 Specifies the number of memtest passes to be
2066 performed. Each pass selects another test
2067 pattern from a given set of patterns. Memtest
2068 fills the memory with this pattern, validates
2069 memory contents and reserves bad memory
2070 regions that are detected.
2071
2072 meye.*= [HW] Set MotionEye Camera parameters
2073 See Documentation/video4linux/meye.txt.
2074
2075 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2076 Multi-Function General Purpose Timers on AMD Geode
2077 platforms.
2078
2079 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2080 the BIOS has incorrectly applied a workaround. TinyBIOS
2081 version 0.98 is known to be affected, 0.99 fixes the
2082 problem by letting the user disable the workaround.
2083
2084 mga= [HW,DRM]
2085
2086 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2087 physical address is ignored.
2088
2089 mini2440= [ARM,HW,KNL]
2090 Format:[0..2][b][c][t]
2091 Default: "0tb"
2092 MINI2440 configuration specification:
2093 0 - The attached screen is the 3.5" TFT
2094 1 - The attached screen is the 7" TFT
2095 2 - The VGA Shield is attached (1024x768)
2096 Leaving out the screen size parameter will not load
2097 the TFT driver, and the framebuffer will be left
2098 unconfigured.
2099 b - Enable backlight. The TFT backlight pin will be
2100 linked to the kernel VESA blanking code and a GPIO
2101 LED. This parameter is not necessary when using the
2102 VGA shield.
2103 c - Enable the s3c camera interface.
2104 t - Reserved for enabling touchscreen support. The
2105 touchscreen support is not enabled in the mainstream
2106 kernel as of 2.6.30, a preliminary port can be found
2107 in the "bleeding edge" mini2440 support kernel at
2108 http://repo.or.cz/w/linux-2.6/mini2440.git
2109
2110 mminit_loglevel=
2111 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2112 parameter allows control of the logging verbosity for
2113 the additional memory initialisation checks. A value
2114 of 0 disables mminit logging and a level of 4 will
2115 log everything. Information is printed at KERN_DEBUG
2116 so loglevel=8 may also need to be specified.
2117
2118 module.sig_enforce
2119 [KNL] When CONFIG_MODULE_SIG is set, this means that
2120 modules without (valid) signatures will fail to load.
2121 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2122 is always true, so this option does nothing.
2123
2124 mousedev.tap_time=
2125 [MOUSE] Maximum time between finger touching and
2126 leaving touchpad surface for touch to be considered
2127 a tap and be reported as a left button click (for
2128 touchpads working in absolute mode only).
2129 Format: <msecs>
2130 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2131 reporting absolute coordinates, such as tablets
2132 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2133 reporting absolute coordinates, such as tablets
2134
2135 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
2136 is similar to kernelcore except it specifies the
2137 amount of memory used for migratable allocations.
2138 If both kernelcore and movablecore is specified,
2139 then kernelcore will be at *least* the specified
2140 value but may be more. If movablecore on its own
2141 is specified, the administrator must be careful
2142 that the amount of memory usable for all allocations
2143 is not too small.
2144
2145 movable_node [KNL,X86] Boot-time switch to enable the effects
2146 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2147
2148 MTD_Partition= [MTD]
2149 Format: <name>,<region-number>,<size>,<offset>
2150
2151 MTD_Region= [MTD] Format:
2152 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2153
2154 mtdparts= [MTD]
2155 See drivers/mtd/cmdlinepart.c.
2156
2157 multitce=off [PPC] This parameter disables the use of the pSeries
2158 firmware feature for updating multiple TCE entries
2159 at a time.
2160
2161 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2162
2163 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2164
2165 boundary - index of last SLC block on Flex-OneNAND.
2166 The remaining blocks are configured as MLC blocks.
2167 lock - Configure if Flex-OneNAND boundary should be locked.
2168 Once locked, the boundary cannot be changed.
2169 1 indicates lock status, 0 indicates unlock status.
2170
2171 mtdset= [ARM]
2172 ARM/S3C2412 JIVE boot control
2173
2174 See arch/arm/mach-s3c2412/mach-jive.c
2175
2176 mtouchusb.raw_coordinates=
2177 [HW] Make the MicroTouch USB driver use raw coordinates
2178 ('y', default) or cooked coordinates ('n')
2179
2180 mtrr_chunk_size=nn[KMG] [X86]
2181 used for mtrr cleanup. It is largest continuous chunk
2182 that could hold holes aka. UC entries.
2183
2184 mtrr_gran_size=nn[KMG] [X86]
2185 Used for mtrr cleanup. It is granularity of mtrr block.
2186 Default is 1.
2187 Large value could prevent small alignment from
2188 using up MTRRs.
2189
2190 mtrr_spare_reg_nr=n [X86]
2191 Format: <integer>
2192 Range: 0,7 : spare reg number
2193 Default : 1
2194 Used for mtrr cleanup. It is spare mtrr entries number.
2195 Set to 2 or more if your graphical card needs more.
2196
2197 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2198
2199 netdev= [NET] Network devices parameters
2200 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2201 Note that mem_start is often overloaded to mean
2202 something different and driver-specific.
2203 This usage is only documented in each driver source
2204 file if at all.
2205
2206 nf_conntrack.acct=
2207 [NETFILTER] Enable connection tracking flow accounting
2208 0 to disable accounting
2209 1 to enable accounting
2210 Default value is 0.
2211
2212 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2213 See Documentation/filesystems/nfs/nfsroot.txt.
2214
2215 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2216 See Documentation/filesystems/nfs/nfsroot.txt.
2217
2218 nfsrootdebug [NFS] enable nfsroot debugging messages.
2219 See Documentation/filesystems/nfs/nfsroot.txt.
2220
2221 nfs.callback_tcpport=
2222 [NFS] set the TCP port on which the NFSv4 callback
2223 channel should listen.
2224
2225 nfs.cache_getent=
2226 [NFS] sets the pathname to the program which is used
2227 to update the NFS client cache entries.
2228
2229 nfs.cache_getent_timeout=
2230 [NFS] sets the timeout after which an attempt to
2231 update a cache entry is deemed to have failed.
2232
2233 nfs.idmap_cache_timeout=
2234 [NFS] set the maximum lifetime for idmapper cache
2235 entries.
2236
2237 nfs.enable_ino64=
2238 [NFS] enable 64-bit inode numbers.
2239 If zero, the NFS client will fake up a 32-bit inode
2240 number for the readdir() and stat() syscalls instead
2241 of returning the full 64-bit number.
2242 The default is to return 64-bit inode numbers.
2243
2244 nfs.max_session_slots=
2245 [NFSv4.1] Sets the maximum number of session slots
2246 the client will attempt to negotiate with the server.
2247 This limits the number of simultaneous RPC requests
2248 that the client can send to the NFSv4.1 server.
2249 Note that there is little point in setting this
2250 value higher than the max_tcp_slot_table_limit.
2251
2252 nfs.nfs4_disable_idmapping=
2253 [NFSv4] When set to the default of '1', this option
2254 ensures that both the RPC level authentication
2255 scheme and the NFS level operations agree to use
2256 numeric uids/gids if the mount is using the
2257 'sec=sys' security flavour. In effect it is
2258 disabling idmapping, which can make migration from
2259 legacy NFSv2/v3 systems to NFSv4 easier.
2260 Servers that do not support this mode of operation
2261 will be autodetected by the client, and it will fall
2262 back to using the idmapper.
2263 To turn off this behaviour, set the value to '0'.
2264 nfs.nfs4_unique_id=
2265 [NFS4] Specify an additional fixed unique ident-
2266 ification string that NFSv4 clients can insert into
2267 their nfs_client_id4 string. This is typically a
2268 UUID that is generated at system install time.
2269
2270 nfs.send_implementation_id =
2271 [NFSv4.1] Send client implementation identification
2272 information in exchange_id requests.
2273 If zero, no implementation identification information
2274 will be sent.
2275 The default is to send the implementation identification
2276 information.
2277
2278 nfs.recover_lost_locks =
2279 [NFSv4] Attempt to recover locks that were lost due
2280 to a lease timeout on the server. Please note that
2281 doing this risks data corruption, since there are
2282 no guarantees that the file will remain unchanged
2283 after the locks are lost.
2284 If you want to enable the kernel legacy behaviour of
2285 attempting to recover these locks, then set this
2286 parameter to '1'.
2287 The default parameter value of '0' causes the kernel
2288 not to attempt recovery of lost locks.
2289
2290 nfs4.layoutstats_timer =
2291 [NFSv4.2] Change the rate at which the kernel sends
2292 layoutstats to the pNFS metadata server.
2293
2294 Setting this to value to 0 causes the kernel to use
2295 whatever value is the default set by the layout
2296 driver. A non-zero value sets the minimum interval
2297 in seconds between layoutstats transmissions.
2298
2299 nfsd.nfs4_disable_idmapping=
2300 [NFSv4] When set to the default of '1', the NFSv4
2301 server will return only numeric uids and gids to
2302 clients using auth_sys, and will accept numeric uids
2303 and gids from such clients. This is intended to ease
2304 migration from NFSv2/v3.
2305
2306 objlayoutdriver.osd_login_prog=
2307 [NFS] [OBJLAYOUT] sets the pathname to the program which
2308 is used to automatically discover and login into new
2309 osd-targets. Please see:
2310 Documentation/filesystems/pnfs.txt for more explanations
2311
2312 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
2313 when a NMI is triggered.
2314 Format: [state][,regs][,debounce][,die]
2315
2316 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2317 Format: [panic,][nopanic,][num]
2318 Valid num: 0 or 1
2319 0 - turn nmi_watchdog off
2320 1 - turn nmi_watchdog on
2321 When panic is specified, panic when an NMI watchdog
2322 timeout occurs (or 'nopanic' to override the opposite
2323 default).
2324 This is useful when you use a panic=... timeout and
2325 need the box quickly up again.
2326
2327 netpoll.carrier_timeout=
2328 [NET] Specifies amount of time (in seconds) that
2329 netpoll should wait for a carrier. By default netpoll
2330 waits 4 seconds.
2331
2332 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2333 emulation library even if a 387 maths coprocessor
2334 is present.
2335
2336 no_console_suspend
2337 [HW] Never suspend the console
2338 Disable suspending of consoles during suspend and
2339 hibernate operations. Once disabled, debugging
2340 messages can reach various consoles while the rest
2341 of the system is being put to sleep (ie, while
2342 debugging driver suspend/resume hooks). This may
2343 not work reliably with all consoles, but is known
2344 to work with serial and VGA consoles.
2345 To facilitate more flexible debugging, we also add
2346 console_suspend, a printk module parameter to control
2347 it. Users could use console_suspend (usually
2348 /sys/module/printk/parameters/console_suspend) to
2349 turn on/off it dynamically.
2350
2351 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2352 caches in the slab allocator. Saves per-node memory,
2353 but will impact performance.
2354
2355 noalign [KNL,ARM]
2356
2357 noapic [SMP,APIC] Tells the kernel to not make use of any
2358 IOAPICs that may be present in the system.
2359
2360 noautogroup Disable scheduler automatic task group creation.
2361
2362 nobats [PPC] Do not use BATs for mapping kernel lowmem
2363 on "Classic" PPC cores.
2364
2365 nocache [ARM]
2366
2367 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2368
2369 nodelayacct [KNL] Disable per-task delay accounting
2370
2371 nodisconnect [HW,SCSI,M68K] Disables SCSI disconnects.
2372
2373 nodsp [SH] Disable hardware DSP at boot time.
2374
2375 noefi Disable EFI runtime services support.
2376
2377 noexec [IA-64]
2378
2379 noexec [X86]
2380 On X86-32 available only on PAE configured kernels.
2381 noexec=on: enable non-executable mappings (default)
2382 noexec=off: disable non-executable mappings
2383
2384 nosmap [X86]
2385 Disable SMAP (Supervisor Mode Access Prevention)
2386 even if it is supported by processor.
2387
2388 nosmep [X86]
2389 Disable SMEP (Supervisor Mode Execution Prevention)
2390 even if it is supported by processor.
2391
2392 noexec32 [X86-64]
2393 This affects only 32-bit executables.
2394 noexec32=on: enable non-executable mappings (default)
2395 read doesn't imply executable mappings
2396 noexec32=off: disable non-executable mappings
2397 read implies executable mappings
2398
2399 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2400
2401 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2402 register save and restore. The kernel will only save
2403 legacy floating-point registers on task switch.
2404
2405 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2406
2407 noxsave [BUGS=X86] Disables x86 extended register state save
2408 and restore using xsave. The kernel will fallback to
2409 enabling legacy floating-point and sse state.
2410
2411 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2412 register states. The kernel will fall back to use
2413 xsave to save the states. By using this parameter,
2414 performance of saving the states is degraded because
2415 xsave doesn't support modified optimization while
2416 xsaveopt supports it on xsaveopt enabled systems.
2417
2418 noxsaves [X86] Disables xsaves and xrstors used in saving and
2419 restoring x86 extended register state in compacted
2420 form of xsave area. The kernel will fall back to use
2421 xsaveopt and xrstor to save and restore the states
2422 in standard form of xsave area. By using this
2423 parameter, xsave area per process might occupy more
2424 memory on xsaves enabled systems.
2425
2426 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2427 wfi(ARM) instruction doesn't work correctly and not to
2428 use it. This is also useful when using JTAG debugger.
2429
2430 no_file_caps Tells the kernel not to honor file capabilities. The
2431 only way then for a file to be executed with privilege
2432 is to be setuid root or executed by root.
2433
2434 nohalt [IA-64] Tells the kernel not to use the power saving
2435 function PAL_HALT_LIGHT when idle. This increases
2436 power-consumption. On the positive side, it reduces
2437 interrupt wake-up latency, which may improve performance
2438 in certain environments such as networked servers or
2439 real-time systems.
2440
2441 nohibernate [HIBERNATION] Disable hibernation and resume.
2442
2443 nohz= [KNL] Boottime enable/disable dynamic ticks
2444 Valid arguments: on, off
2445 Default: on
2446
2447 nohz_full= [KNL,BOOT]
2448 In kernels built with CONFIG_NO_HZ_FULL=y, set
2449 the specified list of CPUs whose tick will be stopped
2450 whenever possible. The boot CPU will be forced outside
2451 the range to maintain the timekeeping.
2452 The CPUs in this range must also be included in the
2453 rcu_nocbs= set.
2454
2455 noiotrap [SH] Disables trapped I/O port accesses.
2456
2457 noirqdebug [X86-32] Disables the code which attempts to detect and
2458 disable unhandled interrupt sources.
2459
2460 no_timer_check [X86,APIC] Disables the code which tests for
2461 broken timer IRQ sources.
2462
2463 noisapnp [ISAPNP] Disables ISA PnP code.
2464
2465 noinitrd [RAM] Tells the kernel not to load any configured
2466 initial RAM disk.
2467
2468 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2469 remapping.
2470 [Deprecated - use intremap=off]
2471
2472 nointroute [IA-64]
2473
2474 nojitter [IA-64] Disables jitter checking for ITC timers.
2475
2476 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2477
2478 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2479 fault handling.
2480
2481 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2482 steal time is computed, but won't influence scheduler
2483 behaviour
2484
2485 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2486
2487 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2488
2489 noltlbs [PPC] Do not use large page/tlb entries for kernel
2490 lowmem mapping on PPC40x.
2491
2492 nomca [IA-64] Disable machine check abort handling
2493
2494 nomce [X86-32] Disable Machine Check Exception
2495
2496 nomfgpt [X86-32] Disable Multi-Function General Purpose
2497 Timer usage (for AMD Geode machines).
2498
2499 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2500 shutdown the other cpus. Instead use the REBOOT_VECTOR
2501 irq.
2502
2503 nomodule Disable module load
2504
2505 nopat [X86] Disable PAT (page attribute table extension of
2506 pagetables) support.
2507
2508 norandmaps Don't use address space randomization. Equivalent to
2509 echo 0 > /proc/sys/kernel/randomize_va_space
2510
2511 noreplace-paravirt [X86,IA-64,PV_OPS] Don't patch paravirt_ops
2512
2513 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2514 with UP alternatives
2515
2516 nordrand [X86] Disable kernel use of the RDRAND and
2517 RDSEED instructions even if they are supported
2518 by the processor. RDRAND and RDSEED are still
2519 available to user space applications.
2520
2521 noresume [SWSUSP] Disables resume and restores original swap
2522 space.
2523
2524 no-scroll [VGA] Disables scrollback.
2525 This is required for the Braillex ib80-piezo Braille
2526 reader made by F.H. Papenmeier (Germany).
2527
2528 nosbagart [IA-64]
2529
2530 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2531
2532 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2533 and disable the IO APIC. legacy for "maxcpus=0".
2534
2535 nosoftlockup [KNL] Disable the soft-lockup detector.
2536
2537 nosync [HW,M68K] Disables sync negotiation for all devices.
2538
2539 notsc [BUGS=X86-32] Disable Time Stamp Counter
2540
2541 nousb [USB] Disable the USB subsystem
2542
2543 nowatchdog [KNL] Disable both lockup detectors, i.e.
2544 soft-lockup and NMI watchdog (hard-lockup).
2545
2546 nowb [ARM]
2547
2548 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2549
2550 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2551 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2552 Some features depend on CPU0. Known dependencies are:
2553 1. Resume from suspend/hibernate depends on CPU0.
2554 Suspend/hibernate will fail if CPU0 is offline and you
2555 need to online CPU0 before suspend/hibernate.
2556 2. PIC interrupts also depend on CPU0. CPU0 can't be
2557 removed if a PIC interrupt is detected.
2558 It's said poweroff/reboot may depend on CPU0 on some
2559 machines although I haven't seen such issues so far
2560 after CPU0 is offline on a few tested machines.
2561 If the dependencies are under your control, you can
2562 turn on cpu0_hotplug.
2563
2564 nptcg= [IA-64] Override max number of concurrent global TLB
2565 purges which is reported from either PAL_VM_SUMMARY or
2566 SAL PALO.
2567
2568 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2569 could support. nr_cpus=n : n >= 1 limits the kernel to
2570 supporting 'n' processors. Later in runtime you can not
2571 use hotplug cpu feature to put more cpu back to online.
2572 just like you compile the kernel NR_CPUS=n
2573
2574 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2575
2576 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2577 Allowed values are enable and disable
2578
2579 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2580 one of ['zone', 'node', 'default'] can be specified
2581 This can be set from sysctl after boot.
2582 See Documentation/sysctl/vm.txt for details.
2583
2584 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2585 See Documentation/debugging-via-ohci1394.txt for more
2586 info.
2587
2588 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2589 Rather than timing out after 20 ms if an EC
2590 command is not properly ACKed, override the length
2591 of the timeout. We have interrupts disabled while
2592 waiting for the ACK, so if this is set too high
2593 interrupts *may* be lost!
2594
2595 omap_mux= [OMAP] Override bootloader pin multiplexing.
2596 Format: <mux_mode0.mode_name=value>...
2597 For example, to override I2C bus2:
2598 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2599
2600 oprofile.timer= [HW]
2601 Use timer interrupt instead of performance counters
2602
2603 oprofile.cpu_type= Force an oprofile cpu type
2604 This might be useful if you have an older oprofile
2605 userland or if you want common events.
2606 Format: { arch_perfmon }
2607 arch_perfmon: [X86] Force use of architectural
2608 perfmon on Intel CPUs instead of the
2609 CPU specific event set.
2610 timer: [X86] Force use of architectural NMI
2611 timer mode (see also oprofile.timer
2612 for generic hr timer mode)
2613 [s390] Force legacy basic mode sampling
2614 (report cpu_type "timer")
2615
2616 oops=panic Always panic on oopses. Default is to just kill the
2617 process, but there is a small probability of
2618 deadlocking the machine.
2619 This will also cause panics on machine check exceptions.
2620 Useful together with panic=30 to trigger a reboot.
2621
2622 OSS [HW,OSS]
2623 See Documentation/sound/oss/oss-parameters.txt
2624
2625 page_owner= [KNL] Boot-time page_owner enabling option.
2626 Storage of the information about who allocated
2627 each page is disabled in default. With this switch,
2628 we can turn it on.
2629 on: enable the feature
2630
2631 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2632 timeout > 0: seconds before rebooting
2633 timeout = 0: wait forever
2634 timeout < 0: reboot immediately
2635 Format: <timeout>
2636
2637 panic_on_warn panic() instead of WARN(). Useful to cause kdump
2638 on a WARN().
2639
2640 crash_kexec_post_notifiers
2641 Run kdump after running panic-notifiers and dumping
2642 kmsg. This only for the users who doubt kdump always
2643 succeeds in any situation.
2644 Note that this also increases risks of kdump failure,
2645 because some panic notifiers can make the crashed
2646 kernel more unstable.
2647
2648 parkbd.port= [HW] Parallel port number the keyboard adapter is
2649 connected to, default is 0.
2650 Format: <parport#>
2651 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2652 0 for XT, 1 for AT (default is AT).
2653 Format: <mode>
2654
2655 parport= [HW,PPT] Specify parallel ports. 0 disables.
2656 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2657 Use 'auto' to force the driver to use any
2658 IRQ/DMA settings detected (the default is to
2659 ignore detected IRQ/DMA settings because of
2660 possible conflicts). You can specify the base
2661 address, IRQ, and DMA settings; IRQ and DMA
2662 should be numbers, or 'auto' (for using detected
2663 settings on that particular port), or 'nofifo'
2664 (to avoid using a FIFO even if it is detected).
2665 Parallel ports are assigned in the order they
2666 are specified on the command line, starting
2667 with parport0.
2668
2669 parport_init_mode= [HW,PPT]
2670 Configure VIA parallel port to operate in
2671 a specific mode. This is necessary on Pegasos
2672 computer where firmware has no options for setting
2673 up parallel port mode and sets it to spp.
2674 Currently this function knows 686a and 8231 chips.
2675 Format: [spp|ps2|epp|ecp|ecpepp]
2676
2677 pause_on_oops=
2678 Halt all CPUs after the first oops has been printed for
2679 the specified number of seconds. This is to be used if
2680 your oopses keep scrolling off the screen.
2681
2682 pcbit= [HW,ISDN]
2683
2684 pcd. [PARIDE]
2685 See header of drivers/block/paride/pcd.c.
2686 See also Documentation/blockdev/paride.txt.
2687
2688 pci=option[,option...] [PCI] various PCI subsystem options:
2689 earlydump [X86] dump PCI config space before the kernel
2690 changes anything
2691 off [X86] don't probe for the PCI bus
2692 bios [X86-32] force use of PCI BIOS, don't access
2693 the hardware directly. Use this if your machine
2694 has a non-standard PCI host bridge.
2695 nobios [X86-32] disallow use of PCI BIOS, only direct
2696 hardware access methods are allowed. Use this
2697 if you experience crashes upon bootup and you
2698 suspect they are caused by the BIOS.
2699 conf1 [X86] Force use of PCI Configuration
2700 Mechanism 1.
2701 conf2 [X86] Force use of PCI Configuration
2702 Mechanism 2.
2703 noaer [PCIE] If the PCIEAER kernel config parameter is
2704 enabled, this kernel boot option can be used to
2705 disable the use of PCIE advanced error reporting.
2706 nodomains [PCI] Disable support for multiple PCI
2707 root domains (aka PCI segments, in ACPI-speak).
2708 nommconf [X86] Disable use of MMCONFIG for PCI
2709 Configuration
2710 check_enable_amd_mmconf [X86] check for and enable
2711 properly configured MMIO access to PCI
2712 config space on AMD family 10h CPU
2713 nomsi [MSI] If the PCI_MSI kernel config parameter is
2714 enabled, this kernel boot option can be used to
2715 disable the use of MSI interrupts system-wide.
2716 noioapicquirk [APIC] Disable all boot interrupt quirks.
2717 Safety option to keep boot IRQs enabled. This
2718 should never be necessary.
2719 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
2720 primary IO-APIC for bridges that cannot disable
2721 boot IRQs. This fixes a source of spurious IRQs
2722 when the system masks IRQs.
2723 noioapicreroute [APIC] Disable workaround that uses the
2724 boot IRQ equivalent of an IRQ that connects to
2725 a chipset where boot IRQs cannot be disabled.
2726 The opposite of ioapicreroute.
2727 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
2728 routing table. These calls are known to be buggy
2729 on several machines and they hang the machine
2730 when used, but on other computers it's the only
2731 way to get the interrupt routing table. Try
2732 this option if the kernel is unable to allocate
2733 IRQs or discover secondary PCI buses on your
2734 motherboard.
2735 rom [X86] Assign address space to expansion ROMs.
2736 Use with caution as certain devices share
2737 address decoders between ROMs and other
2738 resources.
2739 norom [X86] Do not assign address space to
2740 expansion ROMs that do not already have
2741 BIOS assigned address ranges.
2742 nobar [X86] Do not assign address space to the
2743 BARs that weren't assigned by the BIOS.
2744 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
2745 assigned automatically to PCI devices. You can
2746 make the kernel exclude IRQs of your ISA cards
2747 this way.
2748 pirqaddr=0xAAAAA [X86] Specify the physical address
2749 of the PIRQ table (normally generated
2750 by the BIOS) if it is outside the
2751 F0000h-100000h range.
2752 lastbus=N [X86] Scan all buses thru bus #N. Can be
2753 useful if the kernel is unable to find your
2754 secondary buses and you want to tell it
2755 explicitly which ones they are.
2756 assign-busses [X86] Always assign all PCI bus
2757 numbers ourselves, overriding
2758 whatever the firmware may have done.
2759 usepirqmask [X86] Honor the possible IRQ mask stored
2760 in the BIOS $PIR table. This is needed on
2761 some systems with broken BIOSes, notably
2762 some HP Pavilion N5400 and Omnibook XE3
2763 notebooks. This will have no effect if ACPI
2764 IRQ routing is enabled.
2765 noacpi [X86] Do not use ACPI for IRQ routing
2766 or for PCI scanning.
2767 use_crs [X86] Use PCI host bridge window information
2768 from ACPI. On BIOSes from 2008 or later, this
2769 is enabled by default. If you need to use this,
2770 please report a bug.
2771 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2772 If you need to use this, please report a bug.
2773 routeirq Do IRQ routing for all PCI devices.
2774 This is normally done in pci_enable_device(),
2775 so this option is a temporary workaround
2776 for broken drivers that don't call it.
2777 skip_isa_align [X86] do not align io start addr, so can
2778 handle more pci cards
2779 firmware [ARM] Do not re-enumerate the bus but instead
2780 just use the configuration from the
2781 bootloader. This is currently used on
2782 IXP2000 systems where the bus has to be
2783 configured a certain way for adjunct CPUs.
2784 noearly [X86] Don't do any early type 1 scanning.
2785 This might help on some broken boards which
2786 machine check when some devices' config space
2787 is read. But various workarounds are disabled
2788 and some IOMMU drivers will not work.
2789 bfsort Sort PCI devices into breadth-first order.
2790 This sorting is done to get a device
2791 order compatible with older (<= 2.4) kernels.
2792 nobfsort Don't sort PCI devices into breadth-first order.
2793 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
2794 tuning and use the BIOS-configured MPS defaults.
2795 pcie_bus_safe Set every device's MPS to the largest value
2796 supported by all devices below the root complex.
2797 pcie_bus_perf Set device MPS to the largest allowable MPS
2798 based on its parent bus. Also set MRRS (Max
2799 Read Request Size) to the largest supported
2800 value (no larger than the MPS that the device
2801 or bus can support) for best performance.
2802 pcie_bus_peer2peer Set every device's MPS to 128B, which
2803 every device is guaranteed to support. This
2804 configuration allows peer-to-peer DMA between
2805 any pair of devices, possibly at the cost of
2806 reduced performance. This also guarantees
2807 that hot-added devices will work.
2808 cbiosize=nn[KMG] The fixed amount of bus space which is
2809 reserved for the CardBus bridge's IO window.
2810 The default value is 256 bytes.
2811 cbmemsize=nn[KMG] The fixed amount of bus space which is
2812 reserved for the CardBus bridge's memory
2813 window. The default value is 64 megabytes.
2814 resource_alignment=
2815 Format:
2816 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
2817 Specifies alignment and device to reassign
2818 aligned memory resources.
2819 If <order of align> is not specified,
2820 PAGE_SIZE is used as alignment.
2821 PCI-PCI bridge can be specified, if resource
2822 windows need to be expanded.
2823 ecrc= Enable/disable PCIe ECRC (transaction layer
2824 end-to-end CRC checking).
2825 bios: Use BIOS/firmware settings. This is the
2826 the default.
2827 off: Turn ECRC off
2828 on: Turn ECRC on.
2829 hpiosize=nn[KMG] The fixed amount of bus space which is
2830 reserved for hotplug bridge's IO window.
2831 Default size is 256 bytes.
2832 hpmemsize=nn[KMG] The fixed amount of bus space which is
2833 reserved for hotplug bridge's memory window.
2834 Default size is 2 megabytes.
2835 realloc= Enable/disable reallocating PCI bridge resources
2836 if allocations done by BIOS are too small to
2837 accommodate resources required by all child
2838 devices.
2839 off: Turn realloc off
2840 on: Turn realloc on
2841 realloc same as realloc=on
2842 noari do not use PCIe ARI.
2843 pcie_scan_all Scan all possible PCIe devices. Otherwise we
2844 only look for one device below a PCIe downstream
2845 port.
2846
2847 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
2848 Management.
2849 off Disable ASPM.
2850 force Enable ASPM even on devices that claim not to support it.
2851 WARNING: Forcing ASPM on may cause system lockups.
2852
2853 pcie_hp= [PCIE] PCI Express Hotplug driver options:
2854 nomsi Do not use MSI for PCI Express Native Hotplug (this
2855 makes all PCIe ports use INTx for hotplug services).
2856
2857 pcie_ports= [PCIE] PCIe ports handling:
2858 auto Ask the BIOS whether or not to use native PCIe services
2859 associated with PCIe ports (PME, hot-plug, AER). Use
2860 them only if that is allowed by the BIOS.
2861 native Use native PCIe services associated with PCIe ports
2862 unconditionally.
2863 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
2864 ports driver.
2865
2866 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
2867 nomsi Do not use MSI for native PCIe PME signaling (this makes
2868 all PCIe root ports use INTx for all services).
2869
2870 pcmv= [HW,PCMCIA] BadgePAD 4
2871
2872 pd_ignore_unused
2873 [PM]
2874 Keep all power-domains already enabled by bootloader on,
2875 even if no driver has claimed them. This is useful
2876 for debug and development, but should not be
2877 needed on a platform with proper driver support.
2878
2879 pd. [PARIDE]
2880 See Documentation/blockdev/paride.txt.
2881
2882 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
2883 boot time.
2884 Format: { 0 | 1 }
2885 See arch/parisc/kernel/pdc_chassis.c
2886
2887 percpu_alloc= Select which percpu first chunk allocator to use.
2888 Currently supported values are "embed" and "page".
2889 Archs may support subset or none of the selections.
2890 See comments in mm/percpu.c for details on each
2891 allocator. This parameter is primarily for debugging
2892 and performance comparison.
2893
2894 pf. [PARIDE]
2895 See Documentation/blockdev/paride.txt.
2896
2897 pg. [PARIDE]
2898 See Documentation/blockdev/paride.txt.
2899
2900 pirq= [SMP,APIC] Manual mp-table setup
2901 See Documentation/x86/i386/IO-APIC.txt.
2902
2903 plip= [PPT,NET] Parallel port network link
2904 Format: { parport<nr> | timid | 0 }
2905 See also Documentation/parport.txt.
2906
2907 pmtmr= [X86] Manual setup of pmtmr I/O Port.
2908 Override pmtimer IOPort with a hex value.
2909 e.g. pmtmr=0x508
2910
2911 pnp.debug=1 [PNP]
2912 Enable PNP debug messages (depends on the
2913 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
2914 via /sys/module/pnp/parameters/debug. We always show
2915 current resource usage; turning this on also shows
2916 possible settings and some assignment information.
2917
2918 pnpacpi= [ACPI]
2919 { off }
2920
2921 pnpbios= [ISAPNP]
2922 { on | off | curr | res | no-curr | no-res }
2923
2924 pnp_reserve_irq=
2925 [ISAPNP] Exclude IRQs for the autoconfiguration
2926
2927 pnp_reserve_dma=
2928 [ISAPNP] Exclude DMAs for the autoconfiguration
2929
2930 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
2931 Ranges are in pairs (I/O port base and size).
2932
2933 pnp_reserve_mem=
2934 [ISAPNP] Exclude memory regions for the
2935 autoconfiguration.
2936 Ranges are in pairs (memory base and size).
2937
2938 ports= [IP_VS_FTP] IPVS ftp helper module
2939 Default is 21.
2940 Up to 8 (IP_VS_APP_MAX_PORTS) ports
2941 may be specified.
2942 Format: <port>,<port>....
2943
2944 print-fatal-signals=
2945 [KNL] debug: print fatal signals
2946
2947 If enabled, warn about various signal handling
2948 related application anomalies: too many signals,
2949 too many POSIX.1 timers, fatal signals causing a
2950 coredump - etc.
2951
2952 If you hit the warning due to signal overflow,
2953 you might want to try "ulimit -i unlimited".
2954
2955 default: off.
2956
2957 printk.always_kmsg_dump=
2958 Trigger kmsg_dump for cases other than kernel oops or
2959 panics
2960 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2961 default: disabled
2962
2963 printk.time= Show timing data prefixed to each printk message line
2964 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
2965
2966 processor.max_cstate= [HW,ACPI]
2967 Limit processor to maximum C-state
2968 max_cstate=9 overrides any DMI blacklist limit.
2969
2970 processor.nocst [HW,ACPI]
2971 Ignore the _CST method to determine C-states,
2972 instead using the legacy FADT method
2973
2974 profile= [KNL] Enable kernel profiling via /proc/profile
2975 Format: [schedule,]<number>
2976 Param: "schedule" - profile schedule points.
2977 Param: <number> - step/bucket size as a power of 2 for
2978 statistical time based profiling.
2979 Param: "sleep" - profile D-state sleeping (millisecs).
2980 Requires CONFIG_SCHEDSTATS
2981 Param: "kvm" - profile VM exits.
2982
2983 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
2984 before loading.
2985 See Documentation/blockdev/ramdisk.txt.
2986
2987 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
2988 probe for; one of (bare|imps|exps|lifebook|any).
2989 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
2990 per second.
2991 psmouse.resetafter= [HW,MOUSE]
2992 Try to reset the device after so many bad packets
2993 (0 = never).
2994 psmouse.resolution=
2995 [HW,MOUSE] Set desired mouse resolution, in dpi.
2996 psmouse.smartscroll=
2997 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
2998 0 = disabled, 1 = enabled (default).
2999
3000 pstore.backend= Specify the name of the pstore backend to use
3001
3002 pt. [PARIDE]
3003 See Documentation/blockdev/paride.txt.
3004
3005 pty.legacy_count=
3006 [KNL] Number of legacy pty's. Overwrites compiled-in
3007 default number.
3008
3009 quiet [KNL] Disable most log messages
3010
3011 r128= [HW,DRM]
3012
3013 raid= [HW,RAID]
3014 See Documentation/md.txt.
3015
3016 ramdisk_blocksize= [RAM]
3017 See Documentation/blockdev/ramdisk.txt.
3018
3019 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3020 See Documentation/blockdev/ramdisk.txt.
3021
3022 rcu_nocbs= [KNL]
3023 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3024 the specified list of CPUs to be no-callback CPUs.
3025 Invocation of these CPUs' RCU callbacks will
3026 be offloaded to "rcuox/N" kthreads created for
3027 that purpose, where "x" is "b" for RCU-bh, "p"
3028 for RCU-preempt, and "s" for RCU-sched, and "N"
3029 is the CPU number. This reduces OS jitter on the
3030 offloaded CPUs, which can be useful for HPC and
3031 real-time workloads. It can also improve energy
3032 efficiency for asymmetric multiprocessors.
3033
3034 rcu_nocb_poll [KNL]
3035 Rather than requiring that offloaded CPUs
3036 (specified by rcu_nocbs= above) explicitly
3037 awaken the corresponding "rcuoN" kthreads,
3038 make these kthreads poll for callbacks.
3039 This improves the real-time response for the
3040 offloaded CPUs by relieving them of the need to
3041 wake up the corresponding kthread, but degrades
3042 energy efficiency by requiring that the kthreads
3043 periodically wake up to do the polling.
3044
3045 rcutree.blimit= [KNL]
3046 Set maximum number of finished RCU callbacks to
3047 process in one batch.
3048
3049 rcutree.dump_tree= [KNL]
3050 Dump the structure of the rcu_node combining tree
3051 out at early boot. This is used for diagnostic
3052 purposes, to verify correct tree setup.
3053
3054 rcutree.gp_cleanup_delay= [KNL]
3055 Set the number of jiffies to delay each step of
3056 RCU grace-period cleanup. This only has effect
3057 when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3058
3059 rcutree.gp_init_delay= [KNL]
3060 Set the number of jiffies to delay each step of
3061 RCU grace-period initialization. This only has
3062 effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3063 is set.
3064
3065 rcutree.gp_preinit_delay= [KNL]
3066 Set the number of jiffies to delay each step of
3067 RCU grace-period pre-initialization, that is,
3068 the propagation of recent CPU-hotplug changes up
3069 the rcu_node combining tree. This only has effect
3070 when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
3071
3072 rcutree.rcu_fanout_exact= [KNL]
3073 Disable autobalancing of the rcu_node combining
3074 tree. This is used by rcutorture, and might
3075 possibly be useful for architectures having high
3076 cache-to-cache transfer latencies.
3077
3078 rcutree.rcu_fanout_leaf= [KNL]
3079 Increase the number of CPUs assigned to each
3080 leaf rcu_node structure. Useful for very large
3081 systems.
3082
3083 rcutree.jiffies_till_sched_qs= [KNL]
3084 Set required age in jiffies for a
3085 given grace period before RCU starts
3086 soliciting quiescent-state help from
3087 rcu_note_context_switch().
3088
3089 rcutree.jiffies_till_first_fqs= [KNL]
3090 Set delay from grace-period initialization to
3091 first attempt to force quiescent states.
3092 Units are jiffies, minimum value is zero,
3093 and maximum value is HZ.
3094
3095 rcutree.jiffies_till_next_fqs= [KNL]
3096 Set delay between subsequent attempts to force
3097 quiescent states. Units are jiffies, minimum
3098 value is one, and maximum value is HZ.
3099
3100 rcutree.kthread_prio= [KNL,BOOT]
3101 Set the SCHED_FIFO priority of the RCU per-CPU
3102 kthreads (rcuc/N). This value is also used for
3103 the priority of the RCU boost threads (rcub/N)
3104 and for the RCU grace-period kthreads (rcu_bh,
3105 rcu_preempt, and rcu_sched). If RCU_BOOST is
3106 set, valid values are 1-99 and the default is 1
3107 (the least-favored priority). Otherwise, when
3108 RCU_BOOST is not set, valid values are 0-99 and
3109 the default is zero (non-realtime operation).
3110
3111 rcutree.rcu_nocb_leader_stride= [KNL]
3112 Set the number of NOCB kthread groups, which
3113 defaults to the square root of the number of
3114 CPUs. Larger numbers reduces the wakeup overhead
3115 on the per-CPU grace-period kthreads, but increases
3116 that same overhead on each group's leader.
3117
3118 rcutree.qhimark= [KNL]
3119 Set threshold of queued RCU callbacks beyond which
3120 batch limiting is disabled.
3121
3122 rcutree.qlowmark= [KNL]
3123 Set threshold of queued RCU callbacks below which
3124 batch limiting is re-enabled.
3125
3126 rcutree.rcu_idle_gp_delay= [KNL]
3127 Set wakeup interval for idle CPUs that have
3128 RCU callbacks (RCU_FAST_NO_HZ=y).
3129
3130 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3131 Set wakeup interval for idle CPUs that have
3132 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3133 Lazy RCU callbacks are those which RCU can
3134 prove do nothing more than free memory.
3135
3136 rcutorture.cbflood_inter_holdoff= [KNL]
3137 Set holdoff time (jiffies) between successive
3138 callback-flood tests.
3139
3140 rcutorture.cbflood_intra_holdoff= [KNL]
3141 Set holdoff time (jiffies) between successive
3142 bursts of callbacks within a given callback-flood
3143 test.
3144
3145 rcutorture.cbflood_n_burst= [KNL]
3146 Set the number of bursts making up a given
3147 callback-flood test. Set this to zero to
3148 disable callback-flood testing.
3149
3150 rcutorture.cbflood_n_per_burst= [KNL]
3151 Set the number of callbacks to be registered
3152 in a given burst of a callback-flood test.
3153
3154 rcutorture.fqs_duration= [KNL]
3155 Set duration of force_quiescent_state bursts
3156 in microseconds.
3157
3158 rcutorture.fqs_holdoff= [KNL]
3159 Set holdoff time within force_quiescent_state bursts
3160 in microseconds.
3161
3162 rcutorture.fqs_stutter= [KNL]
3163 Set wait time between force_quiescent_state bursts
3164 in seconds.
3165
3166 rcutorture.gp_cond= [KNL]
3167 Use conditional/asynchronous update-side
3168 primitives, if available.
3169
3170 rcutorture.gp_exp= [KNL]
3171 Use expedited update-side primitives, if available.
3172
3173 rcutorture.gp_normal= [KNL]
3174 Use normal (non-expedited) asynchronous
3175 update-side primitives, if available.
3176
3177 rcutorture.gp_sync= [KNL]
3178 Use normal (non-expedited) synchronous
3179 update-side primitives, if available. If all
3180 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3181 rcutorture.gp_normal=, and rcutorture.gp_sync=
3182 are zero, rcutorture acts as if is interpreted
3183 they are all non-zero.
3184
3185 rcutorture.n_barrier_cbs= [KNL]
3186 Set callbacks/threads for rcu_barrier() testing.
3187
3188 rcutorture.nfakewriters= [KNL]
3189 Set number of concurrent RCU writers. These just
3190 stress RCU, they don't participate in the actual
3191 test, hence the "fake".
3192
3193 rcutorture.nreaders= [KNL]
3194 Set number of RCU readers. The value -1 selects
3195 N-1, where N is the number of CPUs. A value
3196 "n" less than -1 selects N-n-2, where N is again
3197 the number of CPUs. For example, -2 selects N
3198 (the number of CPUs), -3 selects N+1, and so on.
3199
3200 rcutorture.object_debug= [KNL]
3201 Enable debug-object double-call_rcu() testing.
3202
3203 rcutorture.onoff_holdoff= [KNL]
3204 Set time (s) after boot for CPU-hotplug testing.
3205
3206 rcutorture.onoff_interval= [KNL]
3207 Set time (s) between CPU-hotplug operations, or
3208 zero to disable CPU-hotplug testing.
3209
3210 rcutorture.shuffle_interval= [KNL]
3211 Set task-shuffle interval (s). Shuffling tasks
3212 allows some CPUs to go into dyntick-idle mode
3213 during the rcutorture test.
3214
3215 rcutorture.shutdown_secs= [KNL]
3216 Set time (s) after boot system shutdown. This
3217 is useful for hands-off automated testing.
3218
3219 rcutorture.stall_cpu= [KNL]
3220 Duration of CPU stall (s) to test RCU CPU stall
3221 warnings, zero to disable.
3222
3223 rcutorture.stall_cpu_holdoff= [KNL]
3224 Time to wait (s) after boot before inducing stall.
3225
3226 rcutorture.stat_interval= [KNL]
3227 Time (s) between statistics printk()s.
3228
3229 rcutorture.stutter= [KNL]
3230 Time (s) to stutter testing, for example, specifying
3231 five seconds causes the test to run for five seconds,
3232 wait for five seconds, and so on. This tests RCU's
3233 ability to transition abruptly to and from idle.
3234
3235 rcutorture.test_boost= [KNL]
3236 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3237 "Maybe" means test if the RCU implementation
3238 under test support RCU priority boosting.
3239
3240 rcutorture.test_boost_duration= [KNL]
3241 Duration (s) of each individual boost test.
3242
3243 rcutorture.test_boost_interval= [KNL]
3244 Interval (s) between each boost test.
3245
3246 rcutorture.test_no_idle_hz= [KNL]
3247 Test RCU's dyntick-idle handling. See also the
3248 rcutorture.shuffle_interval parameter.
3249
3250 rcutorture.torture_runnable= [BOOT]
3251 Start rcutorture running at boot time.
3252
3253 rcutorture.torture_type= [KNL]
3254 Specify the RCU implementation to test.
3255
3256 rcutorture.verbose= [KNL]
3257 Enable additional printk() statements.
3258
3259 rcupdate.rcu_expedited= [KNL]
3260 Use expedited grace-period primitives, for
3261 example, synchronize_rcu_expedited() instead
3262 of synchronize_rcu(). This reduces latency,
3263 but can increase CPU utilization, degrade
3264 real-time latency, and degrade energy efficiency.
3265
3266 rcupdate.rcu_cpu_stall_suppress= [KNL]
3267 Suppress RCU CPU stall warning messages.
3268
3269 rcupdate.rcu_cpu_stall_timeout= [KNL]
3270 Set timeout for RCU CPU stall warning messages.
3271
3272 rcupdate.rcu_task_stall_timeout= [KNL]
3273 Set timeout in jiffies for RCU task stall warning
3274 messages. Disable with a value less than or equal
3275 to zero.
3276
3277 rcupdate.rcu_self_test= [KNL]
3278 Run the RCU early boot self tests
3279
3280 rcupdate.rcu_self_test_bh= [KNL]
3281 Run the RCU bh early boot self tests
3282
3283 rcupdate.rcu_self_test_sched= [KNL]
3284 Run the RCU sched early boot self tests
3285
3286 rdinit= [KNL]
3287 Format: <full_path>
3288 Run specified binary instead of /init from the ramdisk,
3289 used for early userspace startup. See initrd.
3290
3291 reboot= [KNL]
3292 Format (x86 or x86_64):
3293 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3294 [[,]s[mp]#### \
3295 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3296 [[,]f[orce]
3297 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3298 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3299 reboot_force is either force or not specified,
3300 reboot_cpu is s[mp]#### with #### being the processor
3301 to be used for rebooting.
3302
3303 relax_domain_level=
3304 [KNL, SMP] Set scheduler's default relax_domain_level.
3305 See Documentation/cgroups/cpusets.txt.
3306
3307 relative_sleep_states=
3308 [SUSPEND] Use sleep state labeling where the deepest
3309 state available other than hibernation is always "mem".
3310 Format: { "0" | "1" }
3311 0 -- Traditional sleep state labels.
3312 1 -- Relative sleep state labels.
3313
3314 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
3315
3316 reservetop= [X86-32]
3317 Format: nn[KMG]
3318 Reserves a hole at the top of the kernel virtual
3319 address space.
3320
3321 reservelow= [X86]
3322 Format: nn[K]
3323 Set the amount of memory to reserve for BIOS at
3324 the bottom of the address space.
3325
3326 reset_devices [KNL] Force drivers to reset the underlying device
3327 during initialization.
3328
3329 resume= [SWSUSP]
3330 Specify the partition device for software suspend
3331 Format:
3332 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3333
3334 resume_offset= [SWSUSP]
3335 Specify the offset from the beginning of the partition
3336 given by "resume=" at which the swap header is located,
3337 in <PAGE_SIZE> units (needed only for swap files).
3338 See Documentation/power/swsusp-and-swap-files.txt
3339
3340 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3341 read the resume files
3342
3343 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3344 Useful for devices that are detected asynchronously
3345 (e.g. USB and MMC devices).
3346
3347 hibernate= [HIBERNATION]
3348 noresume Don't check if there's a hibernation image
3349 present during boot.
3350 nocompress Don't compress/decompress hibernation images.
3351 no Disable hibernation and resume.
3352
3353 retain_initrd [RAM] Keep initrd memory after extraction
3354
3355 rfkill.default_state=
3356 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3357 etc. communication is blocked by default.
3358 1 Unblocked.
3359
3360 rfkill.master_switch_mode=
3361 0 The "airplane mode" button does nothing.
3362 1 The "airplane mode" button toggles between everything
3363 blocked and the previous configuration.
3364 2 The "airplane mode" button toggles between everything
3365 blocked and everything unblocked.
3366
3367 rhash_entries= [KNL,NET]
3368 Set number of hash buckets for route cache
3369
3370 ro [KNL] Mount root device read-only on boot
3371
3372 root= [KNL] Root filesystem
3373 See name_to_dev_t comment in init/do_mounts.c.
3374
3375 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
3376 mount the root filesystem
3377
3378 rootflags= [KNL] Set root filesystem mount option string
3379
3380 rootfstype= [KNL] Set root filesystem type
3381
3382 rootwait [KNL] Wait (indefinitely) for root device to show up.
3383 Useful for devices that are detected asynchronously
3384 (e.g. USB and MMC devices).
3385
3386 rproc_mem=nn[KMG][@address]
3387 [KNL,ARM,CMA] Remoteproc physical memory block.
3388 Memory area to be used by remote processor image,
3389 managed by CMA.
3390
3391 rw [KNL] Mount root device read-write on boot
3392
3393 S [KNL] Run init in single mode
3394
3395 s390_iommu= [HW,S390]
3396 Set s390 IOTLB flushing mode
3397 strict
3398 With strict flushing every unmap operation will result in
3399 an IOTLB flush. Default is lazy flushing before reuse,
3400 which is faster.
3401
3402 sa1100ir [NET]
3403 See drivers/net/irda/sa1100_ir.c.
3404
3405 sbni= [NET] Granch SBNI12 leased line adapter
3406
3407 sched_debug [KNL] Enables verbose scheduler debug messages.
3408
3409 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
3410 xtime_lock contention on larger systems, and/or RCU lock
3411 contention on all systems with CONFIG_MAXSMP set.
3412 Format: { "0" | "1" }
3413 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3414 1 -- enable.
3415 Note: increases power consumption, thus should only be
3416 enabled if running jitter sensitive (HPC/RT) workloads.
3417
3418 security= [SECURITY] Choose a security module to enable at boot.
3419 If this boot parameter is not specified, only the first
3420 security module asking for security registration will be
3421 loaded. An invalid security module name will be treated
3422 as if no module has been chosen.
3423
3424 selinux= [SELINUX] Disable or enable SELinux at boot time.
3425 Format: { "0" | "1" }
3426 See security/selinux/Kconfig help text.
3427 0 -- disable.
3428 1 -- enable.
3429 Default value is set via kernel config option.
3430 If enabled at boot time, /selinux/disable can be used
3431 later to disable prior to initial policy load.
3432
3433 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3434 Format: { "0" | "1" }
3435 See security/apparmor/Kconfig help text
3436 0 -- disable.
3437 1 -- enable.
3438 Default value is set via kernel config option.
3439
3440 serialnumber [BUGS=X86-32]
3441
3442 shapers= [NET]
3443 Maximal number of shapers.
3444
3445 show_msr= [x86] show boot-time MSR settings
3446 Format: { <integer> }
3447 Show boot-time (BIOS-initialized) MSR settings.
3448 The parameter means the number of CPUs to show,
3449 for example 1 means boot CPU only.
3450
3451 simeth= [IA-64]
3452 simscsi=
3453
3454 slram= [HW,MTD]
3455
3456 slab_nomerge [MM]
3457 Disable merging of slabs with similar size. May be
3458 necessary if there is some reason to distinguish
3459 allocs to different slabs. Debug options disable
3460 merging on their own.
3461 For more information see Documentation/vm/slub.txt.
3462
3463 slab_max_order= [MM, SLAB]
3464 Determines the maximum allowed order for slabs.
3465 A high setting may cause OOMs due to memory
3466 fragmentation. Defaults to 1 for systems with
3467 more than 32MB of RAM, 0 otherwise.
3468
3469 slub_debug[=options[,slabs]] [MM, SLUB]
3470 Enabling slub_debug allows one to determine the
3471 culprit if slab objects become corrupted. Enabling
3472 slub_debug can create guard zones around objects and
3473 may poison objects when not in use. Also tracks the
3474 last alloc / free. For more information see
3475 Documentation/vm/slub.txt.
3476
3477 slub_max_order= [MM, SLUB]
3478 Determines the maximum allowed order for slabs.
3479 A high setting may cause OOMs due to memory
3480 fragmentation. For more information see
3481 Documentation/vm/slub.txt.
3482
3483 slub_min_objects= [MM, SLUB]
3484 The minimum number of objects per slab. SLUB will
3485 increase the slab order up to slub_max_order to
3486 generate a sufficiently large slab able to contain
3487 the number of objects indicated. The higher the number
3488 of objects the smaller the overhead of tracking slabs
3489 and the less frequently locks need to be acquired.
3490 For more information see Documentation/vm/slub.txt.
3491
3492 slub_min_order= [MM, SLUB]
3493 Determines the minimum page order for slabs. Must be
3494 lower than slub_max_order.
3495 For more information see Documentation/vm/slub.txt.
3496
3497 slub_nomerge [MM, SLUB]
3498 Same with slab_nomerge. This is supported for legacy.
3499 See slab_nomerge for more information.
3500
3501 smart2= [HW]
3502 Format: <io1>[,<io2>[,...,<io8>]]
3503
3504 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3505 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3506 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3507 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3508 smsc-ircc2.ircc_irq= [HW] IRQ line
3509 smsc-ircc2.ircc_dma= [HW] DMA channel
3510 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3511 0: Toshiba Satellite 1800 (GP data pin select)
3512 1: Fast pin select (default)
3513 2: ATC IRMode
3514
3515 softlockup_panic=
3516 [KNL] Should the soft-lockup detector generate panics.
3517 Format: <integer>
3518
3519 softlockup_all_cpu_backtrace=
3520 [KNL] Should the soft-lockup detector generate
3521 backtraces on all cpus.
3522 Format: <integer>
3523
3524 sonypi.*= [HW] Sony Programmable I/O Control Device driver
3525 See Documentation/laptops/sonypi.txt
3526
3527 spia_io_base= [HW,MTD]
3528 spia_fio_base=
3529 spia_pedr=
3530 spia_peddr=
3531
3532 stacktrace [FTRACE]
3533 Enabled the stack tracer on boot up.
3534
3535 stacktrace_filter=[function-list]
3536 [FTRACE] Limit the functions that the stack tracer
3537 will trace at boot up. function-list is a comma separated
3538 list of functions. This list can be changed at run
3539 time by the stack_trace_filter file in the debugfs
3540 tracing directory. Note, this enables stack tracing
3541 and the stacktrace above is not needed.
3542
3543 sti= [PARISC,HW]
3544 Format: <num>
3545 Set the STI (builtin display/keyboard on the HP-PARISC
3546 machines) console (graphic card) which should be used
3547 as the initial boot-console.
3548 See also comment in drivers/video/console/sticore.c.
3549
3550 sti_font= [HW]
3551 See comment in drivers/video/console/sticore.c.
3552
3553 stifb= [HW]
3554 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3555
3556 sunrpc.min_resvport=
3557 sunrpc.max_resvport=
3558 [NFS,SUNRPC]
3559 SunRPC servers often require that client requests
3560 originate from a privileged port (i.e. a port in the
3561 range 0 < portnr < 1024).
3562 An administrator who wishes to reserve some of these
3563 ports for other uses may adjust the range that the
3564 kernel's sunrpc client considers to be privileged
3565 using these two parameters to set the minimum and
3566 maximum port values.
3567
3568 sunrpc.pool_mode=
3569 [NFS]
3570 Control how the NFS server code allocates CPUs to
3571 service thread pools. Depending on how many NICs
3572 you have and where their interrupts are bound, this
3573 option will affect which CPUs will do NFS serving.
3574 Note: this parameter cannot be changed while the
3575 NFS server is running.
3576
3577 auto the server chooses an appropriate mode
3578 automatically using heuristics
3579 global a single global pool contains all CPUs
3580 percpu one pool for each CPU
3581 pernode one pool for each NUMA node (equivalent
3582 to global on non-NUMA machines)
3583
3584 sunrpc.tcp_slot_table_entries=
3585 sunrpc.udp_slot_table_entries=
3586 [NFS,SUNRPC]
3587 Sets the upper limit on the number of simultaneous
3588 RPC calls that can be sent from the client to a
3589 server. Increasing these values may allow you to
3590 improve throughput, but will also increase the
3591 amount of memory reserved for use by the client.
3592
3593 suspend.pm_test_delay=
3594 [SUSPEND]
3595 Sets the number of seconds to remain in a suspend test
3596 mode before resuming the system (see
3597 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
3598 is set. Default value is 5.
3599
3600 swapaccount=[0|1]
3601 [KNL] Enable accounting of swap in memory resource
3602 controller if no parameter or 1 is given or disable
3603 it if 0 is given (See Documentation/cgroups/memory.txt)
3604
3605 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
3606 Format: { <int> | force }
3607 <int> -- Number of I/O TLB slabs
3608 force -- force using of bounce buffers even if they
3609 wouldn't be automatically used by the kernel
3610
3611 switches= [HW,M68k]
3612
3613 sysfs.deprecated=0|1 [KNL]
3614 Enable/disable old style sysfs layout for old udev
3615 on older distributions. When this option is enabled
3616 very new udev will not work anymore. When this option
3617 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
3618 in older udev will not work anymore.
3619 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
3620 the kernel configuration.
3621
3622 sysrq_always_enabled
3623 [KNL]
3624 Ignore sysrq setting - this boot parameter will
3625 neutralize any effect of /proc/sys/kernel/sysrq.
3626 Useful for debugging.
3627
3628 tcpmhash_entries= [KNL,NET]
3629 Set the number of tcp_metrics_hash slots.
3630 Default value is 8192 or 16384 depending on total
3631 ram pages. This is used to specify the TCP metrics
3632 cache size. See Documentation/networking/ip-sysctl.txt
3633 "tcp_no_metrics_save" section for more details.
3634
3635 tdfx= [HW,DRM]
3636
3637 test_suspend= [SUSPEND][,N]
3638 Specify "mem" (for Suspend-to-RAM) or "standby" (for
3639 standby suspend) or "freeze" (for suspend type freeze)
3640 as the system sleep state during system startup with
3641 the optional capability to repeat N number of times.
3642 The system is woken from this state using a
3643 wakeup-capable RTC alarm.
3644
3645 thash_entries= [KNL,NET]
3646 Set number of hash buckets for TCP connection
3647
3648 thermal.act= [HW,ACPI]
3649 -1: disable all active trip points in all thermal zones
3650 <degrees C>: override all lowest active trip points
3651
3652 thermal.crt= [HW,ACPI]
3653 -1: disable all critical trip points in all thermal zones
3654 <degrees C>: override all critical trip points
3655
3656 thermal.nocrt= [HW,ACPI]
3657 Set to disable actions on ACPI thermal zone
3658 critical and hot trip points.
3659
3660 thermal.off= [HW,ACPI]
3661 1: disable ACPI thermal control
3662
3663 thermal.psv= [HW,ACPI]
3664 -1: disable all passive trip points
3665 <degrees C>: override all passive trip points to this
3666 value
3667
3668 thermal.tzp= [HW,ACPI]
3669 Specify global default ACPI thermal zone polling rate
3670 <deci-seconds>: poll all this frequency
3671 0: no polling (default)
3672
3673 threadirqs [KNL]
3674 Force threading of all interrupt handlers except those
3675 marked explicitly IRQF_NO_THREAD.
3676
3677 tmem [KNL,XEN]
3678 Enable the Transcendent memory driver if built-in.
3679
3680 tmem.cleancache=0|1 [KNL, XEN]
3681 Default is on (1). Disable the usage of the cleancache
3682 API to send anonymous pages to the hypervisor.
3683
3684 tmem.frontswap=0|1 [KNL, XEN]
3685 Default is on (1). Disable the usage of the frontswap
3686 API to send swap pages to the hypervisor. If disabled
3687 the selfballooning and selfshrinking are force disabled.
3688
3689 tmem.selfballooning=0|1 [KNL, XEN]
3690 Default is on (1). Disable the driving of swap pages
3691 to the hypervisor.
3692
3693 tmem.selfshrinking=0|1 [KNL, XEN]
3694 Default is on (1). Partial swapoff that immediately
3695 transfers pages from Xen hypervisor back to the
3696 kernel based on different criteria.
3697
3698 topology= [S390]
3699 Format: {off | on}
3700 Specify if the kernel should make use of the cpu
3701 topology information if the hardware supports this.
3702 The scheduler will make use of this information and
3703 e.g. base its process migration decisions on it.
3704 Default is on.
3705
3706 topology_updates= [KNL, PPC, NUMA]
3707 Format: {off}
3708 Specify if the kernel should ignore (off)
3709 topology updates sent by the hypervisor to this
3710 LPAR.
3711
3712 tp720= [HW,PS2]
3713
3714 tpm_suspend_pcr=[HW,TPM]
3715 Format: integer pcr id
3716 Specify that at suspend time, the tpm driver
3717 should extend the specified pcr with zeros,
3718 as a workaround for some chips which fail to
3719 flush the last written pcr on TPM_SaveState.
3720 This will guarantee that all the other pcrs
3721 are saved.
3722
3723 trace_buf_size=nn[KMG]
3724 [FTRACE] will set tracing buffer size on each cpu.
3725
3726 trace_event=[event-list]
3727 [FTRACE] Set and start specified trace events in order
3728 to facilitate early boot debugging.
3729 See also Documentation/trace/events.txt
3730
3731 trace_options=[option-list]
3732 [FTRACE] Enable or disable tracer options at boot.
3733 The option-list is a comma delimited list of options
3734 that can be enabled or disabled just as if you were
3735 to echo the option name into
3736
3737 /sys/kernel/debug/tracing/trace_options
3738
3739 For example, to enable stacktrace option (to dump the
3740 stack trace of each event), add to the command line:
3741
3742 trace_options=stacktrace
3743
3744 See also Documentation/trace/ftrace.txt "trace options"
3745 section.
3746
3747 tp_printk[FTRACE]
3748 Have the tracepoints sent to printk as well as the
3749 tracing ring buffer. This is useful for early boot up
3750 where the system hangs or reboots and does not give the
3751 option for reading the tracing buffer or performing a
3752 ftrace_dump_on_oops.
3753
3754 To turn off having tracepoints sent to printk,
3755 echo 0 > /proc/sys/kernel/tracepoint_printk
3756 Note, echoing 1 into this file without the
3757 tracepoint_printk kernel cmdline option has no effect.
3758
3759 ** CAUTION **
3760
3761 Having tracepoints sent to printk() and activating high
3762 frequency tracepoints such as irq or sched, can cause
3763 the system to live lock.
3764
3765 traceoff_on_warning
3766 [FTRACE] enable this option to disable tracing when a
3767 warning is hit. This turns off "tracing_on". Tracing can
3768 be enabled again by echoing '1' into the "tracing_on"
3769 file located in /sys/kernel/debug/tracing/
3770
3771 This option is useful, as it disables the trace before
3772 the WARNING dump is called, which prevents the trace to
3773 be filled with content caused by the warning output.
3774
3775 This option can also be set at run time via the sysctl
3776 option: kernel/traceoff_on_warning
3777
3778 transparent_hugepage=
3779 [KNL]
3780 Format: [always|madvise|never]
3781 Can be used to control the default behavior of the system
3782 with respect to transparent hugepages.
3783 See Documentation/vm/transhuge.txt for more details.
3784
3785 tsc= Disable clocksource stability checks for TSC.
3786 Format: <string>
3787 [x86] reliable: mark tsc clocksource as reliable, this
3788 disables clocksource verification at runtime, as well
3789 as the stability checks done at bootup. Used to enable
3790 high-resolution timer mode on older hardware, and in
3791 virtualized environment.
3792 [x86] noirqtime: Do not use TSC to do irq accounting.
3793 Used to run time disable IRQ_TIME_ACCOUNTING on any
3794 platforms where RDTSC is slow and this accounting
3795 can add overhead.
3796
3797 turbografx.map[2|3]= [HW,JOY]
3798 TurboGraFX parallel port interface
3799 Format:
3800 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
3801 See also Documentation/input/joystick-parport.txt
3802
3803 udbg-immortal [PPC] When debugging early kernel crashes that
3804 happen after console_init() and before a proper
3805 console driver takes over, this boot options might
3806 help "seeing" what's going on.
3807
3808 uhash_entries= [KNL,NET]
3809 Set number of hash buckets for UDP/UDP-Lite connections
3810
3811 uhci-hcd.ignore_oc=
3812 [USB] Ignore overcurrent events (default N).
3813 Some badly-designed motherboards generate lots of
3814 bogus events, for ports that aren't wired to
3815 anything. Set this parameter to avoid log spamming.
3816 Note that genuine overcurrent events won't be
3817 reported either.
3818
3819 unknown_nmi_panic
3820 [X86] Cause panic on unknown NMI.
3821
3822 usbcore.authorized_default=
3823 [USB] Default USB device authorization:
3824 (default -1 = authorized except for wireless USB,
3825 0 = not authorized, 1 = authorized)
3826
3827 usbcore.autosuspend=
3828 [USB] The autosuspend time delay (in seconds) used
3829 for newly-detected USB devices (default 2). This
3830 is the time required before an idle device will be
3831 autosuspended. Devices for which the delay is set
3832 to a negative value won't be autosuspended at all.
3833
3834 usbcore.usbfs_snoop=
3835 [USB] Set to log all usbfs traffic (default 0 = off).
3836
3837 usbcore.blinkenlights=
3838 [USB] Set to cycle leds on hubs (default 0 = off).
3839
3840 usbcore.old_scheme_first=
3841 [USB] Start with the old device initialization
3842 scheme (default 0 = off).
3843
3844 usbcore.usbfs_memory_mb=
3845 [USB] Memory limit (in MB) for buffers allocated by
3846 usbfs (default = 16, 0 = max = 2047).
3847
3848 usbcore.use_both_schemes=
3849 [USB] Try the other device initialization scheme
3850 if the first one fails (default 1 = enabled).
3851
3852 usbcore.initial_descriptor_timeout=
3853 [USB] Specifies timeout for the initial 64-byte
3854 USB_REQ_GET_DESCRIPTOR request in milliseconds
3855 (default 5000 = 5.0 seconds).
3856
3857 usbhid.mousepoll=
3858 [USBHID] The interval which mice are to be polled at.
3859
3860 usb-storage.delay_use=
3861 [UMS] The delay in seconds before a new device is
3862 scanned for Logical Units (default 1).
3863
3864 usb-storage.quirks=
3865 [UMS] A list of quirks entries to supplement or
3866 override the built-in unusual_devs list. List
3867 entries are separated by commas. Each entry has
3868 the form VID:PID:Flags where VID and PID are Vendor
3869 and Product ID values (4-digit hex numbers) and
3870 Flags is a set of characters, each corresponding
3871 to a common usb-storage quirk flag as follows:
3872 a = SANE_SENSE (collect more than 18 bytes
3873 of sense data);
3874 b = BAD_SENSE (don't collect more than 18
3875 bytes of sense data);
3876 c = FIX_CAPACITY (decrease the reported
3877 device capacity by one sector);
3878 d = NO_READ_DISC_INFO (don't use
3879 READ_DISC_INFO command);
3880 e = NO_READ_CAPACITY_16 (don't use
3881 READ_CAPACITY_16 command);
3882 f = NO_REPORT_OPCODES (don't use report opcodes
3883 command, uas only);
3884 g = MAX_SECTORS_240 (don't transfer more than
3885 240 sectors at a time, uas only);
3886 h = CAPACITY_HEURISTICS (decrease the
3887 reported device capacity by one
3888 sector if the number is odd);
3889 i = IGNORE_DEVICE (don't bind to this
3890 device);
3891 l = NOT_LOCKABLE (don't try to lock and
3892 unlock ejectable media);
3893 m = MAX_SECTORS_64 (don't transfer more
3894 than 64 sectors = 32 KB at a time);
3895 n = INITIAL_READ10 (force a retry of the
3896 initial READ(10) command);
3897 o = CAPACITY_OK (accept the capacity
3898 reported by the device);
3899 p = WRITE_CACHE (the device cache is ON
3900 by default);
3901 r = IGNORE_RESIDUE (the device reports
3902 bogus residue values);
3903 s = SINGLE_LUN (the device has only one
3904 Logical Unit);
3905 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
3906 commands, uas only);
3907 u = IGNORE_UAS (don't bind to the uas driver);
3908 w = NO_WP_DETECT (don't test whether the
3909 medium is write-protected).
3910 Example: quirks=0419:aaf5:rl,0421:0433:rc
3911
3912 user_debug= [KNL,ARM]
3913 Format: <int>
3914 See arch/arm/Kconfig.debug help text.
3915 1 - undefined instruction events
3916 2 - system calls
3917 4 - invalid data aborts
3918 8 - SIGSEGV faults
3919 16 - SIGBUS faults
3920 Example: user_debug=31
3921
3922 userpte=
3923 [X86] Flags controlling user PTE allocations.
3924
3925 nohigh = do not allocate PTE pages in
3926 HIGHMEM regardless of setting
3927 of CONFIG_HIGHPTE.
3928
3929 vdso= [X86,SH]
3930 On X86_32, this is an alias for vdso32=. Otherwise:
3931
3932 vdso=1: enable VDSO (the default)
3933 vdso=0: disable VDSO mapping
3934
3935 vdso32= [X86] Control the 32-bit vDSO
3936 vdso32=1: enable 32-bit VDSO
3937 vdso32=0 or vdso32=2: disable 32-bit VDSO
3938
3939 See the help text for CONFIG_COMPAT_VDSO for more
3940 details. If CONFIG_COMPAT_VDSO is set, the default is
3941 vdso32=0; otherwise, the default is vdso32=1.
3942
3943 For compatibility with older kernels, vdso32=2 is an
3944 alias for vdso32=0.
3945
3946 Try vdso32=0 if you encounter an error that says:
3947 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
3948
3949 vector= [IA-64,SMP]
3950 vector=percpu: enable percpu vector domain
3951
3952 video= [FB] Frame buffer configuration
3953 See Documentation/fb/modedb.txt.
3954
3955 video.brightness_switch_enabled= [0,1]
3956 If set to 1, on receiving an ACPI notify event
3957 generated by hotkey, video driver will adjust brightness
3958 level and then send out the event to user space through
3959 the allocated input device; If set to 0, video driver
3960 will only send out the event without touching backlight
3961 brightness level.
3962 default: 1
3963
3964 virtio_mmio.device=
3965 [VMMIO] Memory mapped virtio (platform) device.
3966
3967 <size>@<baseaddr>:<irq>[:<id>]
3968 where:
3969 <size> := size (can use standard suffixes
3970 like K, M and G)
3971 <baseaddr> := physical base address
3972 <irq> := interrupt number (as passed to
3973 request_irq())
3974 <id> := (optional) platform device id
3975 example:
3976 virtio_mmio.device=1K@0x100b0000:48:7
3977
3978 Can be used multiple times for multiple devices.
3979
3980 vga= [BOOT,X86-32] Select a particular video mode
3981 See Documentation/x86/boot.txt and
3982 Documentation/svga.txt.
3983 Use vga=ask for menu.
3984 This is actually a boot loader parameter; the value is
3985 passed to the kernel using a special protocol.
3986
3987 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
3988 size of <nn>. This can be used to increase the
3989 minimum size (128MB on x86). It can also be used to
3990 decrease the size and leave more room for directly
3991 mapped kernel RAM.
3992
3993 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
3994 Format: <command>
3995
3996 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
3997 Format: <command>
3998
3999 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4000 Format: <command>
4001
4002 vsyscall= [X86-64]
4003 Controls the behavior of vsyscalls (i.e. calls to
4004 fixed addresses of 0xffffffffff600x00 from legacy
4005 code). Most statically-linked binaries and older
4006 versions of glibc use these calls. Because these
4007 functions are at fixed addresses, they make nice
4008 targets for exploits that can control RIP.
4009
4010 emulate [default] Vsyscalls turn into traps and are
4011 emulated reasonably safely.
4012
4013 native Vsyscalls are native syscall instructions.
4014 This is a little bit faster than trapping
4015 and makes a few dynamic recompilers work
4016 better than they would in emulation mode.
4017 It also makes exploits much easier to write.
4018
4019 none Vsyscalls don't work at all. This makes
4020 them quite hard to use for exploits but
4021 might break your system.
4022
4023 vt.color= [VT] Default text color.
4024 Format: 0xYX, X = foreground, Y = background.
4025 Default: 0x07 = light gray on black.
4026
4027 vt.cur_default= [VT] Default cursor shape.
4028 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4029 the parameters of the <Esc>[?A;B;Cc escape sequence;
4030 see VGA-softcursor.txt. Default: 2 = underline.
4031
4032 vt.default_blu= [VT]
4033 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4034 Change the default blue palette of the console.
4035 This is a 16-member array composed of values
4036 ranging from 0-255.
4037
4038 vt.default_grn= [VT]
4039 Format: <green0>,<green1>,<green2>,...,<green15>
4040 Change the default green palette of the console.
4041 This is a 16-member array composed of values
4042 ranging from 0-255.
4043
4044 vt.default_red= [VT]
4045 Format: <red0>,<red1>,<red2>,...,<red15>
4046 Change the default red palette of the console.
4047 This is a 16-member array composed of values
4048 ranging from 0-255.
4049
4050 vt.default_utf8=
4051 [VT]
4052 Format=<0|1>
4053 Set system-wide default UTF-8 mode for all tty's.
4054 Default is 1, i.e. UTF-8 mode is enabled for all
4055 newly opened terminals.
4056
4057 vt.global_cursor_default=
4058 [VT]
4059 Format=<-1|0|1>
4060 Set system-wide default for whether a cursor
4061 is shown on new VTs. Default is -1,
4062 i.e. cursors will be created by default unless
4063 overridden by individual drivers. 0 will hide
4064 cursors, 1 will display them.
4065
4066 vt.italic= [VT] Default color for italic text; 0-15.
4067 Default: 2 = green.
4068
4069 vt.underline= [VT] Default color for underlined text; 0-15.
4070 Default: 3 = cyan.
4071
4072 watchdog timers [HW,WDT] For information on watchdog timers,
4073 see Documentation/watchdog/watchdog-parameters.txt
4074 or other driver-specific files in the
4075 Documentation/watchdog/ directory.
4076
4077 workqueue.disable_numa
4078 By default, all work items queued to unbound
4079 workqueues are affine to the NUMA nodes they're
4080 issued on, which results in better behavior in
4081 general. If NUMA affinity needs to be disabled for
4082 whatever reason, this option can be used. Note
4083 that this also can be controlled per-workqueue for
4084 workqueues visible under /sys/bus/workqueue/.
4085
4086 workqueue.power_efficient
4087 Per-cpu workqueues are generally preferred because
4088 they show better performance thanks to cache
4089 locality; unfortunately, per-cpu workqueues tend to
4090 be more power hungry than unbound workqueues.
4091
4092 Enabling this makes the per-cpu workqueues which
4093 were observed to contribute significantly to power
4094 consumption unbound, leading to measurably lower
4095 power usage at the cost of small performance
4096 overhead.
4097
4098 The default value of this parameter is determined by
4099 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4100
4101 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
4102 default x2apic cluster mode on platforms
4103 supporting x2apic.
4104
4105 x86_intel_mid_timer= [X86-32,APBT]
4106 Choose timer option for x86 Intel MID platform.
4107 Two valid options are apbt timer only and lapic timer
4108 plus one apbt timer for broadcast timer.
4109 x86_intel_mid_timer=apbt_only | lapic_and_apbt
4110
4111 xen_512gb_limit [KNL,X86-64,XEN]
4112 Restricts the kernel running paravirtualized under Xen
4113 to use only up to 512 GB of RAM. The reason to do so is
4114 crash analysis tools and Xen tools for doing domain
4115 save/restore/migration must be enabled to handle larger
4116 domains.
4117
4118 xen_emul_unplug= [HW,X86,XEN]
4119 Unplug Xen emulated devices
4120 Format: [unplug0,][unplug1]
4121 ide-disks -- unplug primary master IDE devices
4122 aux-ide-disks -- unplug non-primary-master IDE devices
4123 nics -- unplug network devices
4124 all -- unplug all emulated devices (NICs and IDE disks)
4125 unnecessary -- unplugging emulated devices is
4126 unnecessary even if the host did not respond to
4127 the unplug protocol
4128 never -- do not unplug even if version check succeeds
4129
4130 xen_nopvspin [X86,XEN]
4131 Disables the ticketlock slowpath using Xen PV
4132 optimizations.
4133
4134 xen_nopv [X86]
4135 Disables the PV optimizations forcing the HVM guest to
4136 run as generic HVM guest with no PV drivers.
4137
4138 xirc2ps_cs= [NET,PCMCIA]
4139 Format:
4140 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
4141
4142 ______________________________________________________________________
4143
4144 TODO:
4145
4146 Add more DRM drivers.