blob: ef77c55e87e2f560361e012c865d3a2361001f1e [file] [log] [blame]
Al Stoneb10d79f2015-03-24 14:02:41 +00001 acpi= [HW,ACPI,X86,ARM64]
Bernhard Walle03d926f2007-03-06 02:29:44 -08002 Advanced Configuration and Power Interface
Ard Biesheuvel6a1f5472016-04-12 16:09:11 +02003 Format: { force | on | off | strict | noirq | rsdt |
Rami Rosene58d1542015-09-26 19:27:57 +03004 copy_dsdt }
Linus Torvalds1da177e2005-04-16 15:20:36 -07005 force -- enable ACPI if default was off
Ard Biesheuvel6a1f5472016-04-12 16:09:11 +02006 on -- enable ACPI but allow fallback to DT [arm64]
Linus Torvalds1da177e2005-04-16 15:20:36 -07007 off -- disable ACPI if default was on
8 noirq -- do not use ACPI for IRQ routing
Randy Dunlapa9913042005-10-23 12:57:11 -07009 strict -- Be less tolerant of platforms that are not
Linus Torvalds1da177e2005-04-16 15:20:36 -070010 strictly ACPI specification compliant.
Zhao Yakui237889b2008-12-17 16:55:18 +080011 rsdt -- prefer RSDT over (default) XSDT
Lin Mingaa2110c2010-04-08 14:34:27 +080012 copy_dsdt -- copy DSDT to memory
Ard Biesheuvel6a1f5472016-04-12 16:09:11 +020013 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
14 are available
Linus Torvalds1da177e2005-04-16 15:20:36 -070015
Paul Bolle395cf962011-08-15 02:02:26 +020016 See also Documentation/power/runtime_pm.txt, pci=noacpi
Linus Torvalds1da177e2005-04-16 15:20:36 -070017
Len Browna1fdcc02007-03-11 03:26:14 -040018 acpi_apic_instance= [ACPI, IOAPIC]
19 Format: <int>
20 2: use 2nd APIC table, if available
21 1,0: use 1st APIC table
Len Brown4e381a42007-03-30 14:16:10 -040022 default: 0
Len Browna1fdcc02007-03-11 03:26:14 -040023
Thomas Renningerc3d6de62008-08-01 17:37:55 +020024 acpi_backlight= [HW,ACPI]
25 acpi_backlight=vendor
26 acpi_backlight=video
27 If set to vendor, prefer vendor specific driver
28 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
29 of the ACPI video.ko driver.
30
Colin Ian Kingb2ca5da2016-01-21 17:05:47 +000031 acpi_force_32bit_fadt_addr
32 force FADT to use 32 bit addresses rather than the
33 64 bit X_* addresses. Some firmware have broken 64
34 bit addresses for force ACPI ignore these and use
35 the older legacy 32 bit addresses.
36
Dominik Brodowskief694492015-05-14 15:31:28 +020037 acpica_no_return_repair [HW, ACPI]
38 Disable AML predefined validation mechanism
39 This mechanism can repair the evaluation result to make
40 the return objects more ACPI specification compliant.
41 This option is useful for developers to identify the
42 root cause of an AML interpreter issue when the issue
43 has something to do with the repair mechanism.
44
Bjorn Helgaasa0d84a92008-11-07 16:58:05 -070045 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
46 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
Linus Torvalds1da177e2005-04-16 15:20:36 -070047 Format: <int>
Bjorn Helgaasa0d84a92008-11-07 16:58:05 -070048 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
49 debug output. Bits in debug_layer correspond to a
50 _COMPONENT in an ACPI source file, e.g.,
51 #define _COMPONENT ACPI_PCI_COMPONENT
52 Bits in debug_level correspond to a level in
53 ACPI_DEBUG_PRINT statements, e.g.,
54 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
Bjorn Helgaase76f4272008-11-13 17:30:13 -060055 The debug_level mask defaults to "info". See
56 Documentation/acpi/debug.txt for more information about
57 debug layers and levels.
Linus Torvalds1da177e2005-04-16 15:20:36 -070058
Bjorn Helgaase76f4272008-11-13 17:30:13 -060059 Enable processor driver info messages:
60 acpi.debug_layer=0x20000000
61 Enable PCI/PCI interrupt routing info messages:
62 acpi.debug_layer=0x400000
Bjorn Helgaasa0d84a92008-11-07 16:58:05 -070063 Enable AML "Debug" output, i.e., stores to the Debug
64 object while interpreting AML:
65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
Bjorn Helgaasa0d84a92008-11-07 16:58:05 -070066 Enable all messages related to ACPI hardware:
67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
Thomas Renninger36eec5e2008-08-11 15:03:20 +020068
Bjorn Helgaasa0d84a92008-11-07 16:58:05 -070069 Some values produce so much output that the system is
70 unusable. The "log_buf_len" parameter may be useful
71 if you need to capture more output.
Zhang Ruif9891062007-04-24 13:53:22 +080072
Dominik Brodowskief694492015-05-14 15:31:28 +020073 acpi_enforce_resources= [ACPI]
74 { strict | lax | no }
75 Check for resource conflicts between native drivers
76 and ACPI OperationRegions (SystemIO and SystemMemory
77 only). IO ports and memory declared in ACPI might be
78 used by the ACPI subsystem in arbitrary AML code and
79 can interfere with legacy drivers.
80 strict (default): access to resources claimed by ACPI
81 is denied; legacy drivers trying to access reserved
82 resources will fail to bind to device using them.
83 lax: access to resources claimed by ACPI is allowed;
84 legacy drivers trying to access reserved resources
85 will bind successfully but a warning message is logged.
86 no: ACPI OperationRegions are not marked as reserved,
87 no further checks are performed.
88
Lv Zheng4fc0a7e2014-05-31 08:15:02 +080089 acpi_force_table_verification [HW,ACPI]
90 Enable table checksum verification during early stage.
91 By default, this is disabled due to x86 early mapping
92 size limitation.
93
Randy Dunlap0cb55ad2009-04-05 15:55:22 -070094 acpi_irq_balance [HW,ACPI]
95 ACPI will balance active IRQs
96 default in APIC mode
97
98 acpi_irq_nobalance [HW,ACPI]
99 ACPI will not move active IRQs (default)
100 default in PIC mode
101
102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
103 Format: <irq>,<irq>...
104
105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
106 use by PCI
107 Format: <irq>,<irq>...
108
Lv Zheng9c4aa1e2016-12-16 12:07:57 +0800109 acpi_mask_gpe= [HW,ACPI]
110 Due to the existence of _Lxx/_Exx, some GPEs triggered
111 by unsupported hardware/firmware features can result in
112 GPE floodings that cannot be automatically disabled by
113 the GPE dispatcher.
114 This facility can be used to prevent such uncontrolled
115 GPE floodings.
116 Format: <int>
117 Support masking of GPEs numbered from 0x00 to 0x7f.
118
Lv Zheng08e1d7c2014-03-24 14:49:22 +0800119 acpi_no_auto_serialize [HW,ACPI]
120 Disable auto-serialization of AML methods
Bob Moore22b5afc2014-03-24 14:49:00 +0800121 AML control methods that contain the opcodes to create
122 named objects will be marked as "Serialized" by the
123 auto-serialization feature.
Lv Zheng08e1d7c2014-03-24 14:49:22 +0800124 This feature is enabled by default.
125 This option allows to turn off the feature.
Bob Moore22b5afc2014-03-24 14:49:00 +0800126
Dominik Brodowskief694492015-05-14 15:31:28 +0200127 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
128 kernels.
129
Lv Zhenga94e88c2014-04-04 12:39:11 +0800130 acpi_no_static_ssdt [HW,ACPI]
131 Disable installation of static SSDTs at early boot time
132 By default, SSDTs contained in the RSDT/XSDT will be
133 installed automatically and they will appear under
134 /sys/firmware/acpi/tables.
135 This option turns off this feature.
136 Note that specifying this option does not affect
137 dynamic table installation which will install SSDT
138 tables to /sys/firmware/acpi/tables/dynamic.
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700139
Dominik Brodowskief694492015-05-14 15:31:28 +0200140 acpi_rsdp= [ACPI,EFI,KEXEC]
141 Pass the RSDP address to the kernel, mostly used
142 on machines running EFI runtime service to boot the
143 second kernel for kdump.
Lv Zheng4dde5072014-02-11 11:01:52 +0800144
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700145 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
146 Format: To spoof as Windows 98: ="Microsoft Windows"
147
Rafael J. Wysocki18d78b62015-07-03 01:06:00 +0200148 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
149 of 2 which is mandated by ACPI 6) as the supported ACPI
150 specification revision (when using this switch, it may
151 be necessary to carry out a cold reboot _twice_ in a
152 row to make it take effect on the platform firmware).
153
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700154 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
Lv Zheng5dc17982013-07-22 16:08:25 +0800155 acpi_osi="string1" # add string1
156 acpi_osi="!string2" # remove string2
Lv Zheng741d8122013-07-22 16:08:36 +0800157 acpi_osi=!* # remove all strings
Lv Zheng5dc17982013-07-22 16:08:25 +0800158 acpi_osi=! # disable all built-in OS vendor
159 strings
Lv Zhenga707ede2016-05-03 16:48:32 +0800160 acpi_osi=!! # enable all built-in OS vendor
161 strings
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700162 acpi_osi= # disable all strings
163
Lv Zheng5dc17982013-07-22 16:08:25 +0800164 'acpi_osi=!' can be used in combination with single or
165 multiple 'acpi_osi="string1"' to support specific OS
166 vendor string(s). Note that such command can only
167 affect the default state of the OS vendor strings, thus
168 it cannot affect the default state of the feature group
169 strings and the current state of the OS vendor strings,
170 specifying it multiple times through kernel command line
Lv Zheng741d8122013-07-22 16:08:36 +0800171 is meaningless. This command is useful when one do not
172 care about the state of the feature group strings which
173 should be controlled by the OSPM.
Lv Zheng5dc17982013-07-22 16:08:25 +0800174 Examples:
175 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
176 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
177 can make '_OSI("Windows 2000")' TRUE.
178
179 'acpi_osi=' cannot be used in combination with other
180 'acpi_osi=' command lines, the _OSI method will not
181 exist in the ACPI namespace. NOTE that such command can
182 only affect the _OSI support state, thus specifying it
183 multiple times through kernel command line is also
184 meaningless.
185 Examples:
186 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
187 FALSE.
188
Lv Zheng741d8122013-07-22 16:08:36 +0800189 'acpi_osi=!*' can be used in combination with single or
190 multiple 'acpi_osi="string1"' to support specific
191 string(s). Note that such command can affect the
192 current state of both the OS vendor strings and the
193 feature group strings, thus specifying it multiple times
194 through kernel command line is meaningful. But it may
195 still not able to affect the final state of a string if
196 there are quirks related to this string. This command
197 is useful when one want to control the state of the
198 feature group strings to debug BIOS issues related to
199 the OSPM features.
200 Examples:
201 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
202 '_OSI("Module Device")' FALSE.
203 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
204 '_OSI("Module Device")' TRUE.
205 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
206 equivalent to
207 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
208 and
209 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
210 they all will make '_OSI("Windows 2000")' TRUE.
211
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +0530212 acpi_pm_good [X86]
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700213 Override the pmtimer bug detection: force the kernel
214 to assume that this machine's pmtimer latches its value
215 and always returns good values.
216
Randy Dunlap4af94f32009-04-17 18:30:28 -0700217 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
218 Format: { level | edge | high | low }
219
Randy Dunlap4af94f32009-04-17 18:30:28 -0700220 acpi_skip_timer_override [HW,ACPI]
221 Recognize and ignore IRQ0/pin2 Interrupt Override.
222 For broken nForce2 BIOS resulting in XT-PIC timer.
223
224 acpi_sleep= [HW,ACPI] Sleep options
225 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
Amerigo Wangc3b07952011-05-10 21:09:38 +0200226 old_ordering, nonvs, sci_force_enable }
Randy Dunlap4af94f32009-04-17 18:30:28 -0700227 See Documentation/power/video.txt for information on
228 s3_bios and s3_mode.
229 s3_beep is for debugging; it makes the PC's speaker beep
230 as soon as the kernel's real-mode entry point is called.
231 s4_nohwsig prevents ACPI hardware signature from being
232 used during resume from hibernation.
233 old_ordering causes the ACPI 1.0 ordering of the _PTS
234 control method, with respect to putting devices into
235 low power states, to be enforced (the ACPI 2.0 ordering
236 of _PTS is used by default).
Rafael J. Wysocki72ad5d72010-07-23 22:59:09 +0200237 nonvs prevents the kernel from saving/restoring the
238 ACPI NVS memory during suspend/hibernation and resume.
Zhang Ruid7f0eea2009-12-30 15:36:42 +0800239 sci_force_enable causes the kernel to set SCI_EN directly
240 on resume from S1/S3 (which is against the ACPI spec,
241 but some broken systems don't work without it).
Randy Dunlap4af94f32009-04-17 18:30:28 -0700242
243 acpi_use_timer_override [HW,ACPI]
244 Use timer override. For some broken Nvidia NF5 boards
245 that require a timer override, but don't have HPET
246
Randy Dunlap4af94f32009-04-17 18:30:28 -0700247 add_efi_memmap [EFI; X86] Include EFI memory map in
248 kernel's map of available physical RAM.
249
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700250 agp= [AGP]
251 { off | try_unsupported }
252 off: disable AGP support
253 try_unsupported: try to drive unsupported chipsets
254 (may crash computer or cause data corruption)
255
Randy Dunlapbcfde332010-06-07 17:10:38 -0700256 ALSA [HW,ALSA]
257 See Documentation/sound/alsa/alsa-parameters.txt
258
Russell Kingd944d542010-02-20 16:13:29 +0000259 alignment= [KNL,ARM]
260 Allow the default userspace alignment fault handler
261 behaviour to be specified. Bit 0 enables warnings,
262 bit 1 enables fixups, and bit 2 sends a segfault.
263
Borislav Petkovdfb09f92011-08-05 15:15:08 +0200264 align_va_addr= [X86-64]
265 Align virtual addresses by clearing slice [14:12] when
266 allocating a VMA at process creation time. This option
267 gives you up to 3% performance improvement on AMD F15h
268 machines (where it is enabled by default) for a
269 CPU-intensive style benchmark, and it can vary highly in
270 a microbenchmark depending on workload and compiler.
271
Borislav Petkov8360ee22011-11-21 12:10:19 +0100272 32: only for 32-bit processes
273 64: only for 64-bit processes
Borislav Petkovdfb09f92011-08-05 15:15:08 +0200274 on: enable for both 32- and 64-bit processes
275 off: disable for both 32- and 64-bit processes
276
Steven Rostedt (Red Hat)55034cd2013-03-07 22:48:09 -0500277 alloc_snapshot [FTRACE]
278 Allocate the ftrace snapshot buffer on boot up when the
279 main buffer is allocated. This is handy if debugging
280 and you need to use tracing_snapshot() on boot up, and
281 do not want to use tracing_snapshot_alloc() as it needs
282 to be done where GFP_KERNEL allocations are allowed.
283
Sedat Dilek89e0b9a2011-12-05 23:08:32 +0100284 amd_iommu= [HW,X86-64]
Joerg Roedel54b4cbd2008-06-26 21:28:10 +0200285 Pass parameters to the AMD IOMMU driver in the system.
286 Possible values are:
FUJITA Tomonoriafa9fdc2008-09-20 01:23:30 +0900287 fullflush - enable flushing of IO/TLB entries when
288 they are unmapped. Otherwise they are
289 flushed before they will be reused, which
290 is a lot of faster
Joerg Roedela5235722010-05-11 17:12:33 +0200291 off - do not initialize any AMD IOMMU found in
292 the system
Joerg Roedel5abcdba2011-12-01 15:49:45 +0100293 force_isolation - Force device isolation for all
294 devices. The IOMMU driver is not
295 allowed anymore to lift isolation
296 requirements as needed. This option
297 does not override iommu=pt
FUJITA Tomonoriafa9fdc2008-09-20 01:23:30 +0900298
Shuah Khanc099cf12012-05-24 15:58:25 -0600299 amd_iommu_dump= [HW,X86-64]
300 Enable AMD IOMMU driver option to dump the ACPI table
301 for AMD IOMMU. With this option enabled, AMD IOMMU
302 driver will print ACPI tables for AMD IOMMU during
303 IOMMU initialization.
304
Suravee Suthikulpanit3928aa32016-08-23 13:52:32 -0500305 amd_iommu_intr= [HW,X86-64]
306 Specifies one of the following AMD IOMMU interrupt
307 remapping modes:
308 legacy - Use legacy interrupt remapping mode.
309 vapic - Use virtual APIC mode, which allows IOMMU
310 to inject interrupts directly into guest.
311 This mode requires kvm-amd.avic=1.
312 (Default when IOMMU HW support is present.)
313
Linus Torvalds1da177e2005-04-16 15:20:36 -0700314 amijoy.map= [HW,JOY] Amiga joystick support
315 Map of devices attached to JOY0DAT and JOY1DAT
316 Format: <a>,<b>
Paul Bolle395cf962011-08-15 02:02:26 +0200317 See also Documentation/input/joystick.txt
Linus Torvalds1da177e2005-04-16 15:20:36 -0700318
319 analog.map= [HW,JOY] Analog joystick and gamepad support
320 Specifies type or capabilities of an analog joystick
321 connected to one of 16 gameports
322 Format: <type1>,<type2>,..<type16>
323
Randy Dunlapa9913042005-10-23 12:57:11 -0700324 apc= [HW,SPARC]
325 Power management functions (SPARCstation-4/5 + deriv.)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700326 Format: noidle
327 Disable APC CPU standby support. SPARCstation-Fox does
328 not play well with APC CPU idle - disable it if you have
329 APC and your system crashes randomly.
330
Randy Dunlapca1eda22009-05-06 16:02:58 -0700331 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
Randy Dunlapc8facbb2007-07-31 00:37:40 -0700332 Change the output verbosity whilst booting
Linus Torvalds1da177e2005-04-16 15:20:36 -0700333 Format: { quiet (default) | verbose | debug }
334 Change the amount of debugging information output
335 when initialising the APIC and IO-APIC components.
Randy Dunlapa9913042005-10-23 12:57:11 -0700336
Hidehiro Kawaib7c49482015-12-14 11:19:12 +0100337 apic_extnmi= [APIC,X86] External NMI delivery setting
338 Format: { bsp (default) | all | none }
339 bsp: External NMI is delivered only to CPU 0
340 all: External NMIs are broadcast to all CPUs as a
341 backup of CPU 0
342 none: External NMI is masked for all CPUs. This is
343 useful so that a dump capture kernel won't be
344 shot down by NMI
345
Brian Haleyb0f83b22010-02-04 13:36:50 -0800346 autoconf= [IPV6]
347 See Documentation/networking/ipv6.txt.
348
Cyrill Gorcunov9636bc02009-10-14 19:09:04 +0400349 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
350 Limit apic dumping. The parameter defines the maximal
351 number of local apics being dumped. Also it is possible
352 to set it to "all" by meaning -- no limit here.
353 Format: { 1 (default) | 2 | ... | all }.
354 The parameter valid if only apic=debug or
355 apic=verbose is specified.
356 Example: apic=debug show_lapic=all
357
Linus Torvalds1da177e2005-04-16 15:20:36 -0700358 apm= [APM] Advanced Power Management
Pavel Machek71f77052008-07-04 09:59:43 -0700359 See header of arch/x86/kernel/apm_32.c.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700360
Linus Torvalds1da177e2005-04-16 15:20:36 -0700361 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
362 Format: <io>,<irq>,<nodeID>
363
364 ataflop= [HW,M68k]
365
366 atarimouse= [HW,MOUSE] Atari Mouse
367
Linus Torvalds1da177e2005-04-16 15:20:36 -0700368 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
369 EzKey and similar keyboards
370
371 atkbd.reset= [HW] Reset keyboard during initialization
372
Randy Dunlapa9913042005-10-23 12:57:11 -0700373 atkbd.set= [HW] Select keyboard code set
374 Format: <int> (2 = AT (default), 3 = PS/2)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700375
376 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
377 keyboards
378
379 atkbd.softraw= [HW] Choose between synthetic and real raw mode
380 Format: <bool> (0 = real, 1 = synthetic (default))
Randy Dunlapa9913042005-10-23 12:57:11 -0700381
382 atkbd.softrepeat= [HW]
383 Use software keyboard repeat
Linus Torvalds1da177e2005-04-16 15:20:36 -0700384
Richard Guy Briggsa106fb02013-09-23 21:53:35 -0400385 audit= [KNL] Enable the audit sub-system
386 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
Eric Parisd7961142014-01-13 16:01:06 -0500387 0 - kernel audit is disabled and can not be enabled
388 until the next reboot
389 unset - kernel audit is initialized but disabled and
390 will be fully enabled by the userspace auditd.
391 1 - kernel audit is initialized and partially enabled,
392 storing at most audit_backlog_limit messages in
393 RAM until it is fully enabled by the userspace
394 auditd.
Richard Guy Briggsa106fb02013-09-23 21:53:35 -0400395 Default: unset
Richard Guy Briggsf3411cb2013-09-17 12:34:52 -0400396
Richard Guy Briggsf910fde2013-09-17 12:34:52 -0400397 audit_backlog_limit= [KNL] Set the audit queue size limit.
398 Format: <int> (must be >=0)
399 Default: 64
400
Alex Thorlton1c532e02016-03-31 14:18:29 -0500401 bau= [X86_UV] Enable the BAU on SGI UV. The default
402 behavior is to disable the BAU (i.e. bau=0).
403 Format: { "0" | "1" }
404 0 - Disable the BAU.
405 1 - Enable the BAU.
406 unset - Disable the BAU.
407
Linus Torvalds1da177e2005-04-16 15:20:36 -0700408 baycom_epp= [HW,AX25]
409 Format: <io>,<mode>
Randy Dunlapa9913042005-10-23 12:57:11 -0700410
Linus Torvalds1da177e2005-04-16 15:20:36 -0700411 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
412 Format: <io>,<mode>
413 See header of drivers/net/hamradio/baycom_par.c.
414
Randy Dunlapa9913042005-10-23 12:57:11 -0700415 baycom_ser_fdx= [HW,AX25]
416 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700417 Format: <io>,<irq>,<mode>[,<baud>]
418 See header of drivers/net/hamradio/baycom_ser_fdx.c.
419
Randy Dunlapa9913042005-10-23 12:57:11 -0700420 baycom_ser_hdx= [HW,AX25]
421 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700422 Format: <io>,<irq>,<mode>
423 See header of drivers/net/hamradio/baycom_ser_hdx.c.
424
Paul Gortmaker080506a2013-09-30 13:45:19 -0700425 blkdevparts= Manual partition parsing of block device(s) for
426 embedded devices based on command line input.
427 See Documentation/block/cmdline-partition.txt
428
Randy Dunlapbfe8df32007-10-16 01:23:46 -0700429 boot_delay= Milliseconds to delay each printk during boot.
430 Values larger than 10 seconds (10000) are changed to
431 no delay (0).
432 Format: integer
433
Andreas Herrmann35fc9082008-08-20 14:08:58 -0700434 bootmem_debug [KNL] Enable bootmem allocator debug messages.
435
Huang Yinga3e2acc2016-06-29 13:04:29 -0700436 bert_disable [ACPI]
437 Disable BERT OS support on buggy BIOSes.
438
Linus Torvalds1da177e2005-04-16 15:20:36 -0700439 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
Randy Dunlapa9913042005-10-23 12:57:11 -0700440 bttv.radio= Most important insmod options are available as
441 kernel args too.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700442 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
Paul Bolle395cf962011-08-15 02:02:26 +0200443 bttv.tuner=
Linus Torvalds1da177e2005-04-16 15:20:36 -0700444
Will Schmidt4e89a2d2010-09-28 15:33:12 +0000445 bulk_remove=off [PPC] This parameter disables the use of the pSeries
446 firmware feature for flushing multiple hpte entries
447 at a time.
448
Linus Torvalds1da177e2005-04-16 15:20:36 -0700449 c101= [NET] Moxa C101 synchronous serial card
450
Alan Coxcd4f0ef2007-07-31 00:37:59 -0700451 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700452 Sometimes CPU hardware bugs make them report the cache
453 size incorrectly. The kernel will attempt work arounds
454 to fix known problems, but for some CPUs it is not
455 possible to determine what the correct size should be.
456 This option provides an override for these situations.
457
Dmitry Kasatkinffb70f62014-06-17 11:56:58 +0300458 ca_keys= [KEYS] This parameter identifies a specific key(s) on
459 the system trusted keyring to be used for certificate
460 trust validation.
Dmitry Kasatkin32c47412014-06-17 11:56:59 +0300461 format: { id:<keyid> | builtin }
Dmitry Kasatkinffb70f62014-06-17 11:56:58 +0300462
Florian Fainellifd1bb4c2014-06-25 16:41:13 -0700463 cca= [MIPS] Override the kernel pages' cache coherency
464 algorithm. Accepted values range from 0 to 7
465 inclusive. See arch/mips/include/asm/pgtable-bits.h
466 for platform specific values (SB1, Loongson3 and
467 others).
468
Sebastian Ott14ff56b2008-01-26 14:10:37 +0100469 ccw_timeout_log [S390]
470 See Documentation/s390/CommonIO for details.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700471
Paul Menage8bab8dd2008-04-04 14:29:57 -0700472 cgroup_disable= [KNL] Disable a particular controller
473 Format: {name of the controller(s) to disable}
Qiang Huangca0bdbb2013-11-06 13:18:09 -0800474 The effects of cgroup_disable=foo are:
475 - foo isn't auto-mounted if you mount all cgroups in
476 a single hierarchy
477 - foo isn't visible as an individually mountable
478 subsystem
479 {Currently only "memory" controller deal with this and
480 cut the overhead, others just disable the usage. So
481 only cgroup_disable=memory is actually worthy}
Paul Menage8bab8dd2008-04-04 14:29:57 -0700482
Johannes Weiner1619b6d2016-02-16 13:21:14 -0500483 cgroup_no_v1= [KNL] Disable one, multiple, all cgroup controllers in v1
484 Format: { controller[,controller...] | "all" }
485 Like cgroup_disable, but only applies to cgroup v1;
486 the blacklisted controllers remain available in cgroup2.
487
Johannes Weinerf7e1cb62016-01-14 15:21:29 -0800488 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
489 Format: <string>
490 nosocket -- Disable socket memory accounting.
Vladimir Davydov04823c82016-01-20 15:02:38 -0800491 nokmem -- Disable kernel memory accounting.
Johannes Weinerf7e1cb62016-01-14 15:21:29 -0800492
Linus Torvalds1da177e2005-04-16 15:20:36 -0700493 checkreqprot [SELINUX] Set initial checkreqprot flag value.
494 Format: { "0" | "1" }
495 See security/selinux/Kconfig help text.
Randy Dunlapa9913042005-10-23 12:57:11 -0700496 0 -- check protection applied by kernel (includes
497 any implied execute protection).
Linus Torvalds1da177e2005-04-16 15:20:36 -0700498 1 -- check protection requested by application.
499 Default value is set via a kernel config option.
Randy Dunlapa9913042005-10-23 12:57:11 -0700500 Value can be changed at runtime via
501 /selinux/checkreqprot.
502
Sebastian Ott661ca0d2008-01-26 14:10:36 +0100503 cio_ignore= [S390]
504 See Documentation/s390/CommonIO for details.
Olof Johansson1e435252013-04-27 14:10:18 -0700505 clk_ignore_unused
506 [CLK]
Mike Turquettee156ee52014-09-30 14:24:38 -0700507 Prevents the clock framework from automatically gating
508 clocks that have not been explicitly enabled by a Linux
509 device driver but are enabled in hardware at reset or
510 by the bootloader/firmware. Note that this does not
511 force such clocks to be always-on nor does it reserve
512 those clocks in any way. This parameter is useful for
513 debug and development, but should not be needed on a
514 platform with proper driver support. For more
515 information, see Documentation/clk.txt.
Sebastian Ott661ca0d2008-01-26 14:10:36 +0100516
Alan Coxcd4f0ef2007-07-31 00:37:59 -0700517 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
john stultz734efb42006-06-26 00:25:05 -0700518 [Deprecated]
Matt LaPlante3f6dee92006-10-03 22:45:33 +0200519 Forces specified clocksource (if available) to be used
john stultz734efb42006-06-26 00:25:05 -0700520 when calculating gettimeofday(). If specified
Matt LaPlante3f6dee92006-10-03 22:45:33 +0200521 clocksource is not available, it defaults to PIT.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700522 Format: { pit | tsc | cyclone | pmtmr }
523
John Stultz592913e2010-07-13 17:56:20 -0700524 clocksource= Override the default clocksource
Randy Dunlap3d6ac982007-05-23 13:58:16 -0700525 Format: <string>
526 Override the default clocksource and use the clocksource
527 with the name specified.
528 Some clocksource names to choose from, depending on
529 the platform:
530 [all] jiffies (this is the base, fallback clocksource)
531 [ACPI] acpi_pm
532 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
533 pxa_timer,timer3,32k_counter,timer0_1
534 [AVR32] avr32
Alok Kataria9863c902010-08-23 14:49:11 -0700535 [X86-32] pit,hpet,tsc;
Randy Dunlap3d6ac982007-05-23 13:58:16 -0700536 scx200_hrt on Geode; cyclone on IBM x440
537 [MIPS] MIPS
538 [PARISC] cr16
539 [S390] tod
540 [SH] SuperH
541 [SPARC64] tick
542 [X86-64] hpet,tsc
543
Will Deacon46fd5c62016-06-27 17:30:13 +0100544 clocksource.arm_arch_timer.evtstrm=
545 [ARM,ARM64]
546 Format: <bool>
547 Enable/disable the eventstream feature of the ARM
548 architected timer so that code using WFE-based polling
549 loops can be debugged more effectively on production
550 systems.
551
Scott Woodf6dc1572016-09-22 03:35:17 -0500552 clocksource.arm_arch_timer.fsl-a008585=
553 [ARM64]
554 Format: <bool>
555 Enable/disable the workaround of Freescale/NXP
556 erratum A-008585. This can be useful for KVM
557 guests, if the guest device tree doesn't show the
558 erratum. If unspecified, the workaround is
559 enabled based on the device tree.
560
Andi Kleenac72e782008-01-30 13:33:21 +0100561 clearcpuid=BITNUM [X86]
562 Disable CPUID feature X for the kernel. See
Borislav Petkovcd4d09e2016-01-26 22:12:04 +0100563 arch/x86/include/asm/cpufeatures.h for the valid bit
Randy Dunlap07983f02009-01-06 14:42:41 -0800564 numbers. Note the Linux specific bits are not necessarily
Andi Kleenac72e782008-01-30 13:33:21 +0100565 stable over kernel options, but the vendor specific
566 ones should be.
567 Also note that user programs calling CPUID directly
568 or using the feature without checking anything
569 will still see it. This just prevents it from
570 being used by the kernel or shown in /proc/cpuinfo.
571 Also note the kernel might malfunction if you disable
572 some critical bits.
573
Akinobu Mita5ea3b1b2014-06-04 16:06:54 -0700574 cma=nn[MG]@[start[MG][-end[MG]]]
575 [ARM,X86,KNL]
576 Sets the size of kernel global memory area for
577 contiguous memory allocations and optionally the
578 placement constraint by the physical address range of
Jean Delvaref0d6d1f2014-10-09 15:29:41 -0700579 memory allocations. A value of 0 disables CMA
580 altogether. For more information, see
Marek Szyprowskic64be2b2011-12-29 13:09:51 +0100581 include/linux/dma-contiguous.h
582
Robert Jennings14f966e2009-04-15 05:55:32 +0000583 cmo_free_hint= [PPC] Format: { yes | no }
584 Specify whether pages are marked as being inactive
585 when they are freed. This is used in CMO environments
586 to determine OS memory pressure for page stealing by
587 a hypervisor.
588 Default: yes
589
Marek Szyprowskic7909502011-12-29 13:09:51 +0100590 coherent_pool=nn[KMG] [ARM,KNL]
591 Sets the size of memory pool for coherent, atomic dma
Marek Szyprowskie9da6e92012-07-30 09:11:33 +0200592 allocations, by default set to 256K.
Marek Szyprowskic7909502011-12-29 13:09:51 +0100593
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +0530594 code_bytes [X86] How many bytes of object code to print
Arjan van de Vena25bd942008-01-30 13:33:08 +0100595 in an oops report.
Chuck Ebbert86c41832007-02-13 13:26:25 +0100596 Range: 0 - 8192
597 Default: 64
598
Linus Torvalds1da177e2005-04-16 15:20:36 -0700599 com20020= [HW,NET] ARCnet - COM20020 chipset
Randy Dunlapa9913042005-10-23 12:57:11 -0700600 Format:
601 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700602
603 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
604 Format: <io>[,<irq>]
605
Randy Dunlapa9913042005-10-23 12:57:11 -0700606 com90xx= [HW,NET]
607 ARCnet - COM90xx chipset (memory-mapped buffers)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700608 Format: <io>[,<irq>[,<memstart>]]
609
610 condev= [HW,S390] console device
611 conmode=
Randy Dunlapa9913042005-10-23 12:57:11 -0700612
Linus Torvalds1da177e2005-04-16 15:20:36 -0700613 console= [KNL] Output console device and options.
614
615 tty<n> Use the virtual console device <n>.
616
617 ttyS<n>[,options]
Randy Dunlapf1a1c2d2006-03-25 03:08:17 -0800618 ttyUSB0[,options]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700619 Use the specified serial port. The options are of
Randy Dunlapf1a1c2d2006-03-25 03:08:17 -0800620 the form "bbbbpnf", where "bbbb" is the baud rate,
621 "p" is parity ("n", "o", or "e"), "n" is number of
622 bits, and "f" is flow control ("r" for RTS or
623 omit it). Default is "9600n8".
Linus Torvalds1da177e2005-04-16 15:20:36 -0700624
Jani Nikulae52347b2016-11-03 12:10:10 +0200625 See Documentation/admin-guide/serial-console.rst for more
Randy Dunlapf1a1c2d2006-03-25 03:08:17 -0800626 information. See
627 Documentation/networking/netconsole.txt for an
628 alternative.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700629
Yinghai Lu18a8bd92007-07-15 23:37:59 -0700630 uart[8250],io,<addr>[,options]
631 uart[8250],mmio,<addr>[,options]
Masahiro Yamadabd94c402015-10-28 12:46:05 +0900632 uart[8250],mmio16,<addr>[,options]
Peter Hurleyca782f12015-04-06 10:52:39 -0400633 uart[8250],mmio32,<addr>[,options]
634 uart[8250],0x<addr>[,options]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700635 Start an early, polled-mode console on the 8250/16550
636 UART at the specified I/O port or MMIO address,
Peter Hurleyca782f12015-04-06 10:52:39 -0400637 switching to the matching ttyS device later.
638 MMIO inter-register address stride is either 8-bit
Masahiro Yamadabd94c402015-10-28 12:46:05 +0900639 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
640 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
641 to be equivalent to 'mmio'. 'options' are specified in
642 the same format described for ttyS above; if unspecified,
Peter Hurleyca782f12015-04-06 10:52:39 -0400643 the h/w is not re-initialized.
644
Konrad Rzeszutek Wilka2fd6412013-02-25 15:54:09 -0500645 hvc<n> Use the hypervisor console device <n>. This is for
646 both Xen and PowerPC hypervisors.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700647
Samuel Thibaultf7511d52008-04-30 00:54:51 -0700648 If the device connected to the port is not a TTY but a braille
649 device, prepend "brl," before the device type, for instance
650 console=brl,ttyS0
651 For now, only VisioBraille is supported.
652
Daniel Mackf324edc2009-06-16 15:33:52 -0700653 consoleblank= [KNL] The console blank (screen saver) timeout in
654 seconds. Defaults to 10*60 = 10mins. A value of 0
655 disables the blank timer.
656
Hidehiro Kawai4cb0e112009-01-06 14:42:47 -0800657 coredump_filter=
658 [KNL] Change the default value for
659 /proc/<pid>/coredump_filter.
660 See also Documentation/filesystems/proc.txt.
661
Len Brown62027ae2011-04-01 18:13:10 -0400662 cpuidle.off=1 [CPU_IDLE]
663 disable the cpuidle sub-system
664
Len Brownd68921f2015-05-11 17:27:09 -0400665 cpu_init_udelay=N
666 [X86] Delay for N microsec between assert and de-assert
667 of APIC INIT to start processors. This delay occurs
668 on every CPU online, such as boot, and resume from suspend.
669 Default: 10000
670
Linus Torvalds1da177e2005-04-16 15:20:36 -0700671 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
Randy Dunlapa9913042005-10-23 12:57:11 -0700672 Format:
673 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700674
Ahmed S. Darwish6f21e642011-02-20 20:08:35 -0800675 crashkernel=size[KMG][@offset[KMG]]
676 [KNL] Using kexec, Linux can switch to a 'crash kernel'
677 upon panic. This parameter reserves the physical
678 memory region [offset, offset + size] for that kernel
679 image. If '@offset' is omitted, then a suitable offset
680 is selected automatically. Check
681 Documentation/kdump/kdump.txt for further details.
Eric W. Biedermandc009d92005-06-25 14:57:52 -0700682
Bernhard Wallefb391592007-10-18 23:41:02 -0700683 crashkernel=range1:size1[,range2:size2,...][@offset]
684 [KNL] Same as above, but depends on the memory
685 in the running system. The syntax of range is
686 start-[end] where start and end are both
687 a memory unit (amount[KMG]). See also
Ahmed S. Darwish6f21e642011-02-20 20:08:35 -0800688 Documentation/kdump/kdump.txt for an example.
Bernhard Wallefb391592007-10-18 23:41:02 -0700689
Yinghai Luadbc7422013-04-15 22:23:48 -0700690 crashkernel=size[KMG],high
Yinghai Lu55a20ee2013-04-15 22:23:47 -0700691 [KNL, x86_64] range could be above 4G. Allow kernel
692 to allocate physical memory region from top, so could
693 be above 4G if system have more than 4G ram installed.
694 Otherwise memory region will be allocated below 4G, if
695 available.
696 It will be ignored if crashkernel=X is specified.
Yinghai Luadbc7422013-04-15 22:23:48 -0700697 crashkernel=size[KMG],low
698 [KNL, x86_64] range under 4G. When crashkernel=X,high
699 is passed, kernel could allocate physical memory region
Yinghai Luc729de82013-04-15 22:23:45 -0700700 above 4G, that cause second kernel crash on system
701 that require some amount of low memory, e.g. swiotlb
Baoquan Hec6045032015-09-24 16:51:25 +0800702 requires at least 64M+32K low memory, also enough extra
703 low memory is needed to make sure DMA buffers for 32-bit
704 devices won't run out. Kernel would try to allocate at
705 at least 256M below 4G automatically.
Yinghai Luc729de82013-04-15 22:23:45 -0700706 This one let user to specify own low range under 4G
707 for second kernel instead.
708 0: to disable low allocation.
Yinghai Luadbc7422013-04-15 22:23:48 -0700709 It will be ignored when crashkernel=X,high is not used
Yinghai Lu55a20ee2013-04-15 22:23:47 -0700710 or memory reserved is below 4G.
Yinghai Luc729de82013-04-15 22:23:45 -0700711
Richard W.M. Jones9e5c9fe2016-05-03 10:00:17 +0100712 cryptomgr.notests
713 [KNL] Disable crypto self-tests
714
Linus Torvalds1da177e2005-04-16 15:20:36 -0700715 cs89x0_dma= [HW,NET]
716 Format: <dma>
717
718 cs89x0_media= [HW,NET]
719 Format: { rj45 | aui | bnc }
Randy Dunlapa9913042005-10-23 12:57:11 -0700720
Randy Dunlapa9913042005-10-23 12:57:11 -0700721 dasd= [HW,NET]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700722 See header of drivers/s390/block/dasd_devmap.c.
723
724 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
725 (one device per port)
726 Format: <port#>,<type>
727 See also Documentation/input/joystick-parport.txt
728
Thomas Renningera648ec02010-08-06 16:11:02 +0200729 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
730 time. See Documentation/dynamic-debug-howto.txt for
Jim Cromie29e36c92012-04-27 14:30:41 -0600731 details. Deprecated, see dyndbg.
Thomas Renningera648ec02010-08-06 16:11:02 +0200732
Linus Torvalds1da177e2005-04-16 15:20:36 -0700733 debug [KNL] Enable kernel debugging (events log level).
734
Ingo Molnarcae2ed92006-07-03 00:24:48 -0700735 debug_locks_verbose=
736 [KNL] verbose self-tests
737 Format=<0|1>
738 Print debugging info while doing the locking API
739 self-tests.
740 We default to 0 (no extra messages), setting it to
741 1 will print _a lot_ more information - normally
742 only useful to kernel developers.
743
Thomas Gleixner3ac7fe52008-04-30 00:55:01 -0700744 debug_objects [KNL] Enable object debugging
745
Kyle McMartin3e8ebb52009-03-01 20:41:41 -0500746 no_debug_objects
747 [KNL] Disable object debugging
748
Stanislaw Gruszkac0a32fc2012-01-10 15:07:28 -0800749 debug_guardpage_minorder=
750 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
751 parameter allows control of the order of pages that will
752 be intentionally kept free (and hence protected) by the
753 buddy allocator. Bigger value increase the probability
754 of catching random memory corruption, but reduce the
755 amount of memory for normal system use. The maximum
756 possible value is MAX_ORDER/2. Setting this parameter
757 to 1 or 2 should be enough to identify most random
758 memory corruption problems caused by bugs in kernel or
759 driver code when a CPU writes to (or reads from) a
760 random memory location. Note that there exists a class
761 of memory corruptions problems caused by buggy H/W or
762 F/W or by drivers badly programing DMA (basically when
763 memory is written at bus level and the CPU MMU is
764 bypassed) which are not detectable by
765 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
766 tracking down these problems.
767
Joonsoo Kim031bc572014-12-12 16:55:52 -0800768 debug_pagealloc=
769 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
770 parameter enables the feature at boot time. In
771 default, it is disabled. We can avoid allocating huge
772 chunk of memory for debug pagealloc if we don't enable
773 it at boot time and the system will work mostly same
774 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
775 on: enable the feature
776
Thomas Gleixnerd3af01f2008-07-15 15:04:56 +0200777 debugpat [X86] Enable PAT debugging
778
Robert P. J. Day2d27a962008-02-03 15:18:45 +0200779 decnet.addr= [HW,NET]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700780 Format: <area>[,<node>]
781 See also Documentation/networking/decnet.txt.
782
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700783 default_hugepagesz=
784 [same as hugepagesz=] The size of the default
785 HugeTLB page size. This is the size represented by
786 the legacy /proc/ hugepages APIs, used for SHM, and
787 default size when mounting hugetlbfs filesystems.
788 Defaults to the default architecture's huge page size
789 if not specified.
Antonino A. Daplas55ff9782007-05-08 00:38:53 -0700790
Linus Torvalds1da177e2005-04-16 15:20:36 -0700791 dhash_entries= [KNL]
792 Set number of hash buckets for dentry cache.
Randy Dunlapa9913042005-10-23 12:57:11 -0700793
Oliver O'Halloranfaf78822016-07-05 11:43:21 +1000794 disable_1tb_segments [PPC]
795 Disables the use of 1TB hash page table segments. This
796 causes the kernel to fall back to 256MB segments which
797 can be useful when debugging issues that require an SLB
798 miss to occur.
799
Brian Haleyb0f83b22010-02-04 13:36:50 -0800800 disable= [IPV6]
801 See Documentation/networking/ipv6.txt.
802
Aneesh Kumar K.Vb275bfb2016-07-13 15:05:31 +0530803 disable_radix [PPC]
804 Disable RADIX MMU mode on POWER9
805
HATAYAMA Daisuke151e0c72014-01-15 15:44:58 +0900806 disable_cpu_apicid= [X86,APIC,SMP]
807 Format: <int>
808 The number of initial APIC ID for the
809 corresponding CPU to be disabled at boot,
810 mostly used for the kdump 2nd kernel to
811 disable BSP to wake up multiple CPUs without
812 causing system reset or hang due to sending
813 INIT from AP to BSP.
814
Nishanth Aravamudan4e8b0cf2011-02-10 09:10:47 +0000815 disable_ddw [PPC/PSERIES]
816 Disable Dynamic DMA Window support. Use this if
817 to workaround buggy firmware.
818
Brian Haleyb0f83b22010-02-04 13:36:50 -0800819 disable_ipv6= [IPV6]
820 See Documentation/networking/ipv6.txt.
821
Yinghai Lu95ffa242008-04-29 03:52:33 -0700822 disable_mtrr_cleanup [X86]
Yinghai Lu95ffa242008-04-29 03:52:33 -0700823 The kernel tries to adjust MTRR layout from continuous
824 to discrete, to make X server driver able to add WB
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700825 entry later. This parameter disables that.
Yinghai Lu95ffa242008-04-29 03:52:33 -0700826
Yinghai Lu093af8d2008-01-30 13:33:32 +0100827 disable_mtrr_trim [X86, Intel and AMD only]
Jesse Barnes99fc8d42008-01-30 13:33:18 +0100828 By default the kernel will trim any uncacheable
829 memory out of your available memory pool based on
830 MTRR settings. This parameter disables that behavior,
831 possibly causing your machine to run very slowly.
832
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +0530833 disable_timer_pin_1 [X86]
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700834 Disable PIN 1 of APIC timer
835 Can be useful to work around chipset bugs.
836
Prarit Bhargavace14c582015-08-25 13:34:53 -0400837 dis_ucode_ldr [X86] Disable the microcode loader.
838
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700839 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
840 this option disables the debugging code at boot.
841
842 dma_debug_entries=<number>
843 This option allows to tune the number of preallocated
844 entries for DMA-API debugging code. One entry is
845 required per DMA-API allocation. Use this if the
846 DMA-API debugging code disables itself because the
847 architectural default is too low.
848
Joerg Roedel1745de52009-05-22 21:49:51 +0200849 dma_debug_driver=<driver_name>
850 With this option the DMA-API debugging driver
851 filter feature can be enabled at boot time. Just
852 pass the driver to filter for as the parameter.
853 The filter can be disabled or changed to another
854 driver later using sysfs.
855
Bob Paauwe96206e22015-08-27 10:04:13 -0700856 drm_kms_helper.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
857 Broken monitors, graphic adapters, KVMs and EDIDless
858 panels may send no or incorrect EDID data sets.
859 This parameter allows to specify an EDID data sets
860 in the /lib/firmware directory that are used instead.
Carsten Emdeda0df922012-03-18 22:37:33 +0100861 Generic built-in EDID data sets are used, if one of
862 edid/1024x768.bin, edid/1280x1024.bin,
863 edid/1680x1050.bin, or edid/1920x1080.bin is given
864 and no file with the same name exists. Details and
865 instructions how to build your own EDID data are
866 available in Documentation/EDID/HOWTO.txt. An EDID
867 data set will only be used for a particular connector,
868 if its name and a colon are prepended to the EDID
Bob Paauwe96206e22015-08-27 10:04:13 -0700869 name. Each connector may use a unique EDID data
870 set by separating the files with a comma. An EDID
871 data set with no connector name will be used for
872 any connectors not explicitly specified.
Carsten Emdeda0df922012-03-18 22:37:33 +0100873
Linus Torvalds1da177e2005-04-16 15:20:36 -0700874 dscc4.setup= [NET]
875
Lukas Wunner58c54752016-11-12 21:32:36 +0000876 dump_apple_properties [X86]
877 Dump name and content of EFI device properties on
878 x86 Macs. Useful for driver authors to determine
879 what data is available or for reverse-engineering.
880
Jim Cromie29e36c92012-04-27 14:30:41 -0600881 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
882 module.dyndbg[="val"]
883 Enable debug messages at boot time. See
884 Documentation/dynamic-debug-howto.txt for details.
885
Dave Hansen8c3641e2015-06-07 11:37:02 -0700886 nompx [X86] Disables Intel Memory Protection Extensions.
887 See Documentation/x86/intel_mpx.txt for more
888 information about the feature.
889
Dave Hansen06976942016-02-12 13:02:29 -0800890 nopku [X86] Disable Memory Protection Keys CPU feature found
891 in some Intel CPUs.
892
Luis R. Rodriguezf2411da2015-03-30 16:20:05 -0700893 module.async_probe [KNL]
894 Enable asynchronous probe on this module.
895
Mark Salter56aeeba2014-04-07 15:39:53 -0700896 early_ioremap_debug [KNL]
897 Enable debug messages in early_ioremap support. This
898 is useful for tracking down temporary early mappings
899 which are not unmapped.
900
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700901 earlycon= [KNL] Output early console device and options.
Rob Herring0d3c6732014-04-18 17:19:57 -0500902
Scott Wood5664f762015-09-14 19:54:07 -0500903 When used with no options, the early console is
904 determined by the stdout-path property in device
905 tree's chosen node.
906
Scott Telfordc41251b2016-09-22 16:58:16 +0100907 cdns,<addr>[,options]
908 Start an early, polled-mode console on a Cadence
909 (xuartps) serial port at the specified address. Only
910 supported option is baud rate. If baud rate is not
911 specified, the serial port must already be setup and
912 configured.
Michal Simek6fa62fc2014-09-10 12:43:02 +0200913
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700914 uart[8250],io,<addr>[,options]
915 uart[8250],mmio,<addr>[,options]
Samium Gromoff1917ac72010-07-20 15:26:51 -0700916 uart[8250],mmio32,<addr>[,options]
Noam Camus6e63be32015-05-25 06:54:28 +0300917 uart[8250],mmio32be,<addr>[,options]
Peter Hurleyca782f12015-04-06 10:52:39 -0400918 uart[8250],0x<addr>[,options]
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700919 Start an early, polled-mode console on the 8250/16550
920 UART at the specified I/O port or MMIO address.
Randy Dunlap16290242011-08-13 12:34:52 -0700921 MMIO inter-register address stride is either 8-bit
Noam Camus6e63be32015-05-25 06:54:28 +0300922 (mmio) or 32-bit (mmio32 or mmio32be).
923 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
924 to be equivalent to 'mmio'. 'options' are specified
925 in the same format described for "console=ttyS<n>"; if
Peter Hurleyca782f12015-04-06 10:52:39 -0400926 unspecified, the h/w is not initialized.
Randy Dunlap0cb55ad2009-04-05 15:55:22 -0700927
Rob Herring0d3c6732014-04-18 17:19:57 -0500928 pl011,<addr>
Timur Tabi3b78fae2016-01-04 15:37:42 -0600929 pl011,mmio32,<addr>
Rob Herring0d3c6732014-04-18 17:19:57 -0500930 Start an early, polled-mode console on a pl011 serial
931 port at the specified address. The pl011 serial port
932 must already be setup and configured. Options are not
Timur Tabi3b78fae2016-01-04 15:37:42 -0600933 yet supported. If 'mmio32' is specified, then only
934 the driver will use only 32-bit accessors to read/write
935 the device registers.
Rob Herring0d3c6732014-04-18 17:19:57 -0500936
Andreas Färber736d5532016-03-06 12:21:24 +0100937 meson,<addr>
938 Start an early, polled-mode console on a meson serial
939 port at the specified address. The serial port must
940 already be setup and configured. Options are not yet
941 supported.
942
Stephen Boyd0efe7292014-09-15 17:22:51 -0700943 msm_serial,<addr>
944 Start an early, polled-mode console on an msm serial
945 port at the specified address. The serial port
946 must already be setup and configured. Options are not
947 yet supported.
948
949 msm_serial_dm,<addr>
950 Start an early, polled-mode console on an msm serial
951 dm port at the specified address. The serial port
952 must already be setup and configured. Options are not
953 yet supported.
954
Rob Herringd50d7262014-04-18 17:19:58 -0500955 smh Use ARM semihosting calls for early console.
956
Tomasz Figab94ba032015-01-23 14:47:41 +0100957 s3c2410,<addr>
958 s3c2412,<addr>
959 s3c2440,<addr>
960 s3c6400,<addr>
961 s5pv210,<addr>
962 exynos4210,<addr>
963 Use early console provided by serial driver available
964 on Samsung SoCs, requires selecting proper type and
965 a correct base address of the selected UART port. The
966 serial port must already be setup and configured.
967 Options are not yet supported.
968
Stefan Agner1d59b382015-10-17 00:45:55 -0700969 lpuart,<addr>
970 lpuart32,<addr>
971 Use early console provided by Freescale LP UART driver
972 found on Freescale Vybrid and QorIQ LS1021A processors.
973 A valid base address must be provided, and the serial
974 port must already be setup and configured.
975
Wilson Ding30530792016-02-16 19:14:53 +0100976 armada3700_uart,<addr>
977 Start an early, polled-mode console on the
978 Armada 3700 serial port at the specified
979 address. The serial port must already be setup
980 and configured. Options are not yet supported.
981
Heiko Carstens89175cf2017-01-11 09:14:52 +0100982 earlyprintk= [X86,SH,BLACKFIN,ARM,M68k,S390]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700983 earlyprintk=vga
Matt Fleming72548e82013-10-04 09:36:56 +0100984 earlyprintk=efi
Heiko Carstens89175cf2017-01-11 09:14:52 +0100985 earlyprintk=sclp
Konrad Rzeszutek Wilk2482a922013-02-25 15:54:08 -0500986 earlyprintk=xen
Linus Torvalds1da177e2005-04-16 15:20:36 -0700987 earlyprintk=serial[,ttySn[,baudrate]]
Dave Hansen147ea092013-04-10 14:03:38 -0700988 earlyprintk=serial[,0x...[,baudrate]]
Jason Wesselea3acb12009-09-24 09:08:30 -0500989 earlyprintk=ttySn[,baudrate]
Jason Wessel9780bc42009-08-20 15:39:57 -0500990 earlyprintk=dbgp[debugController#]
Alexander Kuleshovc43088e2015-06-10 23:21:15 +0600991 earlyprintk=pciserial,bus:device.function[,baudrate]
Linus Torvalds1da177e2005-04-16 15:20:36 -0700992
Dave Hansen147ea092013-04-10 14:03:38 -0700993 earlyprintk is useful when the kernel crashes before
994 the normal console is initialized. It is not enabled by
995 default because it has some cosmetic problems.
996
Randy Dunlapa9913042005-10-23 12:57:11 -0700997 Append ",keep" to not disable it when the real console
Linus Torvalds1da177e2005-04-16 15:20:36 -0700998 takes over.
999
Matt Fleming72548e82013-10-04 09:36:56 +01001000 Only one of vga, efi, serial, or usb debug port can
1001 be used at a time.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001002
Dave Hansen147ea092013-04-10 14:03:38 -07001003 Currently only ttyS0 and ttyS1 may be specified by
1004 name. Other I/O ports may be explicitly specified
1005 on some architectures (x86 and arm at least) by
1006 replacing ttySn with an I/O port address, like this:
1007 earlyprintk=serial,0x1008,115200
1008 You can find the port for a given device in
1009 /proc/tty/driver/serial:
1010 2: uart:ST16650V2 port:00001008 irq:18 ...
Linus Torvalds1da177e2005-04-16 15:20:36 -07001011
1012 Interaction with the standard serial driver is not
1013 very good.
1014
Matt Fleming72548e82013-10-04 09:36:56 +01001015 The VGA and EFI output is eventually overwritten by
1016 the real console.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001017
Konrad Rzeszutek Wilk2482a922013-02-25 15:54:08 -05001018 The xen output can only be used by Xen PV guests.
1019
Heiko Carstens89175cf2017-01-11 09:14:52 +01001020 The sclp output can only be used on s390.
1021
Chen, Gongc700f012013-12-06 01:17:08 -05001022 edac_report= [HW,EDAC] Control how to report EDAC event
1023 Format: {"on" | "off" | "force"}
1024 on: enable EDAC to report H/W event. May be overridden
1025 by other higher priority error reporting module.
1026 off: disable H/W event reporting through EDAC.
1027 force: enforce the use of EDAC to report H/W event.
1028 default: on.
1029
Jason Wessel97311912010-05-20 21:04:30 -05001030 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1031 ekgdboc=kbd
1032
Lucas De Marchi25985ed2011-03-30 22:57:33 -03001033 This is designed to be used in conjunction with
Jason Wessel97311912010-05-20 21:04:30 -05001034 the boot argument: earlyprintk=vga
1035
Linus Torvalds1da177e2005-04-16 15:20:36 -07001036 edd= [EDD]
Tim Gardner8c4dd602008-04-29 01:02:45 -07001037 Format: {"off" | "on" | "skip[mbr]"}
Linus Torvalds1da177e2005-04-16 15:20:36 -07001038
Borislav Petkovd2f7cbe2013-10-31 17:25:08 +01001039 efi= [EFI]
Borislav Petkovfed6cef2015-02-05 11:44:41 +01001040 Format: { "old_map", "nochunk", "noruntime", "debug" }
Borislav Petkovd2f7cbe2013-10-31 17:25:08 +01001041 old_map [X86-64]: switch to the old ioremap-based EFI
1042 runtime services mapping. 32-bit still uses this one by
1043 default.
Matt Fleming5a17dae2014-08-05 11:52:11 +01001044 nochunk: disable reading files in "chunks" in the EFI
1045 boot stub, as chunking can cause problems with some
1046 firmware implementations.
Dave Young5ae36832014-08-14 17:15:28 +08001047 noruntime : disable EFI runtime services support
Borislav Petkovfed6cef2015-02-05 11:44:41 +01001048 debug: enable misc debug output
Borislav Petkovd2f7cbe2013-10-31 17:25:08 +01001049
Richard Weinberger8c58bf32013-04-17 01:00:53 +02001050 efi_no_storage_paranoia [EFI; X86]
1051 Using this parameter you can use more than 50% of
1052 your efi variable storage. Use this parameter only if
1053 you are really sure that your UEFI does sane gc and
1054 fulfills the spec otherwise your board may brick.
1055
Taku Izumi0f96a992015-09-30 23:01:56 +09001056 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1057 Add arbitrary attribute to specific memory range by
1058 updating original EFI memory map.
1059 Region of memory which aa attribute is added to is
1060 from ss to ss+nn.
1061 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1062 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1063 attribute is added to range 0x100000000-0x180000000 and
1064 0x10a0000000-0x1120000000.
1065
1066 Using this parameter you can do debugging of EFI memmap
1067 related feature. For example, you can do debugging of
1068 Address Range Mirroring feature even if your box
1069 doesn't support it.
1070
Octavian Purdila475fb4e2016-07-08 19:13:12 +03001071 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1072 that is to be dynamically loaded by Linux. If there are
1073 multiple variables with the same name but with different
1074 vendor GUIDs, all of them will be loaded. See
1075 Documentation/acpi/ssdt-overlays.txt for details.
1076
1077
Linus Torvalds1da177e2005-04-16 15:20:36 -07001078 eisa_irq_edge= [PARISC,HW]
1079 See header of drivers/parisc/eisa.c.
1080
Alan Coxcd4f0ef2007-07-31 00:37:59 -07001081 elanfreq= [X86-32]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001082 See comment before function elanfreq_setup() in
Pavel Machek71f77052008-07-04 09:59:43 -07001083 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001084
1085 elevator= [IOSCHED]
Randy Dunlap17a9e7b2010-11-11 12:09:59 +01001086 Format: {"cfq" | "deadline" | "noop"}
Paul Bolle395cf962011-08-15 02:02:26 +02001087 See Documentation/block/cfq-iosched.txt and
Randy Dunlapa9913042005-10-23 12:57:11 -07001088 Documentation/block/deadline-iosched.txt for details.
1089
Michael Holzheud3bf37952011-10-30 15:16:37 +01001090 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
Randy Dunlapa9913042005-10-23 12:57:11 -07001091 Specifies physical address of start of kernel core
Michael Holzheud3bf37952011-10-30 15:16:37 +01001092 image elf header and optionally the size. Generally
1093 kexec loader will pass this option to capture kernel.
Vivek Goyalaac04b32006-01-09 20:51:47 -08001094 See Documentation/kdump/kdump.txt for details.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001095
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001096 enable_mtrr_cleanup [X86]
1097 The kernel tries to adjust MTRR layout from continuous
1098 to discrete, to make X server driver able to add WB
1099 entry later. This parameter enables that.
1100
Randy Dunlapca1eda22009-05-06 16:02:58 -07001101 enable_timer_pin_1 [X86]
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001102 Enable PIN 1 of APIC timer
1103 Can be useful to work around chipset bugs
1104 (in particular on some ATI chipsets).
1105 The kernel tries to set a reasonable default.
1106
Linus Torvalds1da177e2005-04-16 15:20:36 -07001107 enforcing [SELINUX] Set initial enforcing status.
1108 Format: {"0" | "1"}
1109 See security/selinux/Kconfig help text.
1110 0 -- permissive (log only, no denials).
1111 1 -- enforcing (deny and log).
1112 Default value is 0.
1113 Value can be changed at runtime via /selinux/enforce.
1114
Huang Yinga08f82d2010-05-18 14:35:21 +08001115 erst_disable [ACPI]
1116 Disable Error Record Serialization Table (ERST)
1117 support.
1118
Linus Torvalds1da177e2005-04-16 15:20:36 -07001119 ether= [HW,NET] Ethernet cards parameters
1120 This option is obsoleted by the "netdev=" option, which
1121 has equivalent usage. See its documentation for details.
1122
Mimi Zohar7102ebc2011-05-12 18:33:20 -04001123 evm= [EVM]
1124 Format: { "fix" }
1125 Permit 'security.evm' to be updated regardless of
1126 current integrity status.
1127
Akinobu Mitade1ba092006-12-08 02:39:42 -08001128 failslab=
1129 fail_page_alloc=
1130 fail_make_request=[KNL]
1131 General fault injection mechanism.
1132 Format: <interval>,<probability>,<space>,<times>
Paul Bolle395cf962011-08-15 02:02:26 +02001133 See also Documentation/fault-injection/.
Akinobu Mitade1ba092006-12-08 02:39:42 -08001134
Linus Torvalds1da177e2005-04-16 15:20:36 -07001135 floppy= [HW]
Randy Dunlap31c00fc2008-11-13 21:33:24 +00001136 See Documentation/blockdev/floppy.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001137
Alex Chiangf13ae302008-05-08 14:03:23 -06001138 force_pal_cache_flush
1139 [IA-64] Avoid check_sal_cache_flush which may hang on
1140 buggy SAL_CACHE_FLUSH implementations. Using this
1141 parameter will force ia64_sal_cache_flush to call
1142 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1143
Chris Bainbridge69f23662014-03-07 18:40:42 +07001144 forcepae [X86-32]
1145 Forcefully enable Physical Address Extension (PAE).
1146 Many Pentium M systems disable PAE but may have a
1147 functionally usable PAE implementation.
1148 Warning: use of this parameter will taint the kernel
1149 and may cause unknown problems.
1150
Peter Zijlstrad9e54072008-11-01 19:57:37 +01001151 ftrace=[tracer]
Steven Rostedt2af15d62009-05-28 13:37:24 -04001152 [FTRACE] will set and start the specified tracer
Peter Zijlstrad9e54072008-11-01 19:57:37 +01001153 as early as possible in order to facilitate early
1154 boot debugging.
1155
Frederic Weisbeckercecbca92010-04-18 19:08:41 +02001156 ftrace_dump_on_oops[=orig_cpu]
Steven Rostedt2af15d62009-05-28 13:37:24 -04001157 [FTRACE] will dump the trace buffers on oops.
Frederic Weisbeckercecbca92010-04-18 19:08:41 +02001158 If no parameter is passed, ftrace will dump
1159 buffers of all CPUs, but if you pass orig_cpu, it will
1160 dump only the buffer of the CPU that triggered the
1161 oops.
Steven Rostedt2af15d62009-05-28 13:37:24 -04001162
1163 ftrace_filter=[function-list]
1164 [FTRACE] Limit the functions traced by the function
1165 tracer at boot up. function-list is a comma separated
1166 list of functions. This list can be changed at run
1167 time by the set_ftrace_filter file in the debugfs
Randy Dunlap16290242011-08-13 12:34:52 -07001168 tracing directory.
Steven Rostedt2af15d62009-05-28 13:37:24 -04001169
1170 ftrace_notrace=[function-list]
1171 [FTRACE] Do not trace the functions specified in
1172 function-list. This list can be changed at run time
1173 by the set_ftrace_notrace file in the debugfs
1174 tracing directory.
Peter Zijlstrad9e54072008-11-01 19:57:37 +01001175
Stefan Assmann369bc182009-10-12 22:17:21 +02001176 ftrace_graph_filter=[function-list]
1177 [FTRACE] Limit the top level callers functions traced
1178 by the function graph tracer at boot up.
1179 function-list is a comma separated list of functions
1180 that can be changed at run time by the
1181 set_graph_function file in the debugfs tracing directory.
1182
Namhyung Kim0d7d9a12014-06-13 01:23:50 +09001183 ftrace_graph_notrace=[function-list]
1184 [FTRACE] Do not trace from the functions specified in
1185 function-list. This list is a comma separated list of
1186 functions that can be changed at run time by the
1187 set_graph_notrace file in the debugfs tracing directory.
1188
Linus Torvalds1da177e2005-04-16 15:20:36 -07001189 gamecon.map[2|3]=
1190 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1191 support via parallel port (up to 5 devices per port)
1192 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1193 See also Documentation/input/joystick-parport.txt
1194
1195 gamma= [HW,DRM]
1196
Yinghai Luaaf23042008-01-30 13:33:09 +01001197 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1198 Format: off | on
1199 default: on
1200
Peter Oberparleiter2521f2c2009-06-17 16:28:08 -07001201 gcov_persist= [GCOV] When non-zero (default), profiling data for
1202 kernel modules is saved and remains accessible via
1203 debugfs, even when the module is unloaded/reloaded.
1204 When zero, profiling data is discarded and associated
1205 debugfs files are removed at module unload time.
1206
Linus Torvalds1da177e2005-04-16 15:20:36 -07001207 gpt [EFI] Forces disk with valid GPT signature but
Davidlohr Bueso6c5de792014-01-23 15:56:03 -08001208 invalid Protective MBR to be treated as GPT. If the
1209 primary GPT is corrupted, it enables the backup/alternate
1210 GPT to be used instead.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001211
Andreas Larsson6cec9b02012-11-15 08:47:14 +01001212 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1213 the "Enable 0" bit of the configuration register.
1214 Format: 0 | 1
1215 Default: 0
1216 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1217 the "Enable 0" bit of the configuration register.
1218 Format: 0 | 1
1219 Default: 0
1220 grcan.select= [HW] Select which physical interface to use.
1221 Format: 0 | 1
1222 Default: 0
1223 grcan.txsize= [HW] Sets the size of the tx buffer.
1224 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1225 Default: 1024
1226 grcan.rxsize= [HW] Sets the size of the rx buffer.
1227 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1228 Default: 1024
1229
Bamvor Jian Zhang0f98dd12016-08-31 11:45:46 +02001230 gpio-mockup.gpio_mockup_ranges
1231 [HW] Sets the ranges of gpiochip of for this device.
1232 Format: <start1>,<end1>,<start2>,<end2>...
1233
Jiri Kosina55537872015-11-05 18:44:41 -08001234 hardlockup_all_cpu_backtrace=
1235 [KNL] Should the hard-lockup detector generate
1236 backtraces on all cpus.
1237 Format: <integer>
1238
Linus Torvalds1da177e2005-04-16 15:20:36 -07001239 hashdist= [KNL,NUMA] Large hashes allocated during boot
1240 are distributed across NUMA nodes. Defaults on
Randy Dunlap16290242011-08-13 12:34:52 -07001241 for 64-bit NUMA, off otherwise.
Randy Dunlapa9913042005-10-23 12:57:11 -07001242 Format: 0 | 1 (for off | on)
Linus Torvalds1da177e2005-04-16 15:20:36 -07001243
1244 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1245
1246 hd= [EIDE] (E)IDE hard drive subsystem geometry
1247 Format: <cyl>,<head>,<sect>
1248
Huang Yingea8c0712010-05-18 14:35:15 +08001249 hest_disable [ACPI]
1250 Disable Hardware Error Source Table (HEST) support;
1251 corresponding firmware-first mode error processing
1252 logic will be disabled.
1253
Linus Torvalds1da177e2005-04-16 15:20:36 -07001254 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1255 size of <nn>. This works even on boxes that have no
1256 highmem otherwise. This also works to reduce highmem
1257 size on bigger boxes.
1258
Thomas Gleixner54cdfdb2007-02-16 01:28:11 -08001259 highres= [KNL] Enable/disable high resolution timer mode.
1260 Valid parameters: "on", "off"
1261 Default: "on"
1262
Linus Torvalds1da177e2005-04-16 15:20:36 -07001263 hisax= [HW,ISDN]
1264 See Documentation/isdn/README.HiSax.
1265
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001266 hlt [BUGS=ARM,SH]
1267
1268 hpet= [X86-32,HPET] option to control HPET usage
1269 Format: { enable (default) | disable | force |
1270 verbose }
1271 disable: disable HPET and use PIT instead
1272 force: allow force enabled of undocumented chips (ICH4,
1273 VIA, nVidia)
1274 verbose: show contents of HPET registers during setup
1275
Prarit Bhargava3d035f52013-11-12 15:08:33 -08001276 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1277 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1278
Andi Kleenb4718e62008-07-23 21:27:51 -07001279 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1280 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
Jon Tollefson0d9ea752008-07-23 21:27:56 -07001281 On x86-64 and powerpc, this option can be specified
1282 multiple times interleaved with hugepages= to reserve
1283 huge pages of different sizes. Valid pages sizes on
1284 x86-64 are 2M (when the CPU supports "pse") and 1G
Luiz Capitulino27ec26e2014-12-12 16:55:18 -08001285 (when the CPU supports the "pdpe1gb" cpuinfo flag).
Paul Mundt6902aa82008-09-21 17:14:42 +09001286
Hendrik Brueckner555d61d2009-01-09 12:15:02 +01001287 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1288 terminal devices. Valid values: 0..8
Hendrik Brueckner431429f2009-03-26 15:23:55 +01001289 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1290 If specified, z/VM IUCV HVC accepts connections
1291 from listed z/VM user IDs only.
Hendrik Bruecknercef71252008-12-25 13:39:55 +01001292
James Hoganfdabf522012-10-09 11:00:27 +01001293 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1294 hardware thread id mappings.
1295 Format: <cpu>:<hwthread>
1296
Fabio M. Di Nitto7bf69392011-03-22 16:34:20 -07001297 keep_bootcon [KNL]
1298 Do not unregister boot console at start. This is only
1299 useful for debugging when something happens in the window
1300 between unregistering the boot console and initializing
1301 the real console.
1302
Jarkko Nikula3a853fb2009-03-23 18:07:47 -07001303 i2c_bus= [HW] Override the default board specific I2C bus speed
Jarkko Nikula79547632009-03-23 18:07:48 -07001304 or register an additional I2C bus that is not
1305 registered from board initialization code.
Jarkko Nikula3a853fb2009-03-23 18:07:47 -07001306 Format:
1307 <bus_id>,<clkrate>
1308
Jiri Kosina36d95732008-10-06 02:51:09 -04001309 i8042.debug [HW] Toggle i8042 debug mode
Stephen Chandler Paule1443d22015-07-15 10:20:17 -07001310 i8042.unmask_kbd_data
1311 [HW] Enable printing of interrupt data from the KBD port
1312 (disabled by default, and as a pre-condition
1313 requires that i8042.debug=1 be enabled)
Linus Torvalds1da177e2005-04-16 15:20:36 -07001314 i8042.direct [HW] Put keyboard port into non-translated mode
Matt LaPlante84eb8d02006-10-03 22:53:09 +02001315 i8042.dumbkbd [HW] Pretend that controller can only read data from
1316 keyboard and cannot control its state
Linus Torvalds1da177e2005-04-16 15:20:36 -07001317 (Don't attempt to blink the leds)
1318 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
Dmitry Torokhov945ef0d2005-09-04 01:42:00 -05001319 i8042.nokbd [HW] Don't check/create keyboard port
Jiri Kosina75d08c72008-03-13 16:13:59 -04001320 i8042.noloop [HW] Disable the AUX Loopback command while probing
1321 for the AUX port
Linus Torvalds1da177e2005-04-16 15:20:36 -07001322 i8042.nomux [HW] Don't check presence of an active multiplexing
Dmitry Torokhove55a3362014-10-31 09:35:53 -07001323 controller
Linus Torvalds1da177e2005-04-16 15:20:36 -07001324 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1325 controllers
Masanari Iida24775d62012-02-15 00:26:42 +09001326 i8042.notimeout [HW] Ignore timeout condition signalled by controller
Marcos Paulo de Souza930e19242016-10-01 12:07:35 -07001327 i8042.reset [HW] Reset the controller during init, cleanup and
1328 suspend-to-ram transitions, only during s2r
1329 transitions, or never reset
1330 Format: { 1 | Y | y | 0 | N | n }
1331 1, Y, y: always reset controller
1332 0, N, n: don't ever reset controller
1333 Default: only on s2r transitions on x86; most other
1334 architectures force reset to be always executed
Linus Torvalds1da177e2005-04-16 15:20:36 -07001335 i8042.unlock [HW] Unlock (ignore) the keylock
Srihari Vijayaraghavan148e9a72015-01-07 16:25:53 -08001336 i8042.kbdreset [HW] Reset device connected to KBD port
Linus Torvalds1da177e2005-04-16 15:20:36 -07001337
1338 i810= [HW,DRM]
1339
Dmitry Torokhove70c9d52005-06-25 14:54:25 -07001340 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1341 indicates that the driver is running on unsupported
1342 hardware.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001343 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1344 does not match list of supported models.
1345 i8k.power_status
1346 [HW] Report power status in /proc/i8k
1347 (disabled by default)
1348 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1349 capability is set.
1350
Carsten Emde4dca20e2012-03-15 15:56:26 +01001351 i915.invert_brightness=
Carsten Emde7bd90902012-03-15 15:56:25 +01001352 [DRM] Invert the sense of the variable that is used to
1353 set the brightness of the panel backlight. Normally a
Carsten Emde4dca20e2012-03-15 15:56:26 +01001354 brightness value of 0 indicates backlight switched off,
1355 and the maximum of the brightness value sets the backlight
1356 to maximum brightness. If this parameter is set to 0
1357 (default) and the machine requires it, or this parameter
1358 is set to 1, a brightness value of 0 sets the backlight
1359 to maximum brightness, and the maximum of the brightness
1360 value switches the backlight off.
1361 -1 -- never invert brightness
1362 0 -- machine default
1363 1 -- force brightness inversion
Carsten Emde7bd90902012-03-15 15:56:25 +01001364
Linus Torvalds1da177e2005-04-16 15:20:36 -07001365 icn= [HW,ISDN]
1366 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1367
David Fries0af80c02009-02-25 20:28:21 +01001368 ide-core.nodma= [HW] (E)IDE subsystem
1369 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
Bartlomiej Zolnierkiewicz075affc2009-06-07 13:52:52 +02001370 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1371 .cdrom .chs .ignore_cable are additional options
Randy Dunlap1c10e932008-03-07 21:53:50 +01001372 See Documentation/ide/ide.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001373
Maciej W. Rozycki0f8b7f52014-10-25 17:03:52 +01001374 ide-generic.probe-mask= [HW] (E)IDE subsystem
1375 Format: <int>
1376 Probe mask for legacy ISA IDE ports. Depending on
1377 platform up to 6 ports are supported, enabled by
1378 setting corresponding bits in the mask to 1. The
1379 default value is 0x0, which has a special meaning.
1380 On systems that have PCI, it triggers scanning the
1381 PCI bus for the first and the second port, which
1382 are then probed. On systems without PCI the value
1383 of 0x0 enables probing the two first ports as if it
1384 was 0x3.
1385
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001386 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1387 Claim all unknown PCI IDE storage controllers.
1388
Andi Kleenf039b752007-05-02 19:27:12 +02001389 idle= [X86]
Len Brown69fb3672013-02-10 01:38:39 -05001390 Format: idle=poll, idle=halt, idle=nomwait
Randy Dunlapada9cfd2008-12-19 10:57:32 -08001391 Poll forces a polling idle loop that can slightly
1392 improve the performance of waking up a idle CPU, but
1393 will use a lot of power and make the system run hot.
1394 Not recommended.
Randy Dunlapada9cfd2008-12-19 10:57:32 -08001395 idle=halt: Halt is forced to be used for CPU idle.
Zhao Yakuic1e3b372008-06-24 17:58:53 +08001396 In such case C2/C3 won't be used again.
Randy Dunlapada9cfd2008-12-19 10:57:32 -08001397 idle=nomwait: Disable mwait for CPU C-states
Randy Dunlapa9913042005-10-23 12:57:11 -07001398
Maciej W. Rozycki503943e2015-11-13 00:48:29 +00001399 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1400 Format: { strict | legacy | 2008 | relaxed }
1401 Default: strict
1402
1403 Choose which programs will be accepted for execution
1404 based on the IEEE 754 NaN encoding(s) supported by
1405 the FPU and the NaN encoding requested with the value
1406 of an ELF file header flag individually set by each
1407 binary. Hardware implementations are permitted to
1408 support either or both of the legacy and the 2008 NaN
1409 encoding mode.
1410
1411 Available settings are as follows:
1412 strict accept binaries that request a NaN encoding
1413 supported by the FPU
1414 legacy only accept legacy-NaN binaries, if supported
1415 by the FPU
1416 2008 only accept 2008-NaN binaries, if supported
1417 by the FPU
1418 relaxed accept any binaries regardless of whether
1419 supported by the FPU
1420
1421 The FPU emulator is always able to support both NaN
1422 encodings, so if no FPU hardware is present or it has
1423 been disabled with 'nofpu', then the settings of
1424 'legacy' and '2008' strap the emulator accordingly,
1425 'relaxed' straps the emulator for both legacy-NaN and
1426 2008-NaN, whereas 'strict' enables legacy-NaN only on
1427 legacy processors and both NaN encodings on MIPS32 or
1428 MIPS64 CPUs.
1429
1430 The setting for ABS.fmt/NEG.fmt instruction execution
1431 mode generally follows that for the NaN encoding,
1432 except where unsupported by hardware.
1433
Ingo Molnar792908222006-12-06 20:40:51 -08001434 ignore_loglevel [KNL]
1435 Ignore loglevel setting - this will print /all/
1436 kernel messages to the console. Useful for debugging.
Yanmin Zhang0eca6b72011-10-31 17:11:25 -07001437 We also add it as printk module parameter, so users
1438 could change it dynamically, usually by
1439 /sys/module/printk/parameters/ignore_loglevel.
Ingo Molnar792908222006-12-06 20:40:51 -08001440
Konstantin Khlebnikovd977d562016-02-02 16:57:43 -08001441 ignore_rlimit_data
1442 Ignore RLIMIT_DATA setting for data mappings,
1443 print warning at first misuse. Can be changed via
1444 /sys/module/kernel/parameters/ignore_rlimit_data.
1445
Linus Torvalds1da177e2005-04-16 15:20:36 -07001446 ihash_entries= [KNL]
1447 Set number of hash buckets for inode cache.
1448
Mimi Zohar2fe5d6d2012-02-13 10:15:05 -05001449 ima_appraise= [IMA] appraise integrity measurements
Dmitry Kasatkin2faa6ef2014-05-08 13:11:29 +03001450 Format: { "off" | "enforce" | "fix" | "log" }
Mimi Zohar2fe5d6d2012-02-13 10:15:05 -05001451 default: "enforce"
1452
Mimi Zohar07f6a792011-03-09 22:25:48 -05001453 ima_appraise_tcb [IMA]
1454 The builtin appraise policy appraises all files
1455 owned by uid=0.
1456
Mimi Zohard68a6fe2016-12-19 16:22:57 -08001457 ima_canonical_fmt [IMA]
1458 Use the canonical format for the binary runtime
1459 measurements, instead of host native format.
1460
Mimi Zohar3323eec2009-02-04 09:06:58 -05001461 ima_hash= [IMA]
Mimi Zohare7a2ad72013-06-07 12:16:37 +02001462 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1463 | sha512 | ... }
Mimi Zohar3323eec2009-02-04 09:06:58 -05001464 default: "sha1"
1465
Mimi Zohare7a2ad72013-06-07 12:16:37 +02001466 The list of supported hash algorithms is defined
1467 in crypto/hash_info.h.
1468
Mimi Zohar24fd03c2015-06-11 20:48:33 -04001469 ima_policy= [IMA]
1470 The builtin measurement policy to load during IMA
1471 setup. Specyfing "tcb" as the value, measures all
1472 programs exec'd, files mmap'd for exec, and all files
1473 opened with the read mode bit set by either the
1474 effective uid (euid=0) or uid=0.
1475 Format: "tcb"
1476
1477 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
Eric Paris5789ba32009-05-21 15:47:06 -04001478 Load a policy which meets the needs of the Trusted
1479 Computing Base. This means IMA will measure all
1480 programs exec'd, files mmap'd for exec, and all files
1481 opened for read by uid=0.
1482
Roberto Sassu9b9d4ce2013-06-07 12:16:35 +02001483 ima_template= [IMA]
1484 Select one of defined IMA measurements template formats.
Roberto Sassu8265a2f2015-04-11 17:07:03 +02001485 Formats: { "ima" | "ima-ng" | "ima-sig" }
Roberto Sassu9b9d4ce2013-06-07 12:16:35 +02001486 Default: "ima-ng"
1487
Roberto Sassuc2426d22014-10-13 14:08:42 +02001488 ima_template_fmt=
1489 [IMA] Define a custom template format.
1490 Format: { "field1|...|fieldN" }
1491
Dmitry Kasatkin3bcced32014-02-26 17:05:20 +02001492 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1493 Format: <min_file_size>
1494 Set the minimal file size for using asynchronous hash.
1495 If left unspecified, ahash usage is disabled.
1496
1497 ahash performance varies for different data sizes on
1498 different crypto accelerators. This option can be used
1499 to achieve the best performance for a particular HW.
1500
Dmitry Kasatkin6edf7a82014-05-06 14:47:13 +03001501 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1502 Format: <bufsize>
1503 Set hashing buffer size. Default: 4k.
1504
1505 ahash performance varies for different chunk sizes on
1506 different crypto accelerators. This option can be used
1507 to achieve best performance for particular HW.
1508
Linus Torvalds1da177e2005-04-16 15:20:36 -07001509 init= [KNL]
1510 Format: <full_path>
1511 Run specified binary instead of /sbin/init as init
1512 process.
1513
1514 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1515 for working out where the kernel is dying during
1516 startup.
1517
Prarit Bhargava7b0b73d2014-06-04 16:12:17 -07001518 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1519 initcall functions. Useful for debugging built-in
1520 modules and initcalls.
1521
Linus Torvalds1da177e2005-04-16 15:20:36 -07001522 initrd= [BOOT] Specify the location of the initial ramdisk
1523
Dave Hansenacd547b2016-07-29 09:30:21 -07001524 init_pkru= [x86] Specify the default memory protection keys rights
1525 register contents for all processes. 0x55555554 by
1526 default (disallow access to all but pkey 0). Can
1527 override in debugfs after boot.
1528
Linus Torvalds1da177e2005-04-16 15:20:36 -07001529 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1530 Format: <irq>
1531
Fenghua Yu6bb2ff82013-05-21 15:35:17 -04001532 int_pln_enable [x86] Enable power limit notification interrupt
1533
Mimi Zohard726d8d2013-03-18 14:48:02 -04001534 integrity_audit=[IMA]
1535 Format: { "0" | "1" }
1536 0 -- basic integrity auditing messages. (Default)
1537 1 -- additional integrity auditing messages.
1538
Keshavamurthy, Anil Sba395922007-10-21 16:41:49 -07001539 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
Kyle McMartin0cd5c3c2009-02-04 14:29:19 -08001540 on
1541 Enable intel iommu driver.
Keshavamurthy, Anil Sba395922007-10-21 16:41:49 -07001542 off
1543 Disable intel iommu driver.
1544 igfx_off [Default Off]
1545 By default, gfx is mapped as normal device. If a gfx
1546 device has a dedicated DMAR unit, the DMAR unit is
1547 bypassed by not enabling DMAR with this option. In
1548 this case, gfx device will use physical address for
1549 DMA.
Keshavamurthy, Anil S7d3b03c2007-10-21 16:41:53 -07001550 forcedac [x86_64]
1551 With this option iommu will not optimize to look
Randy Dunlap16290242011-08-13 12:34:52 -07001552 for io virtual address below 32-bit forcing dual
Keshavamurthy, Anil S7d3b03c2007-10-21 16:41:53 -07001553 address cycle on pci bus for cards supporting greater
Randy Dunlap16290242011-08-13 12:34:52 -07001554 than 32-bit addressing. The default is to look
1555 for translation below 32-bit and if not available
Keshavamurthy, Anil S7d3b03c2007-10-21 16:41:53 -07001556 then look in the higher range.
mark gross5e0d2a62008-03-04 15:22:08 -08001557 strict [Default Off]
1558 With this option on every unmap_single operation will
1559 result in a hardware IOTLB flush operation as opposed
1560 to batching them for performance.
Youquan Song6dd9a7c2011-05-25 19:13:49 +01001561 sp_off [Default Off]
1562 By default, super page will be supported if Intel IOMMU
1563 has the capability. With this option, super page will
1564 not be supported.
David Woodhousec83b2f22015-06-12 10:15:49 +01001565 ecs_off [Default Off]
1566 By default, extended context tables will be supported if
1567 the hardware advertises that it has support both for the
1568 extended tables themselves, and also PASID support. With
1569 this option set, extended tables will not be used even
1570 on hardware which claims to support them.
Masanari Iida2e92c7a2011-12-15 01:18:52 +09001571
1572 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1573 0 disables intel_idle and fall back on acpi_idle.
baolex.ni22c6bbe2016-07-11 09:57:37 +08001574 1 to 9 specify maximum depth of C-state.
Masanari Iida2e92c7a2011-12-15 01:18:52 +09001575
Dirk Brandewie6be26492013-02-15 22:55:10 +01001576 intel_pstate= [X86]
1577 disable
1578 Do not enable intel_pstate as the default
1579 scaling driver for the supported processors
Rafael J. Wysocki001c76f2016-11-17 23:34:17 +01001580 passive
1581 Use intel_pstate as a scaling driver, but configure it
1582 to work with generic cpufreq governors (instead of
1583 enabling its internal governor). This mode cannot be
1584 used along with the hardware-managed P-states (HWP)
1585 feature.
Ethan Zhaoaa4ea342014-12-09 10:43:19 +09001586 force
1587 Enable intel_pstate on systems that prohibit it by default
1588 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1589 instead of acpi-cpufreq may disable platform features, such
1590 as thermal controls and power capping, that rely on ACPI
1591 P-States information being indicated to OSPM and therefore
1592 should be used with caution. This option does not work with
1593 processors that aren't supported by the intel_pstate driver
1594 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
Dirk Brandewie2f86dc42014-11-06 09:40:47 -08001595 no_hwp
1596 Do not enable hardware P state control (HWP)
1597 if available.
Kristen Carlson Accardid64c3b02015-02-06 13:41:55 -08001598 hwp_only
1599 Only load intel_pstate on systems which support
1600 hardware P state control (HWP) if available.
Srinivas Pandruvada9522a2f2016-04-27 15:48:06 -07001601 support_acpi_ppc
Srinivas Pandruvada2b3ec762016-04-27 15:48:08 -07001602 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1603 Description Table, specifies preferred power management
1604 profile as "Enterprise Server" or "Performance Server",
1605 then this feature is turned on by default.
Srinivas Pandruvada22449c02016-10-25 13:20:43 -07001606 per_cpu_perf_limits
1607 Allow per-logical-CPU P-State performance control limits using
1608 cpufreq sysfs interface
Dirk Brandewie6be26492013-02-15 22:55:10 +01001609
Chris Wrightd1423d52010-07-20 11:06:49 -07001610 intremap= [X86-64, Intel-IOMMU]
Chris Wrightd1423d52010-07-20 11:06:49 -07001611 on enable Interrupt Remapping (default)
1612 off disable Interrupt Remapping
1613 nosid disable Source ID checking
Suresh Siddha41750d32011-08-23 17:05:18 -07001614 no_x2apic_optout
1615 BIOS x2APIC opt-out request will be ignored
Feng Wub7d20632015-09-18 22:29:56 +08001616 nopost disable Interrupt Posting
Chris Wrightd1423d52010-07-20 11:06:49 -07001617
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001618 iomem= Disable strict checking of access to MMIO memory
1619 strict regions from userspace.
1620 relaxed
1621
1622 iommu= [x86]
1623 off
1624 force
1625 noforce
1626 biomerge
1627 panic
1628 nopanic
1629 merge
1630 nomerge
1631 forcesac
1632 soft
Alex Williamsonbcb71ab2011-10-21 15:56:24 -04001633 pt [x86, IA-64]
Thadeu Lima de Souza Cascardo4e287842014-10-23 19:19:35 -02001634 nobypass [PPC/POWERNV]
1635 Disable IOMMU bypass, using IOMMU for PCI devices.
Alex Williamsonbcb71ab2011-10-21 15:56:24 -04001636
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001637
1638 io7= [HW] IO7 for Marvel based alpha systems
1639 See comment before marvel_specify_io7 in
1640 arch/alpha/kernel/core_marvel.c.
1641
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05301642 io_delay= [X86] I/O delay method
Ingo Molnar6e7c4022008-01-30 13:30:05 +01001643 0x80
1644 Standard port 0x80 based delay
1645 0xed
1646 Alternate port 0xed based delay (needed on some systems)
Rene Hermanb02aae92008-01-30 13:30:05 +01001647 udelay
Ingo Molnar6e7c4022008-01-30 13:30:05 +01001648 Simple two microseconds delay
1649 none
1650 No delay
Rene Hermanb02aae92008-01-30 13:30:05 +01001651
Linus Torvalds1da177e2005-04-16 15:20:36 -07001652 ip= [IP_PNP]
J. Bruce Fieldsdc7a0812009-10-27 14:41:35 -04001653 See Documentation/filesystems/nfs/nfsroot.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001654
Thomas Gleixnerfbf19802016-02-03 19:52:23 +01001655 irqaffinity= [SMP] Set the default irq affinity mask
Noam Camus2d13e6c2016-10-11 13:51:35 -07001656 The argument is a cpu list, as described above.
Thomas Gleixnerfbf19802016-02-03 19:52:23 +01001657
Alan Cox200803d2005-06-28 20:45:18 -07001658 irqfixup [HW]
1659 When an interrupt is not handled search all handlers
1660 for it. Intended to get systems with badly broken
1661 firmware running.
1662
1663 irqpoll [HW]
1664 When an interrupt is not handled search all handlers
1665 for it. Also check all handlers each timer
1666 interrupt. Intended to get systems with badly broken
1667 firmware running.
1668
Linus Torvalds1da177e2005-04-16 15:20:36 -07001669 isapnp= [ISAPNP]
Randy Dunlapa9913042005-10-23 12:57:11 -07001670 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
Linus Torvalds1da177e2005-04-16 15:20:36 -07001671
1672 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
Noam Camus2d13e6c2016-10-11 13:51:35 -07001673 The argument is a cpu list, as described above.
Li Zefanb225d442008-11-06 12:53:39 -08001674
Linus Torvalds1da177e2005-04-16 15:20:36 -07001675 This option can be used to specify one or more CPUs
1676 to isolate from the general SMP balancing and scheduling
Li Zefanb225d442008-11-06 12:53:39 -08001677 algorithms. You can move a process onto or off an
1678 "isolated" CPU via the CPU affinity syscalls or cpuset.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001679 <cpu number> begins at 0 and the maximum value is
1680 "number of CPUs in system - 1".
1681
1682 This option is the preferred way to isolate CPUs. The
Randy Dunlapa9913042005-10-23 12:57:11 -07001683 alternative -- manually setting the CPU mask of all
1684 tasks in the system -- can cause problems and
1685 suboptimal load balancer performance.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001686
Randy Dunlapa9913042005-10-23 12:57:11 -07001687 iucv= [HW,NET]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001688
Joerg Roedel7d8bfa22013-04-09 21:27:19 +02001689 ivrs_ioapic [HW,X86_64]
1690 Provide an override to the IOAPIC-ID<->DEVICE-ID
1691 mapping provided in the IVRS ACPI table. For
1692 example, to map IOAPIC-ID decimal 10 to
1693 PCI device 00:14.0 write the parameter as:
1694 ivrs_ioapic[10]=00:14.0
1695
1696 ivrs_hpet [HW,X86_64]
1697 Provide an override to the HPET-ID<->DEVICE-ID
1698 mapping provided in the IVRS ACPI table. For
1699 example, to map HPET-ID decimal 0 to
1700 PCI device 00:14.0 write the parameter as:
1701 ivrs_hpet[0]=00:14.0
1702
Suravee Suthikulpanitca3bf5d2016-04-01 09:06:01 -04001703 ivrs_acpihid [HW,X86_64]
1704 Provide an override to the ACPI-HID:UID<->DEVICE-ID
1705 mapping provided in the IVRS ACPI table. For
1706 example, to map UART-HID:UID AMD0020:0 to
1707 PCI device 00:14.5 write the parameter as:
1708 ivrs_acpihid[00:14.5]=AMD0020:0
1709
Linus Torvalds1da177e2005-04-16 15:20:36 -07001710 js= [HW,JOY] Analog joystick
1711 See Documentation/input/joystick.txt.
1712
Kees Cook65fe9352016-06-13 15:10:02 -07001713 nokaslr [KNL]
1714 When CONFIG_RANDOMIZE_BASE is set, this disables
1715 kernel and module base offset ASLR (Address Space
1716 Layout Randomization).
Kees Cook24f2e022014-06-13 13:30:36 -07001717
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07001718 keepinitrd [HW,ARM]
1719
Taku Izumi342332e2016-03-15 14:55:22 -07001720 kernelcore= [KNL,X86,IA-64,PPC]
1721 Format: nn[KMGTPE] | "mirror"
1722 This parameter
Mel Gormaned7ed362007-07-17 04:03:14 -07001723 specifies the amount of memory usable by the kernel
1724 for non-movable allocations. The requested amount is
1725 spread evenly throughout all nodes in the system. The
1726 remaining memory in each node is used for Movable
1727 pages. In the event, a node is too small to have both
1728 kernelcore and Movable pages, kernelcore pages will
1729 take priority and other nodes will have a larger number
Weiping Pan675217f2013-09-30 13:45:10 -07001730 of Movable pages. The Movable zone is used for the
Mel Gormaned7ed362007-07-17 04:03:14 -07001731 allocation of pages that may be reclaimed or moved
1732 by the page migration subsystem. This means that
1733 HugeTLB pages may not be allocated from this zone.
1734 Note that allocations like PTEs-from-HighMem still
1735 use the HighMem zone if it exists, and the Normal
1736 zone if it does not.
1737
Taku Izumi342332e2016-03-15 14:55:22 -07001738 Instead of specifying the amount of memory (nn[KMGTPE]),
1739 you can specify "mirror" option. In case "mirror"
1740 option is specified, mirrored (reliable) memory is used
1741 for non-movable allocations and remaining memory is used
1742 for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1743 so you can NOT specify nn[KMGTPE] and "mirror" at the same
1744 time.
1745
Jason Wessel4fe1da42010-05-20 21:04:31 -05001746 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1747 Format: <Controller#>[,poll interval]
1748 The controller # is the number of the ehci usb debug
1749 port as it is probed via PCI. The poll interval is
1750 optional and is the number seconds in between
1751 each poll cycle to the debug port in case you need
1752 the functionality for interrupting the kernel with
1753 gdb or control-c on the dbgp connection. When
1754 not using this parameter you use sysrq-g to break into
1755 the kernel debugger.
1756
Jason Wessel84c08fd2010-05-20 21:04:24 -05001757 kgdboc= [KGDB,HW] kgdb over consoles.
Jason Wesselada64e42010-05-20 21:04:24 -05001758 Requires a tty driver that supports console polling,
1759 or a supported polling keyboard driver (non-usb).
Jason Wessel65b5ac12010-08-05 09:22:33 -05001760 Serial only format: <serial_device>[,baud]
1761 keyboard only format: kbd
1762 keyboard and serial format: kbd,<serial_device>[,baud]
1763 Optional Kernel mode setting:
1764 kms, kbd format: kms,kbd
1765 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
Jason Wessel6cdf6e02008-04-17 20:05:38 +02001766
Jason Wessel84c08fd2010-05-20 21:04:24 -05001767 kgdbwait [KGDB] Stop kernel execution and enter the
1768 kernel debugger at the earliest opportunity.
1769
Florian Fainelli9bed90c62008-08-23 18:54:37 +02001770 kmac= [MIPS] korina ethernet MAC address.
1771 Configure the RouterBoard 532 series on-chip
1772 Ethernet adapter MAC address.
1773
Catalin Marinas04f70332009-06-11 13:22:39 +01001774 kmemleak= [KNL] Boot-time kmemleak enable/disable
1775 Valid arguments: on, off
1776 Default: on
Masanari Iida47aeedd2014-10-24 21:24:59 +09001777 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1778 the default is off.
Catalin Marinas04f70332009-06-11 13:22:39 +01001779
Xishi Qiuc3ac14b2014-01-23 15:53:33 -08001780 kmemcheck= [X86] Boot-time kmemcheck enable/disable/one-shot mode
1781 Valid arguments: 0, 1, 2
1782 kmemcheck=0 (disabled)
1783 kmemcheck=1 (enabled)
1784 kmemcheck=2 (one-shot mode)
1785 Default: 2 (one-shot mode)
1786
Andre Przywarafef07aa2009-07-10 14:20:35 +02001787 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1788 Default is 0 (don't ignore, but inject #GP)
1789
Xiao Guangronga182d872010-09-20 22:17:48 +08001790 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1791 KVM MMU at runtime.
Andre Przywarafef07aa2009-07-10 14:20:35 +02001792 Default is 0 (off)
1793
Andre Przywarafef07aa2009-07-10 14:20:35 +02001794 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
Xiao Guangrong8475f942010-09-20 22:16:45 +08001795 Default is 1 (enabled)
Andre Przywarafef07aa2009-07-10 14:20:35 +02001796
1797 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1798 for all guests.
Randy Dunlap16290242011-08-13 12:34:52 -07001799 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
Andre Przywarafef07aa2009-07-10 14:20:35 +02001800
Andre Przywarafef07aa2009-07-10 14:20:35 +02001801 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1802 (virtualized MMU) support on capable Intel chips.
1803 Default is 1 (enabled)
1804
1805 kvm-intel.emulate_invalid_guest_state=
1806 [KVM,Intel] Enable emulation of invalid guest states
1807 Default is 0 (disabled)
1808
1809 kvm-intel.flexpriority=
1810 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1811 Default is 1 (enabled)
1812
Sasha Levine1a72ae2011-08-09 14:28:35 +03001813 kvm-intel.nested=
1814 [KVM,Intel] Enable VMX nesting (nVMX).
1815 Default is 0 (disabled)
1816
Andre Przywarafef07aa2009-07-10 14:20:35 +02001817 kvm-intel.unrestricted_guest=
1818 [KVM,Intel] Disable unrestricted guest feature
1819 (virtualized real and unpaged mode) on capable
1820 Intel chips. Default is 1 (enabled)
1821
1822 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1823 feature (tagged TLBs) on capable Intel chips.
1824 Default is 1 (enabled)
1825
Linus Torvalds1da177e2005-04-16 15:20:36 -07001826 l2cr= [PPC]
1827
Robert Brosea78bfbf2008-03-29 07:20:23 +11001828 l3cr= [PPC]
1829
Alan Coxcd4f0ef2007-07-31 00:37:59 -07001830 lapic [X86-32,APIC] Enable the local APIC even if BIOS
Randy Dunlapa9913042005-10-23 12:57:11 -07001831 disabled it.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001832
Suresh Siddha279f1462012-10-22 14:37:58 -07001833 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1834 value for LAPIC timer one-shot implementation. Default
1835 back to the programmable timer unit in the LAPIC.
1836
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05301837 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
Randy Dunlapada9cfd2008-12-19 10:57:32 -08001838 in C2 power state.
Thomas Gleixnere585bef2007-03-23 16:08:01 +01001839
FD Camifcb71f62008-01-06 19:08:56 +01001840 libata.dma= [LIBATA] DMA control
1841 libata.dma=0 Disable all PATA and SATA DMA
1842 libata.dma=1 PATA and SATA Disk DMA only
1843 libata.dma=2 ATAPI (CDROM) DMA only
Randy Dunlap16290242011-08-13 12:34:52 -07001844 libata.dma=4 Compact Flash DMA only
FD Camifcb71f62008-01-06 19:08:56 +01001845 Combinations also work, so libata.dma=3 enables DMA
1846 for disks and CDROMs, but not CFs.
Randy Dunlap16290242011-08-13 12:34:52 -07001847
Michael Prokop20308872009-08-06 00:14:10 +02001848 libata.ignore_hpa= [LIBATA] Ignore HPA limit
1849 libata.ignore_hpa=0 keep BIOS limits (default)
1850 libata.ignore_hpa=1 ignore limits, using full disk
FD Camifcb71f62008-01-06 19:08:56 +01001851
Dave Jones78e70c22007-09-27 11:50:13 -04001852 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
1853 when set.
1854 Format: <int>
1855
Tejun Heo33267322008-02-13 09:15:09 +09001856 libata.force= [LIBATA] Force configurations. The format is comma
1857 separated list of "[ID:]VAL" where ID is
Roman Fietze4c44f302010-04-21 12:17:12 +02001858 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
Tejun Heo33267322008-02-13 09:15:09 +09001859 matching port, link or device. Basically, it matches
1860 the ATA ID string printed on console by libata. If
1861 the whole ID part is omitted, the last PORT and DEVICE
1862 values are used. If ID hasn't been specified yet, the
1863 configuration applies to all ports, links and devices.
1864
1865 If only DEVICE is omitted, the parameter applies to
1866 the port and all links and devices behind it. DEVICE
1867 number of 0 either selects the first device or the
1868 first fan-out link behind PMP device. It does not
1869 select the host link. DEVICE number of 15 selects the
1870 host link and device attached to it.
1871
1872 The VAL specifies the configuration to force. As long
1873 as there's no ambiguity shortcut notation is allowed.
1874 For example, both 1.5 and 1.5G would work for 1.5Gbps.
1875 The following configurations can be forced.
1876
1877 * Cable type: 40c, 80c, short40c, unk, ign or sata.
1878 Any ID with matching PORT is used.
1879
1880 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
1881
1882 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
1883 udma[/][16,25,33,44,66,100,133] notation is also
1884 allowed.
1885
1886 * [no]ncq: Turn on or off NCQ.
1887
Martin K. Petersend7b16e42015-05-04 21:54:18 -04001888 * [no]ncqtrim: Turn off queued DSM TRIM.
1889
Tejun Heo05944bd2008-08-13 20:19:09 +09001890 * nohrst, nosrst, norst: suppress hard, soft
1891 and both resets.
1892
Dan Williamsca6d43b2012-06-21 23:41:41 -07001893 * rstonce: only attempt one reset during
1894 hot-unplug link recovery
1895
Tejun Heo43c9c592010-05-23 12:59:11 +02001896 * dump_id: dump IDENTIFY data.
1897
Vincent Pelletier966fbe12013-05-21 22:30:58 +02001898 * atapi_dmadir: Enable ATAPI DMADIR bridge support
1899
Robin H. Johnsonb8bd6dc2013-12-16 09:31:19 -08001900 * disable: Disable this device.
1901
Tejun Heo33267322008-02-13 09:15:09 +09001902 If there are multiple matching configurations changing
1903 the same attribute, the last one is used.
1904
Yinghai Lu95f72d12010-07-12 14:36:09 +10001905 memblock=debug [KNL] Enable memblock debug messages.
Randy Dunlap7c4be252009-01-06 14:42:44 -08001906
Linus Torvalds1da177e2005-04-16 15:20:36 -07001907 load_ramdisk= [RAM] List of ramdisks to load from floppy
Randy Dunlap31c00fc2008-11-13 21:33:24 +00001908 See Documentation/blockdev/ramdisk.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001909
Randy Dunlapa6b25b62006-01-14 13:21:19 -08001910 lockd.nlm_grace_period=P [NFS] Assign grace period.
1911 Format: <integer>
Linus Torvalds1da177e2005-04-16 15:20:36 -07001912
Randy Dunlapa6b25b62006-01-14 13:21:19 -08001913 lockd.nlm_tcpport=N [NFS] Assign TCP port.
1914 Format: <integer>
1915
1916 lockd.nlm_timeout=T [NFS] Assign timeout value.
1917 Format: <integer>
1918
1919 lockd.nlm_udpport=M [NFS] Assign UDP port.
1920 Format: <integer>
Linus Torvalds1da177e2005-04-16 15:20:36 -07001921
Paul E. McKenneyec4518a2014-09-12 10:50:01 -07001922 locktorture.nreaders_stress= [KNL]
1923 Set the number of locking read-acquisition kthreads.
1924 Defaults to being automatically set based on the
1925 number of online CPUs.
1926
1927 locktorture.nwriters_stress= [KNL]
1928 Set the number of locking write-acquisition kthreads.
1929
1930 locktorture.onoff_holdoff= [KNL]
1931 Set time (s) after boot for CPU-hotplug testing.
1932
1933 locktorture.onoff_interval= [KNL]
1934 Set time (s) between CPU-hotplug operations, or
1935 zero to disable CPU-hotplug testing.
1936
1937 locktorture.shuffle_interval= [KNL]
1938 Set task-shuffle interval (jiffies). Shuffling
1939 tasks allows some CPUs to go into dyntick-idle
1940 mode during the locktorture test.
1941
1942 locktorture.shutdown_secs= [KNL]
1943 Set time (s) after boot system shutdown. This
1944 is useful for hands-off automated testing.
1945
1946 locktorture.stat_interval= [KNL]
1947 Time (s) between statistics printk()s.
1948
1949 locktorture.stutter= [KNL]
1950 Time (s) to stutter testing, for example,
1951 specifying five seconds causes the test to run for
1952 five seconds, wait for five seconds, and so on.
1953 This tests the locking primitive's ability to
1954 transition abruptly to and from idle.
1955
1956 locktorture.torture_runnable= [BOOT]
1957 Start locktorture running at boot time.
1958
1959 locktorture.torture_type= [KNL]
1960 Specify the locking implementation to test.
1961
1962 locktorture.verbose= [KNL]
1963 Enable additional printk() statements.
1964
Linus Torvalds1da177e2005-04-16 15:20:36 -07001965 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
1966 Format: <irq>
1967
1968 loglevel= All Kernel Messages with a loglevel smaller than the
1969 console loglevel will be printed to the console. It can
1970 also be changed with klogd or other programs. The
1971 loglevels are defined as follows:
1972
1973 0 (KERN_EMERG) system is unusable
1974 1 (KERN_ALERT) action must be taken immediately
1975 2 (KERN_CRIT) critical conditions
1976 3 (KERN_ERR) error conditions
1977 4 (KERN_WARNING) warning conditions
1978 5 (KERN_NOTICE) normal but significant condition
1979 6 (KERN_INFO) informational
1980 7 (KERN_DEBUG) debug-level messages
1981
Randy Dunlapc756d082011-02-20 20:08:35 -08001982 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
Luis R. Rodriguez23b28992014-08-06 16:08:56 -07001983 in bytes. n must be a power of two and greater
1984 than the minimal size. The minimal size is defined
1985 by LOG_BUF_SHIFT kernel config parameter. There is
1986 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
1987 that allows to increase the default size depending on
1988 the number of CPUs. See init/Kconfig for more details.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001989
Randy Dunlapaccaa242007-10-16 01:29:37 -07001990 logo.nologo [FB] Disables display of the built-in Linux logo.
1991 This may be used to provide more screen space for
1992 kernel log messages and is useful when debugging
1993 kernel boot problems.
1994
Linus Torvalds1da177e2005-04-16 15:20:36 -07001995 lp=0 [LP] Specify parallel ports to use, e.g,
1996 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
1997 lp=reset first parallel port). 'lp=0' disables the
1998 lp=auto printer driver. 'lp=reset' (which can be
1999 specified in addition to the ports) causes
2000 attached printers to be reset. Using
2001 lp=port1,port2,... specifies the parallel ports
2002 to associate lp devices with, starting with
2003 lp0. A port specification may be 'none' to skip
2004 that lp device, or a parport name such as
2005 'parport0'. Specifying 'lp=auto' instead of a
2006 port specification list means that device IDs
2007 from each port should be examined, to see if
2008 an IEEE 1284-compliant printer is attached; if
2009 so, the driver will manage that printer.
2010 See also header of drivers/char/lp.c.
2011
2012 lpj=n [KNL]
2013 Sets loops_per_jiffy to given constant, thus avoiding
2014 time-consuming boot-time autodetection (up to 250 ms per
2015 CPU). 0 enables autodetection (default). To determine
2016 the correct value for your kernel, boot with normal
2017 autodetection and see what value is printed. Note that
2018 on SMP systems the preset will be applied to all CPUs,
2019 which is likely to cause problems if your CPUs need
2020 significantly divergent settings. An incorrect value
2021 will cause delays in the kernel to be wrong, leading to
2022 unpredictable I/O errors and other breakage. Although
2023 unlikely, in the extreme case this might damage your
2024 hardware.
2025
2026 ltpc= [NET]
2027 Format: <io>,<irq>,<dma>
2028
Randy Dunlap16290242011-08-13 12:34:52 -07002029 machvec= [IA-64] Force the use of a particular machine-vector
Randy Dunlapa9913042005-10-23 12:57:11 -07002030 (machvec) in a generic kernel.
2031 Example: machvec=hpzx1_swiotlb
Linus Torvalds1da177e2005-04-16 15:20:36 -07002032
Wu Zhangjin3209e702009-07-02 23:27:12 +08002033 machtype= [Loongson] Share the same kernel image file between different
2034 yeeloong laptop.
2035 Example: machtype=lemote-yeeloong-2f-7inch
2036
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002037 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2038 than or equal to this physical address is ignored.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002039
2040 maxcpus= [SMP] Maximum number of processors that an SMP kernel
Baoquan He7c142bf2016-08-24 13:06:45 +08002041 will bring up during bootup. maxcpus=n : n >= 0 limits
2042 the kernel to bring up 'n' processors. Surely after
2043 bootup you can bring up the other plugged cpu by executing
2044 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2045 only takes effect during system bootup.
2046 While n=0 is a special case, it is equivalent to "nosmp",
2047 which also disables the IO APIC.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002048
Kay Sieversd134b002011-07-31 22:08:04 +02002049 max_loop= [LOOP] The number of loop block devices that get
2050 (loop.max_loop) unconditionally pre-created at init time. The default
2051 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2052 of statically allocating a predefined number, loop
2053 devices can be requested on-demand with the
2054 /dev/loop-control interface.
Bob Picco2b2c3752005-06-29 18:00:00 -07002055
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002056 mce [X86-32] Machine Check Exception
Linus Torvalds1da177e2005-04-16 15:20:36 -07002057
Uwe Hermann71cced62008-10-20 09:32:21 -07002058 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
Andi Kleen909dd322007-10-17 18:04:38 +02002059
Linus Torvalds1da177e2005-04-16 15:20:36 -07002060 md= [HW] RAID subsystems devices and level
Jani Nikulae52347b2016-11-03 12:10:10 +02002061 See Documentation/admin-guide/md.rst.
Randy Dunlapa9913042005-10-23 12:57:11 -07002062
Linus Torvalds1da177e2005-04-16 15:20:36 -07002063 mdacon= [MDA]
2064 Format: <first>,<last>
2065 Specifies range of consoles to be captured by the MDA.
Randy Dunlapa9913042005-10-23 12:57:11 -07002066
Linus Torvalds1da177e2005-04-16 15:20:36 -07002067 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2068 Amount of memory to be used when the kernel is not able
2069 to see the whole system memory or for test.
Wen Congyangfbb97d82012-12-17 15:59:29 -08002070 [X86] Work as limiting max address. Use together
2071 with memmap= to avoid physical address space collisions.
2072 Without memmap= PCI devices could be placed at addresses
2073 belonging to unused RAM.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002074
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002075 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
Linus Torvalds1da177e2005-04-16 15:20:36 -07002076 memory.
2077
Paul Mundt6902aa82008-09-21 17:14:42 +09002078 memchunk=nn[KMG]
2079 [KNL,SH] Allow user to override the default size for
2080 per-device physically contiguous DMA buffers.
2081
Vitaly Kuznetsov86dd9952016-05-19 17:13:06 -07002082 memhp_default_state=online/offline
2083 [KNL] Set the initial state for the memory hotplug
2084 onlining policy. If not specified, the default value is
2085 set according to the
2086 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2087 option.
2088 See Documentation/memory-hotplug.txt.
2089
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05302090 memmap=exactmap [KNL,X86] Enable setting of an exact
Linus Torvalds1da177e2005-04-16 15:20:36 -07002091 E820 memory map, as specified by the user.
2092 Such memmap=exactmap lines can be constructed based on
2093 BIOS output or other requirements. See the memmap=nn@ss
2094 option description.
2095
2096 memmap=nn[KMG]@ss[KMG]
Randy Dunlap277cba12014-02-06 12:04:19 -08002097 [KNL] Force usage of a specific region of memory.
2098 Region of memory to be used is from ss to ss+nn.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002099
2100 memmap=nn[KMG]#ss[KMG]
2101 [KNL,ACPI] Mark specific memory as ACPI data.
Randy Dunlap277cba12014-02-06 12:04:19 -08002102 Region of memory to be marked is from ss to ss+nn.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002103
2104 memmap=nn[KMG]$ss[KMG]
2105 [KNL,ACPI] Mark specific memory as reserved.
Randy Dunlap277cba12014-02-06 12:04:19 -08002106 Region of memory to be reserved is from ss to ss+nn.
Pavel Machek13128482008-03-24 12:29:43 -07002107 Example: Exclude memory from 0x18690000-0x1869ffff
2108 memmap=64K$0x18690000
2109 or
2110 memmap=0x10000$0x18690000
Linus Torvalds1da177e2005-04-16 15:20:36 -07002111
Christoph Hellwigec776ef2015-04-01 09:12:18 +02002112 memmap=nn[KMG]!ss[KMG]
2113 [KNL,X86] Mark specific memory as protected.
2114 Region of memory to be used, from ss to ss+nn.
2115 The memory region may be marked as e820 type 12 (0xc)
2116 and is NVDIMM or ADR memory.
2117
Jeremy Fitzhardinge9f077872008-09-07 01:51:34 -07002118 memory_corruption_check=0/1 [X86]
2119 Some BIOSes seem to corrupt the first 64k of
2120 memory when doing things like suspend/resume.
2121 Setting this option will scan the memory
2122 looking for corruption. Enabling this will
2123 both detect corruption and prevent the kernel
2124 from using the memory being corrupted.
2125 However, its intended as a diagnostic tool; if
2126 repeatable BIOS-originated corruption always
2127 affects the same memory, you can use memmap=
2128 to prevent the kernel from using that memory.
2129
2130 memory_corruption_check_size=size [X86]
2131 By default it checks for corruption in the low
2132 64k, making this memory unavailable for normal
2133 use. Use this parameter to scan for
2134 corruption in more or less memory.
2135
2136 memory_corruption_check_period=seconds [X86]
2137 By default it checks for corruption every 60
2138 seconds. Use this parameter to check at some
2139 other rate. 0 disables periodic checking.
2140
Vladimir Murzine4b0db72015-04-14 15:48:43 -07002141 memtest= [KNL,X86,ARM] Enable memtest
Yinghai Luc64df702008-03-21 18:56:19 -07002142 Format: <integer>
Yinghai Luc64df702008-03-21 18:56:19 -07002143 default : 0 <disable>
Andreas Herrmann9e5f6cf2009-02-25 11:30:45 +01002144 Specifies the number of memtest passes to be
2145 performed. Each pass selects another test
2146 pattern from a given set of patterns. Memtest
2147 fills the memory with this pattern, validates
2148 memory contents and reserves bad memory
2149 regions that are detected.
Yinghai Luc64df702008-03-21 18:56:19 -07002150
Rafael J. Wysocki406e7932016-11-21 22:45:40 +01002151 mem_sleep_default= [SUSPEND] Default system suspend mode:
2152 s2idle - Suspend-To-Idle
2153 shallow - Power-On Suspend or equivalent (if supported)
2154 deep - Suspend-To-RAM or equivalent (if supported)
2155 See Documentation/power/states.txt.
2156
Linus Torvalds1da177e2005-04-16 15:20:36 -07002157 meye.*= [HW] Set MotionEye Camera parameters
2158 See Documentation/video4linux/meye.txt.
2159
Andres Salomon8f368812007-10-12 23:04:06 +02002160 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2161 Multi-Function General Purpose Timers on AMD Geode
2162 platforms.
2163
Willy Tarreaue6c4dc62008-01-30 13:33:33 +01002164 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2165 the BIOS has incorrectly applied a workaround. TinyBIOS
2166 version 0.98 is known to be affected, 0.99 fixes the
2167 problem by letting the user disable the workaround.
2168
Linus Torvalds1da177e2005-04-16 15:20:36 -07002169 mga= [HW,DRM]
2170
Randy Dunlap1c207f952008-11-19 15:36:16 -08002171 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2172 physical address is ignored.
2173
Michel Pollet39f45d72009-05-20 11:10:31 +01002174 mini2440= [ARM,HW,KNL]
2175 Format:[0..2][b][c][t]
2176 Default: "0tb"
2177 MINI2440 configuration specification:
2178 0 - The attached screen is the 3.5" TFT
2179 1 - The attached screen is the 7" TFT
2180 2 - The VGA Shield is attached (1024x768)
2181 Leaving out the screen size parameter will not load
2182 the TFT driver, and the framebuffer will be left
2183 unconfigured.
2184 b - Enable backlight. The TFT backlight pin will be
2185 linked to the kernel VESA blanking code and a GPIO
2186 LED. This parameter is not necessary when using the
2187 VGA shield.
2188 c - Enable the s3c camera interface.
2189 t - Reserved for enabling touchscreen support. The
2190 touchscreen support is not enabled in the mainstream
2191 kernel as of 2.6.30, a preliminary port can be found
2192 in the "bleeding edge" mini2440 support kernel at
2193 http://repo.or.cz/w/linux-2.6/mini2440.git
2194
Mel Gorman6b74ab92008-07-23 21:26:49 -07002195 mminit_loglevel=
2196 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2197 parameter allows control of the logging verbosity for
2198 the additional memory initialisation checks. A value
2199 of 0 disables mminit logging and a level of 4 will
2200 log everything. Information is printed at KERN_DEBUG
2201 so loglevel=8 may also need to be specified.
2202
Rusty Russell106a4ee2012-09-26 10:09:40 +01002203 module.sig_enforce
2204 [KNL] When CONFIG_MODULE_SIG is set, this means that
2205 modules without (valid) signatures will fail to load.
Paul Bolle2a039be2013-03-25 20:42:06 +01002206 Note that if CONFIG_MODULE_SIG_FORCE is set, that
Rusty Russell106a4ee2012-09-26 10:09:40 +01002207 is always true, so this option does nothing.
2208
Prarit Bhargavabe7de5f2016-07-21 15:37:56 +09302209 module_blacklist= [KNL] Do not load a comma-separated list of
2210 modules. Useful for debugging problem modules.
2211
Linus Torvalds1da177e2005-04-16 15:20:36 -07002212 mousedev.tap_time=
2213 [MOUSE] Maximum time between finger touching and
2214 leaving touchpad surface for touch to be considered
2215 a tap and be reported as a left button click (for
2216 touchpads working in absolute mode only).
2217 Format: <msecs>
2218 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2219 reporting absolute coordinates, such as tablets
2220 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2221 reporting absolute coordinates, such as tablets
2222
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05302223 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002224 is similar to kernelcore except it specifies the
2225 amount of memory used for migratable allocations.
2226 If both kernelcore and movablecore is specified,
2227 then kernelcore will be at *least* the specified
2228 value but may be more. If movablecore on its own
2229 is specified, the administrator must be careful
2230 that the amount of memory usable for all allocations
2231 is not too small.
2232
Reza Arbab39fa104d2016-12-12 16:42:55 -08002233 movable_node [KNL] Boot-time switch to enable the effects
Tang Chenc5320922013-11-12 15:08:10 -08002234 of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2235
Linus Torvalds1da177e2005-04-16 15:20:36 -07002236 MTD_Partition= [MTD]
2237 Format: <name>,<region-number>,<size>,<offset>
2238
Randy Dunlapa9913042005-10-23 12:57:11 -07002239 MTD_Region= [MTD] Format:
2240 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
Linus Torvalds1da177e2005-04-16 15:20:36 -07002241
2242 mtdparts= [MTD]
Randy Dunlapc8facbb2007-07-31 00:37:40 -07002243 See drivers/mtd/cmdlinepart.c.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002244
Will Schmidt4e89a2d2010-09-28 15:33:12 +00002245 multitce=off [PPC] This parameter disables the use of the pSeries
2246 firmware feature for updating multiple TCE entries
2247 at a time.
2248
Rohit Hagargundgi5988af22009-05-12 13:46:57 -07002249 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2250
2251 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2252
2253 boundary - index of last SLC block on Flex-OneNAND.
2254 The remaining blocks are configured as MLC blocks.
2255 lock - Configure if Flex-OneNAND boundary should be locked.
2256 Once locked, the boundary cannot be changed.
2257 1 indicates lock status, 0 indicates unlock status.
2258
Ben Dooks9db829f2008-07-03 11:24:29 +01002259 mtdset= [ARM]
2260 ARM/S3C2412 JIVE boot control
2261
2262 See arch/arm/mach-s3c2412/mach-jive.c
2263
Linus Torvalds1da177e2005-04-16 15:20:36 -07002264 mtouchusb.raw_coordinates=
Randy Dunlapa9913042005-10-23 12:57:11 -07002265 [HW] Make the MicroTouch USB driver use raw coordinates
2266 ('y', default) or cooked coordinates ('n')
Linus Torvalds1da177e2005-04-16 15:20:36 -07002267
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002268 mtrr_chunk_size=nn[KMG] [X86]
Matt LaPlante19f59462009-04-27 15:06:31 +02002269 used for mtrr cleanup. It is largest continuous chunk
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002270 that could hold holes aka. UC entries.
2271
2272 mtrr_gran_size=nn[KMG] [X86]
2273 Used for mtrr cleanup. It is granularity of mtrr block.
2274 Default is 1.
2275 Large value could prevent small alignment from
2276 using up MTRRs.
2277
2278 mtrr_spare_reg_nr=n [X86]
2279 Format: <integer>
2280 Range: 0,7 : spare reg number
2281 Default : 1
2282 Used for mtrr cleanup. It is spare mtrr entries number.
2283 Set to 2 or more if your graphical card needs more.
2284
Linus Torvalds1da177e2005-04-16 15:20:36 -07002285 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2286
Linus Torvalds1da177e2005-04-16 15:20:36 -07002287 netdev= [NET] Network devices parameters
2288 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2289 Note that mem_start is often overloaded to mean
2290 something different and driver-specific.
Randy Dunlapa9913042005-10-23 12:57:11 -07002291 This usage is only documented in each driver source
2292 file if at all.
2293
Krzysztof Piotr Oledzki58401572008-07-21 10:01:34 -07002294 nf_conntrack.acct=
2295 [NETFILTER] Enable connection tracking flow accounting
2296 0 to disable accounting
2297 1 to enable accounting
Tim Gardnerd70a0112010-06-25 14:46:56 +02002298 Default value is 0.
Krzysztof Piotr Oledzki58401572008-07-21 10:01:34 -07002299
Chuck Lever306a0752010-09-17 10:54:37 -04002300 nfsaddrs= [NFS] Deprecated. Use ip= instead.
J. Bruce Fieldsdc7a0812009-10-27 14:41:35 -04002301 See Documentation/filesystems/nfs/nfsroot.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002302
2303 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
J. Bruce Fieldsdc7a0812009-10-27 14:41:35 -04002304 See Documentation/filesystems/nfs/nfsroot.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002305
Chuck Lever306a0752010-09-17 10:54:37 -04002306 nfsrootdebug [NFS] enable nfsroot debugging messages.
2307 See Documentation/filesystems/nfs/nfsroot.txt.
2308
Trond Myklebust5405fc42016-08-29 20:03:52 -04002309 nfs.callback_nr_threads=
2310 [NFSv4] set the total number of threads that the
2311 NFS client will assign to service NFSv4 callback
2312 requests.
2313
Trond Myklebusta72b4422006-01-03 09:55:41 +01002314 nfs.callback_tcpport=
2315 [NFS] set the TCP port on which the NFSv4 callback
2316 channel should listen.
2317
Trond Myklebuste571cbf2009-08-19 18:12:27 -04002318 nfs.cache_getent=
2319 [NFS] sets the pathname to the program which is used
2320 to update the NFS client cache entries.
2321
2322 nfs.cache_getent_timeout=
2323 [NFS] sets the timeout after which an attempt to
2324 update a cache entry is deemed to have failed.
2325
Trond Myklebust58df0952006-01-03 09:55:57 +01002326 nfs.idmap_cache_timeout=
2327 [NFS] set the maximum lifetime for idmapper cache
2328 entries.
2329
Trond Myklebustf43bf0b2007-10-09 12:01:04 -04002330 nfs.enable_ino64=
2331 [NFS] enable 64-bit inode numbers.
2332 If zero, the NFS client will fake up a 32-bit inode
2333 number for the readdir() and stat() syscalls instead
2334 of returning the full 64-bit number.
2335 The default is to return 64-bit inode numbers.
2336
Trond Myklebust5405fc42016-08-29 20:03:52 -04002337 nfs.max_session_cb_slots=
2338 [NFSv4.1] Sets the maximum number of session
2339 slots the client will assign to the callback
2340 channel. This determines the maximum number of
2341 callbacks the client will process in parallel for
2342 a particular server.
2343
Trond Myklebustef159e92012-02-06 19:50:40 -05002344 nfs.max_session_slots=
2345 [NFSv4.1] Sets the maximum number of session slots
2346 the client will attempt to negotiate with the server.
2347 This limits the number of simultaneous RPC requests
2348 that the client can send to the NFSv4.1 server.
2349 Note that there is little point in setting this
2350 value higher than the max_tcp_slot_table_limit.
2351
Trond Myklebustb064eca22011-02-22 15:44:32 -08002352 nfs.nfs4_disable_idmapping=
Trond Myklebust074b1d12012-01-09 13:46:26 -05002353 [NFSv4] When set to the default of '1', this option
2354 ensures that both the RPC level authentication
2355 scheme and the NFS level operations agree to use
2356 numeric uids/gids if the mount is using the
2357 'sec=sys' security flavour. In effect it is
2358 disabling idmapping, which can make migration from
2359 legacy NFSv2/v3 systems to NFSv4 easier.
2360 Servers that do not support this mode of operation
2361 will be autodetected by the client, and it will fall
2362 back to using the idmapper.
2363 To turn off this behaviour, set the value to '0'.
Chuck Lever6f2ea7f2012-09-14 17:24:41 -04002364 nfs.nfs4_unique_id=
2365 [NFS4] Specify an additional fixed unique ident-
2366 ification string that NFSv4 clients can insert into
2367 their nfs_client_id4 string. This is typically a
2368 UUID that is generated at system install time.
Trond Myklebustb064eca22011-02-22 15:44:32 -08002369
Weston Andros Adamsondb8ac8b2012-02-17 15:20:24 -05002370 nfs.send_implementation_id =
2371 [NFSv4.1] Send client implementation identification
2372 information in exchange_id requests.
2373 If zero, no implementation identification information
2374 will be sent.
2375 The default is to send the implementation identification
2376 information.
Jani Nikulae52347b2016-11-03 12:10:10 +02002377
Trond Myklebustf6de7a32013-09-04 10:08:54 -04002378 nfs.recover_lost_locks =
2379 [NFSv4] Attempt to recover locks that were lost due
2380 to a lease timeout on the server. Please note that
2381 doing this risks data corruption, since there are
2382 no guarantees that the file will remain unchanged
2383 after the locks are lost.
2384 If you want to enable the kernel legacy behaviour of
2385 attempting to recover these locks, then set this
2386 parameter to '1'.
2387 The default parameter value of '0' causes the kernel
2388 not to attempt recovery of lost locks.
Weston Andros Adamsondb8ac8b2012-02-17 15:20:24 -05002389
Trond Myklebustbbf58bf2015-08-24 20:39:18 -04002390 nfs4.layoutstats_timer =
2391 [NFSv4.2] Change the rate at which the kernel sends
2392 layoutstats to the pNFS metadata server.
2393
2394 Setting this to value to 0 causes the kernel to use
2395 whatever value is the default set by the layout
2396 driver. A non-zero value sets the minimum interval
2397 in seconds between layoutstats transmissions.
2398
J. Bruce Fieldse9541ce2012-03-22 16:07:18 -04002399 nfsd.nfs4_disable_idmapping=
2400 [NFSv4] When set to the default of '1', the NFSv4
2401 server will return only numeric uids and gids to
2402 clients using auth_sys, and will accept numeric uids
2403 and gids from such clients. This is intended to ease
2404 migration from NFSv2/v3.
Weston Andros Adamsondb8ac8b2012-02-17 15:20:24 -05002405
Sachin Bhamare18d98f62012-03-19 20:47:58 -07002406 objlayoutdriver.osd_login_prog=
2407 [NFS] [OBJLAYOUT] sets the pathname to the program which
2408 is used to automatically discover and login into new
2409 osd-targets. Please see:
2410 Documentation/filesystems/pnfs.txt for more explanations
2411
Paul Mundt1e1030d2009-09-01 17:38:32 +09002412 nmi_debug= [KNL,AVR32,SH] Specify one or more actions to take
Haavard Skinnemoene7ba1762007-10-10 14:58:29 +02002413 when a NMI is triggered.
2414 Format: [state][,regs][,debounce][,die]
2415
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05302416 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
Don Zickusfef2c9b2011-03-22 16:34:16 -07002417 Format: [panic,][nopanic,][num]
Ulrich Obergfell195daf62015-04-14 15:44:13 -07002418 Valid num: 0 or 1
Pranith Kumar334bb792015-10-10 15:40:42 -04002419 0 - turn hardlockup detector in nmi_watchdog off
2420 1 - turn hardlockup detector in nmi_watchdog on
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002421 When panic is specified, panic when an NMI watchdog
Don Zickusfef2c9b2011-03-22 16:34:16 -07002422 timeout occurs (or 'nopanic' to override the opposite
Pranith Kumar334bb792015-10-10 15:40:42 -04002423 default). To disable both hard and soft lockup detectors,
2424 please see 'nowatchdog'.
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002425 This is useful when you use a panic=... timeout and
2426 need the box quickly up again.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002427
Anton Vorontsovbff38772009-07-08 11:10:56 -07002428 netpoll.carrier_timeout=
2429 [NET] Specifies amount of time (in seconds) that
2430 netpoll should wait for a carrier. By default netpoll
2431 waits 4 seconds.
2432
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002433 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
Linus Torvalds1da177e2005-04-16 15:20:36 -07002434 emulation library even if a 387 maths coprocessor
2435 is present.
2436
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002437 no_console_suspend
2438 [HW] Never suspend the console
2439 Disable suspending of consoles during suspend and
2440 hibernate operations. Once disabled, debugging
2441 messages can reach various consoles while the rest
2442 of the system is being put to sleep (ie, while
2443 debugging driver suspend/resume hooks). This may
2444 not work reliably with all consoles, but is known
2445 to work with serial and VGA consoles.
Yanmin Zhang134620f2011-10-31 17:11:27 -07002446 To facilitate more flexible debugging, we also add
2447 console_suspend, a printk module parameter to control
2448 it. Users could use console_suspend (usually
2449 /sys/module/printk/parameters/console_suspend) to
2450 turn on/off it dynamically.
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002451
Christoph Lameterc1aee212007-05-31 00:40:47 -07002452 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2453 caches in the slab allocator. Saves per-node memory,
2454 but will impact performance.
Paul Menage3395ee02006-12-06 20:32:16 -08002455
Randy Dunlapa9913042005-10-23 12:57:11 -07002456 noalign [KNL,ARM]
2457
Linus Torvalds1da177e2005-04-16 15:20:36 -07002458 noapic [SMP,APIC] Tells the kernel to not make use of any
2459 IOAPICs that may be present in the system.
2460
Mike Galbraith5091faa2010-11-30 14:18:03 +01002461 noautogroup Disable scheduler automatic task group creation.
2462
Linus Torvalds1da177e2005-04-16 15:20:36 -07002463 nobats [PPC] Do not use BATs for mapping kernel lowmem
2464 on "Classic" PPC cores.
2465
2466 nocache [ARM]
Randy Dunlapa9913042005-10-23 12:57:11 -07002467
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002468 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2469
Shailabh Nagar163ecdf2006-07-30 03:03:11 -07002470 nodelayacct [KNL] Disable per-task delay accounting
2471
Paul Mundt6902aa82008-09-21 17:14:42 +09002472 nodsp [SH] Disable hardware DSP at boot time.
2473
Dave Youngb2e0a542014-08-14 17:15:26 +08002474 noefi Disable EFI runtime services support.
Huang, Ying8b2cb7a2008-01-30 13:32:11 +01002475
Linus Torvalds1da177e2005-04-16 15:20:36 -07002476 noexec [IA-64]
2477
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05302478 noexec [X86]
Jiri Slabyf5a1b192008-04-12 10:28:25 +02002479 On X86-32 available only on PAE configured kernels.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002480 noexec=on: enable non-executable mappings (default)
Jiri Slabyf5a1b192008-04-12 10:28:25 +02002481 noexec=off: disable non-executable mappings
2482
H. Peter Anvin52b61792012-09-21 12:43:13 -07002483 nosmap [X86]
2484 Disable SMAP (Supervisor Mode Access Prevention)
2485 even if it is supported by processor.
2486
Fenghua Yude5397a2011-05-11 16:51:05 -07002487 nosmep [X86]
H. Peter Anvin52b61792012-09-21 12:43:13 -07002488 Disable SMEP (Supervisor Mode Execution Prevention)
Fenghua Yude5397a2011-05-11 16:51:05 -07002489 even if it is supported by processor.
2490
Jiri Slabyf5a1b192008-04-12 10:28:25 +02002491 noexec32 [X86-64]
2492 This affects only 32-bit executables.
2493 noexec32=on: enable non-executable mappings (default)
2494 read doesn't imply executable mappings
2495 noexec32=off: disable non-executable mappings
2496 read implies executable mappings
Linus Torvalds1da177e2005-04-16 15:20:36 -07002497
Maciej W. Rozyckifab43ef2015-04-03 23:23:34 +01002498 nofpu [MIPS,SH] Disable hardware FPU at boot time.
Paul Mundt6902aa82008-09-21 17:14:42 +09002499
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002500 nofxsr [BUGS=X86-32] Disables x86 floating point extended
Chuck Ebbert4f886512006-03-23 02:59:34 -08002501 register save and restore. The kernel will only save
2502 legacy floating-point registers on task switch.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002503
Toshi Kani0ddab1d2015-04-14 15:47:20 -07002504 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2505
Sascha Silbe52c48c52016-04-05 12:53:38 +02002506 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
2507 Equivalent to smt=1.
2508
Suresh Siddha0c752a92009-05-22 12:17:45 -07002509 noxsave [BUGS=X86] Disables x86 extended register state save
2510 and restore using xsave. The kernel will fallback to
2511 enabling legacy floating-point and sse state.
2512
Fenghua Yub6f42a42014-05-29 11:12:31 -07002513 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2514 register states. The kernel will fall back to use
2515 xsave to save the states. By using this parameter,
2516 performance of saving the states is degraded because
2517 xsave doesn't support modified optimization while
2518 xsaveopt supports it on xsaveopt enabled systems.
2519
2520 noxsaves [X86] Disables xsaves and xrstors used in saving and
2521 restoring x86 extended register state in compacted
2522 form of xsave area. The kernel will fall back to use
2523 xsaveopt and xrstor to save and restore the states
2524 in standard form of xsave area. By using this
2525 parameter, xsave area per process might occupy more
2526 memory on xsaves enabled systems.
2527
Paulius Zaleckas01a24d22009-03-31 13:55:44 +01002528 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2529 wfi(ARM) instruction doesn't work correctly and not to
2530 use it. This is also useful when using JTAG debugger.
Randy Dunlapa9913042005-10-23 12:57:11 -07002531
Serge E. Hallyn1f29fae2008-11-05 16:08:52 -06002532 no_file_caps Tells the kernel not to honor file capabilities. The
2533 only way then for a file to be executed with privilege
2534 is to be setuid root or executed by root.
2535
Linus Torvalds1da177e2005-04-16 15:20:36 -07002536 nohalt [IA-64] Tells the kernel not to use the power saving
2537 function PAL_HALT_LIGHT when idle. This increases
2538 power-consumption. On the positive side, it reduces
2539 interrupt wake-up latency, which may improve performance
2540 in certain environments such as networked servers or
2541 real-time systems.
2542
Kees Cooka6e15a32014-06-13 13:30:35 -07002543 nohibernate [HIBERNATION] Disable hibernation and resume.
2544
Thomas Gleixner79bf2bb2007-02-16 01:28:03 -08002545 nohz= [KNL] Boottime enable/disable dynamic ticks
2546 Valid arguments: on, off
2547 Default: on
2548
Frederic Weisbeckerc5bfece2013-04-12 16:45:34 +02002549 nohz_full= [KNL,BOOT]
Noam Camus2d13e6c2016-10-11 13:51:35 -07002550 The argument is a cpu list, as described above.
Frederic Weisbeckerc5bfece2013-04-12 16:45:34 +02002551 In kernels built with CONFIG_NO_HZ_FULL=y, set
Frederic Weisbeckera8318812012-12-18 17:32:19 +01002552 the specified list of CPUs whose tick will be stopped
Frederic Weisbecker0453b432013-03-27 02:18:34 +01002553 whenever possible. The boot CPU will be forced outside
2554 the range to maintain the timekeeping.
Frederic Weisbeckerd1e43fa2013-03-26 23:47:24 +01002555 The CPUs in this range must also be included in the
2556 rcu_nocbs= set.
Frederic Weisbeckera8318812012-12-18 17:32:19 +01002557
Paul Mundteeee7852009-04-02 12:31:16 +09002558 noiotrap [SH] Disables trapped I/O port accesses.
2559
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002560 noirqdebug [X86-32] Disables the code which attempts to detect and
Linus Torvalds1da177e2005-04-16 15:20:36 -07002561 disable unhandled interrupt sources.
2562
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05302563 no_timer_check [X86,APIC] Disables the code which tests for
Zachary Amsden8542b202006-12-07 02:14:09 +01002564 broken timer IRQ sources.
2565
Linus Torvalds1da177e2005-04-16 15:20:36 -07002566 noisapnp [ISAPNP] Disables ISA PnP code.
2567
2568 noinitrd [RAM] Tells the kernel not to load any configured
2569 initial RAM disk.
2570
Weidong Han03ea8152009-04-17 16:42:15 +08002571 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2572 remapping.
Chris Wrightd1423d52010-07-20 11:06:49 -07002573 [Deprecated - use intremap=off]
Weidong Han03ea8152009-04-17 16:42:15 +08002574
Linus Torvalds1da177e2005-04-16 15:20:36 -07002575 nointroute [IA-64]
2576
Andy Lutomirskid12a72b2016-01-29 11:42:58 -08002577 noinvpcid [X86] Disable the INVPCID cpu feature.
2578
Randy Dunlap16290242011-08-13 12:34:52 -07002579 nojitter [IA-64] Disables jitter checking for ITC timers.
Tony Luck0aa366f2007-07-20 11:22:30 -07002580
Jiri Kosina9cf4c4f2010-08-16 17:51:20 +02002581 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2582
Gleb Natapovfd10cde2010-10-14 11:22:51 +02002583 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2584 fault handling.
2585
Alexey Makhalov80e9a4f2016-10-28 00:54:32 -07002586 no-vmw-sched-clock
2587 [X86,PV_OPS] Disable paravirtualized VMware scheduler
2588 clock and use the default one.
2589
Glauber Costad910f5c2011-07-11 15:28:19 -04002590 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2591 steal time is computed, but won't influence scheduler
2592 behaviour
2593
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002594 nolapic [X86-32,APIC] Do not enable or use the local APIC.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002595
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002596 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
Thomas Gleixnerad62ca22007-03-22 00:11:21 -08002597
Linus Torvalds1da177e2005-04-16 15:20:36 -07002598 noltlbs [PPC] Do not use large page/tlb entries for kernel
Christophe Leroyf15eea62016-02-09 17:07:52 +01002599 lowmem mapping on PPC40x and PPC8xx
Linus Torvalds1da177e2005-04-16 15:20:36 -07002600
Horms312f1f02006-02-22 09:57:55 +09002601 nomca [IA-64] Disable machine check abort handling
2602
Masanari Iida13696e02015-05-16 02:16:43 +09002603 nomce [X86-32] Disable Machine Check Exception
Hormsabe37e52006-04-01 01:36:09 +02002604
Andres Salomon83d73842007-10-12 23:04:06 +02002605 nomfgpt [X86-32] Disable Multi-Function General Purpose
2606 Timer usage (for AMD Geode machines).
2607
Don Zickusbda62632011-10-13 15:14:27 -04002608 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2609 shutdown the other cpus. Instead use the REBOOT_VECTOR
2610 irq.
2611
Dave Young02608be2012-02-01 10:33:14 +08002612 nomodule Disable module load
2613
Jiri Kosina016ddd92010-01-18 17:05:40 +01002614 nopat [X86] Disable PAT (page attribute table extension of
2615 pagetables) support.
2616
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002617 norandmaps Don't use address space randomization. Equivalent to
2618 echo 0 > /proc/sys/kernel/randomize_va_space
2619
Jiri Kosinabbff2162010-08-11 18:34:09 +02002620 noreplace-paravirt [X86,IA-64,PV_OPS] Don't patch paravirt_ops
Jeremy Fitzhardinge959b4fd2007-05-02 19:27:16 +02002621
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002622 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
Jeremy Fitzhardingeb7fb4af2007-05-02 19:27:13 +02002623 with UP alternatives
2624
H. Peter Anvin7a5091d2014-05-11 20:25:20 -07002625 nordrand [X86] Disable kernel use of the RDRAND and
2626 RDSEED instructions even if they are supported
2627 by the processor. RDRAND and RDSEED are still
2628 available to user space applications.
H. Peter Anvin49d859d2011-07-31 14:02:19 -07002629
Randy Dunlapa9913042005-10-23 12:57:11 -07002630 noresume [SWSUSP] Disables resume and restores original swap
2631 space.
2632
Linus Torvalds1da177e2005-04-16 15:20:36 -07002633 no-scroll [VGA] Disables scrollback.
2634 This is required for the Braillex ib80-piezo Braille
2635 reader made by F.H. Papenmeier (Germany).
2636
2637 nosbagart [IA-64]
2638
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002639 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
Chuck Ebbert4f886512006-03-23 02:59:34 -08002640
Len Brown61ec7562007-08-16 03:34:22 -04002641 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2642 and disable the IO APIC. legacy for "maxcpus=0".
Linus Torvalds1da177e2005-04-16 15:20:36 -07002643
Dave Jones97842212007-07-15 23:41:05 -07002644 nosoftlockup [KNL] Disable the soft-lockup detector.
2645
Linus Torvalds1da177e2005-04-16 15:20:36 -07002646 nosync [HW,M68K] Disables sync negotiation for all devices.
2647
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002648 notsc [BUGS=X86-32] Disable Time Stamp Counter
Linus Torvalds1da177e2005-04-16 15:20:36 -07002649
Ulrich Obergfell195daf62015-04-14 15:44:13 -07002650 nowatchdog [KNL] Disable both lockup detectors, i.e.
2651 soft-lockup and NMI watchdog (hard-lockup).
Don Zickus58687ac2010-05-07 17:11:44 -04002652
Linus Torvalds1da177e2005-04-16 15:20:36 -07002653 nowb [ARM]
Randy Dunlapa9913042005-10-23 12:57:11 -07002654
Weidong Han2b2fd872009-04-17 16:42:12 +08002655 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2656
Fenghua Yuf78cff42012-11-13 11:32:38 -08002657 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2658 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2659 Some features depend on CPU0. Known dependencies are:
2660 1. Resume from suspend/hibernate depends on CPU0.
2661 Suspend/hibernate will fail if CPU0 is offline and you
2662 need to online CPU0 before suspend/hibernate.
2663 2. PIC interrupts also depend on CPU0. CPU0 can't be
2664 removed if a PIC interrupt is detected.
2665 It's said poweroff/reboot may depend on CPU0 on some
2666 machines although I haven't seen such issues so far
2667 after CPU0 is offline on a few tested machines.
2668 If the dependencies are under your control, you can
2669 turn on cpu0_hotplug.
2670
Randy Dunlap16290242011-08-13 12:34:52 -07002671 nptcg= [IA-64] Override max number of concurrent global TLB
Fenghua Yua6c75b862008-03-14 13:57:08 -07002672 purges which is reported from either PAL_VM_SUMMARY or
2673 SAL PALO.
2674
Yinghai Lu2b633e32010-02-10 01:20:37 -08002675 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2676 could support. nr_cpus=n : n >= 1 limits the kernel to
Baoquan He7c142bf2016-08-24 13:06:45 +08002677 support 'n' processors. It could be larger than the
2678 number of already plugged CPU during bootup, later in
2679 runtime you can physically add extra cpu until it reaches
2680 n. So during boot up some boot time memory for per-cpu
2681 variables need be pre-allocated for later physical cpu
2682 hot plugging.
Yinghai Lu2b633e32010-02-10 01:20:37 -08002683
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07002684 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2685
Mel Gorman1a687c22012-11-22 11:16:36 +00002686 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2687 Allowed values are enable and disable
2688
KAMEZAWA Hiroyukif0c0b2b2007-07-15 23:38:01 -07002689 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2690 one of ['zone', 'node', 'default'] can be specified
2691 This can be set from sysctl after boot.
2692 See Documentation/sysctl/vm.txt for details.
2693
Randy Dunlap7c4be252009-01-06 14:42:44 -08002694 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2695 See Documentation/debugging-via-ohci1394.txt for more
2696 info.
2697
Andres Salomon3ef0e1f2008-04-29 00:59:53 -07002698 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2699 Rather than timing out after 20 ms if an EC
2700 command is not properly ACKed, override the length
2701 of the timeout. We have interrupts disabled while
2702 waiting for the ACK, so if this is set too high
2703 interrupts *may* be lost!
2704
Tony Lindgren15ac7af2009-12-11 16:16:32 -08002705 omap_mux= [OMAP] Override bootloader pin multiplexing.
2706 Format: <mux_mode0.mode_name=value>...
2707 For example, to override I2C bus2:
2708 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2709
Linus Torvalds1da177e2005-04-16 15:20:36 -07002710 oprofile.timer= [HW]
2711 Use timer interrupt instead of performance counters
2712
Robert Richter7e4e0bd2009-05-06 12:10:23 +02002713 oprofile.cpu_type= Force an oprofile cpu type
2714 This might be useful if you have an older oprofile
2715 userland or if you want common events.
Robert Richter8d7ff4f2009-06-23 11:48:14 +02002716 Format: { arch_perfmon }
2717 arch_perfmon: [X86] Force use of architectural
Robert Richter7e4e0bd2009-05-06 12:10:23 +02002718 perfmon on Intel CPUs instead of the
2719 CPU specific event set.
Robert Richter159a80b2011-10-11 19:39:16 +02002720 timer: [X86] Force use of architectural NMI
2721 timer mode (see also oprofile.timer
2722 for generic hr timer mode)
Andi Kleen1dcdb5a2009-04-27 17:44:11 +02002723
Randy Dunlap44a4dcf2011-04-04 15:02:24 -07002724 oops=panic Always panic on oopses. Default is to just kill the
2725 process, but there is a small probability of
2726 deadlocking the machine.
Olaf Heringd404ab02011-03-22 16:34:04 -07002727 This will also cause panics on machine check exceptions.
2728 Useful together with panic=30 to trigger a reboot.
2729
Randy Dunlapbcfde332010-06-07 17:10:38 -07002730 OSS [HW,OSS]
2731 See Documentation/sound/oss/oss-parameters.txt
2732
Joonsoo Kim48c96a32014-12-12 16:56:01 -08002733 page_owner= [KNL] Boot-time page_owner enabling option.
2734 Storage of the information about who allocated
2735 each page is disabled in default. With this switch,
2736 we can turn it on.
2737 on: enable the feature
2738
Laura Abbott8823b1d2016-03-15 14:56:27 -07002739 page_poison= [KNL] Boot-time parameter changing the state of
2740 poisoning on the buddy allocator.
2741 off: turn off poisoning
2742 on: turn on poisoning
2743
Randy Dunlap44a4dcf2011-04-04 15:02:24 -07002744 panic= [KNL] Kernel behaviour on panic: delay <timeout>
Hugh Dickins4302fbc2011-07-26 16:08:52 -07002745 timeout > 0: seconds before rebooting
2746 timeout = 0: wait forever
2747 timeout < 0: reboot immediately
Linus Torvalds1da177e2005-04-16 15:20:36 -07002748 Format: <timeout>
2749
Prarit Bhargava9e3961a2014-12-10 15:45:50 -08002750 panic_on_warn panic() instead of WARN(). Useful to cause kdump
2751 on a WARN().
2752
Masami Hiramatsuf06e5152014-06-06 14:37:07 -07002753 crash_kexec_post_notifiers
2754 Run kdump after running panic-notifiers and dumping
2755 kmsg. This only for the users who doubt kdump always
2756 succeeds in any situation.
2757 Note that this also increases risks of kdump failure,
2758 because some panic notifiers can make the crashed
2759 kernel more unstable.
2760
Linus Torvalds1da177e2005-04-16 15:20:36 -07002761 parkbd.port= [HW] Parallel port number the keyboard adapter is
2762 connected to, default is 0.
2763 Format: <parport#>
2764 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2765 0 for XT, 1 for AT (default is AT).
Randy Dunlapa9913042005-10-23 12:57:11 -07002766 Format: <mode>
Linus Torvalds1da177e2005-04-16 15:20:36 -07002767
Randy Dunlapa9913042005-10-23 12:57:11 -07002768 parport= [HW,PPT] Specify parallel ports. 0 disables.
2769 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2770 Use 'auto' to force the driver to use any
2771 IRQ/DMA settings detected (the default is to
2772 ignore detected IRQ/DMA settings because of
2773 possible conflicts). You can specify the base
2774 address, IRQ, and DMA settings; IRQ and DMA
2775 should be numbers, or 'auto' (for using detected
2776 settings on that particular port), or 'nofifo'
2777 (to avoid using a FIFO even if it is detected).
2778 Parallel ports are assigned in the order they
2779 are specified on the command line, starting
2780 with parport0.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002781
Randy Dunlapa9913042005-10-23 12:57:11 -07002782 parport_init_mode= [HW,PPT]
2783 Configure VIA parallel port to operate in
2784 a specific mode. This is necessary on Pegasos
2785 computer where firmware has no options for setting
2786 up parallel port mode and sets it to spp.
2787 Currently this function knows 686a and 8231 chips.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002788 Format: [spp|ps2|epp|ecp|ecpepp]
2789
Andrew Mortondd287792006-03-23 03:00:57 -08002790 pause_on_oops=
2791 Halt all CPUs after the first oops has been printed for
2792 the specified number of seconds. This is to be used if
2793 your oopses keep scrolling off the screen.
2794
Linus Torvalds1da177e2005-04-16 15:20:36 -07002795 pcbit= [HW,ISDN]
2796
2797 pcd. [PARIDE]
2798 See header of drivers/block/paride/pcd.c.
Randy Dunlap31c00fc2008-11-13 21:33:24 +00002799 See also Documentation/blockdev/paride.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002800
Randy Dunlapa9913042005-10-23 12:57:11 -07002801 pci=option[,option...] [PCI] various PCI subsystem options:
Bjorn Helgaas1cc0ca22009-01-14 10:04:36 -07002802 earlydump [X86] dump PCI config space before the kernel
2803 changes anything
Andi Kleenc0115602008-08-22 09:53:39 +02002804 off [X86] don't probe for the PCI bus
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002805 bios [X86-32] force use of PCI BIOS, don't access
Randy Dunlapa9913042005-10-23 12:57:11 -07002806 the hardware directly. Use this if your machine
2807 has a non-standard PCI host bridge.
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002808 nobios [X86-32] disallow use of PCI BIOS, only direct
Randy Dunlapa9913042005-10-23 12:57:11 -07002809 hardware access methods are allowed. Use this
2810 if you experience crashes upon bootup and you
2811 suspect they are caused by the BIOS.
Borislav Petkovafd8c082016-01-13 16:48:51 +01002812 conf1 [X86] Force use of PCI Configuration Access
2813 Mechanism 1 (config address in IO port 0xCF8,
2814 data in IO port 0xCFC, both 32-bit).
2815 conf2 [X86] Force use of PCI Configuration Access
2816 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
2817 the function, IO port 0xCFA, also 8-bit, sets
2818 bus number. The config space is then accessed
2819 through ports 0xC000-0xCFFF).
2820 See http://wiki.osdev.org/PCI for more info
2821 on the configuration access mechanisms.
Randy Dunlap7f785762007-10-05 13:17:58 -07002822 noaer [PCIE] If the PCIEAER kernel config parameter is
2823 enabled, this kernel boot option can be used to
2824 disable the use of PCIE advanced error reporting.
Jeff Garzik32a2eea2007-10-11 16:57:27 -04002825 nodomains [PCI] Disable support for multiple PCI
2826 root domains (aka PCI segments, in ACPI-speak).
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05302827 nommconf [X86] Disable use of MMCONFIG for PCI
Bjorn Helgaas61be6d62006-02-15 15:17:43 -08002828 Configuration
Andreas Herrmann12983072009-06-07 16:15:16 +02002829 check_enable_amd_mmconf [X86] check for and enable
2830 properly configured MMIO access to PCI
2831 config space on AMD family 10h CPU
Matthew Wilcox309e57d2006-03-05 22:33:34 -07002832 nomsi [MSI] If the PCI_MSI kernel config parameter is
2833 enabled, this kernel boot option can be used to
2834 disable the use of MSI interrupts system-wide.
Stefan Assmanna9322f62008-06-11 16:35:14 +02002835 noioapicquirk [APIC] Disable all boot interrupt quirks.
2836 Safety option to keep boot IRQs enabled. This
2837 should never be necessary.
Stefan Assmann91979792008-06-11 16:35:15 +02002838 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
2839 primary IO-APIC for bridges that cannot disable
2840 boot IRQs. This fixes a source of spurious IRQs
2841 when the system masks IRQs.
Stefan Assmann41b9eb22008-07-15 13:48:55 +02002842 noioapicreroute [APIC] Disable workaround that uses the
2843 boot IRQ equivalent of an IRQ that connects to
2844 a chipset where boot IRQs cannot be disabled.
2845 The opposite of ioapicreroute.
Alan Coxcd4f0ef2007-07-31 00:37:59 -07002846 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
Randy Dunlapa9913042005-10-23 12:57:11 -07002847 routing table. These calls are known to be buggy
2848 on several machines and they hang the machine
2849 when used, but on other computers it's the only
2850 way to get the interrupt routing table. Try
2851 this option if the kernel is unable to allocate
2852 IRQs or discover secondary PCI buses on your
2853 motherboard.
Andi Kleenc0115602008-08-22 09:53:39 +02002854 rom [X86] Assign address space to expansion ROMs.
Randy Dunlapa9913042005-10-23 12:57:11 -07002855 Use with caution as certain devices share
2856 address decoders between ROMs and other
2857 resources.
Andi Kleenc0115602008-08-22 09:53:39 +02002858 norom [X86] Do not assign address space to
Gary Hadebb71ad82008-05-12 13:57:46 -07002859 expansion ROMs that do not already have
2860 BIOS assigned address ranges.
Mike Habeck7bd1c362010-05-12 11:14:32 -07002861 nobar [X86] Do not assign address space to the
2862 BARs that weren't assigned by the BIOS.
Andi Kleenc0115602008-08-22 09:53:39 +02002863 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
Randy Dunlapa9913042005-10-23 12:57:11 -07002864 assigned automatically to PCI devices. You can
2865 make the kernel exclude IRQs of your ISA cards
2866 this way.
Andi Kleenc0115602008-08-22 09:53:39 +02002867 pirqaddr=0xAAAAA [X86] Specify the physical address
Randy Dunlapa9913042005-10-23 12:57:11 -07002868 of the PIRQ table (normally generated
2869 by the BIOS) if it is outside the
2870 F0000h-100000h range.
Andi Kleenc0115602008-08-22 09:53:39 +02002871 lastbus=N [X86] Scan all buses thru bus #N. Can be
Randy Dunlapa9913042005-10-23 12:57:11 -07002872 useful if the kernel is unable to find your
2873 secondary buses and you want to tell it
2874 explicitly which ones they are.
Andi Kleenc0115602008-08-22 09:53:39 +02002875 assign-busses [X86] Always assign all PCI bus
Randy Dunlapa9913042005-10-23 12:57:11 -07002876 numbers ourselves, overriding
2877 whatever the firmware may have done.
Andi Kleenc0115602008-08-22 09:53:39 +02002878 usepirqmask [X86] Honor the possible IRQ mask stored
Randy Dunlapa9913042005-10-23 12:57:11 -07002879 in the BIOS $PIR table. This is needed on
2880 some systems with broken BIOSes, notably
2881 some HP Pavilion N5400 and Omnibook XE3
2882 notebooks. This will have no effect if ACPI
2883 IRQ routing is enabled.
Andi Kleenc0115602008-08-22 09:53:39 +02002884 noacpi [X86] Do not use ACPI for IRQ routing
Randy Dunlapa9913042005-10-23 12:57:11 -07002885 or for PCI scanning.
Bjorn Helgaas7bc5e3f2010-02-23 10:24:41 -07002886 use_crs [X86] Use PCI host bridge window information
2887 from ACPI. On BIOSes from 2008 or later, this
2888 is enabled by default. If you need to use this,
2889 please report a bug.
2890 nocrs [X86] Ignore PCI host bridge windows from ACPI.
2891 If you need to use this, please report a bug.
Randy Dunlapa9913042005-10-23 12:57:11 -07002892 routeirq Do IRQ routing for all PCI devices.
2893 This is normally done in pci_enable_device(),
2894 so this option is a temporary workaround
2895 for broken drivers that don't call it.
Yinghai Lu13a6ddb2008-03-27 01:31:18 -07002896 skip_isa_align [X86] do not align io start addr, so can
2897 handle more pci cards
Andi Kleen0637a702006-09-26 10:52:41 +02002898 noearly [X86] Don't do any early type 1 scanning.
2899 This might help on some broken boards which
2900 machine check when some devices' config space
2901 is read. But various workarounds are disabled
2902 and some IOMMU drivers will not work.
Matt Domsch6b4b78f2006-09-29 15:23:23 -05002903 bfsort Sort PCI devices into breadth-first order.
2904 This sorting is done to get a device
2905 order compatible with older (<= 2.4) kernels.
2906 nobfsort Don't sort PCI devices into breadth-first order.
Yijing Wangfa238712013-01-30 09:40:52 +08002907 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
2908 tuning and use the BIOS-configured MPS defaults.
2909 pcie_bus_safe Set every device's MPS to the largest value
2910 supported by all devices below the root complex.
2911 pcie_bus_perf Set device MPS to the largest allowable MPS
2912 based on its parent bus. Also set MRRS (Max
2913 Read Request Size) to the largest supported
2914 value (no larger than the MPS that the device
2915 or bus can support) for best performance.
2916 pcie_bus_peer2peer Set every device's MPS to 128B, which
2917 every device is guaranteed to support. This
2918 configuration allows peer-to-peer DMA between
2919 any pair of devices, possibly at the cost of
2920 reduced performance. This also guarantees
2921 that hot-added devices will work.
Atsushi Nemoto4516a612007-02-05 16:36:06 -08002922 cbiosize=nn[KMG] The fixed amount of bus space which is
2923 reserved for the CardBus bridge's IO window.
2924 The default value is 256 bytes.
2925 cbmemsize=nn[KMG] The fixed amount of bus space which is
2926 reserved for the CardBus bridge's memory
2927 window. The default value is 64 megabytes.
Yuji Shimada32a9a6822009-03-16 17:13:39 +09002928 resource_alignment=
2929 Format:
2930 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
Koehrer Mathias (ETAS/ESW5)644a5442016-06-07 14:24:17 +00002931 [<order of align>@]pci:<vendor>:<device>\
2932 [:<subvendor>:<subdevice>][; ...]
Yuji Shimada32a9a6822009-03-16 17:13:39 +09002933 Specifies alignment and device to reassign
2934 aligned memory resources.
2935 If <order of align> is not specified,
2936 PAGE_SIZE is used as alignment.
2937 PCI-PCI bridge can be specified, if resource
2938 windows need to be expanded.
Mathias Koehrer8b078c62016-08-09 10:33:31 +02002939 To specify the alignment for several
2940 instances of a device, the PCI vendor,
2941 device, subvendor, and subdevice may be
2942 specified, e.g., 4096@pci:8086:9c22:103c:198f
Andrew Patterson43c16402009-04-22 16:52:09 -06002943 ecrc= Enable/disable PCIe ECRC (transaction layer
2944 end-to-end CRC checking).
2945 bios: Use BIOS/firmware settings. This is the
2946 the default.
2947 off: Turn ECRC off
2948 on: Turn ECRC on.
Yijing Wang8c8803c2013-01-23 20:29:06 +08002949 hpiosize=nn[KMG] The fixed amount of bus space which is
2950 reserved for hotplug bridge's IO window.
2951 Default size is 256 bytes.
2952 hpmemsize=nn[KMG] The fixed amount of bus space which is
2953 reserved for hotplug bridge's memory window.
2954 Default size is 2 megabytes.
Keith Busche16b4662016-07-21 21:40:28 -06002955 hpbussize=nn The minimum amount of additional bus numbers
2956 reserved for buses below a hotplug bridge.
2957 Default is 1.
Yinghai Lub55438f2012-02-23 19:23:30 -08002958 realloc= Enable/disable reallocating PCI bridge resources
2959 if allocations done by BIOS are too small to
2960 accommodate resources required by all child
2961 devices.
2962 off: Turn realloc off
2963 on: Turn realloc on
2964 realloc same as realloc=on
Rafael J. Wysocki6748dcc2012-03-01 00:06:33 +01002965 noari do not use PCIe ARI.
Bjorn Helgaas284f5f92012-04-30 15:21:02 -06002966 pcie_scan_all Scan all possible PCIe devices. Otherwise we
2967 only look for one device below a PCIe downstream
2968 port.
Matt Domsch6b4b78f2006-09-29 15:23:23 -05002969
Chuck Ebberte5665a42008-09-24 20:40:34 -04002970 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
2971 Management.
2972 off Disable ASPM.
2973 force Enable ASPM even on devices that claim not to support it.
2974 WARNING: Forcing ASPM on may cause system lockups.
2975
MUNEDA Takahiro7570a332012-02-02 11:09:22 -05002976 pcie_hp= [PCIE] PCI Express Hotplug driver options:
2977 nomsi Do not use MSI for PCI Express Native Hotplug (this
2978 makes all PCIe ports use INTx for hotplug services).
2979
Rafael J. Wysocki79dd9182010-08-21 01:51:44 +02002980 pcie_ports= [PCIE] PCIe ports handling:
Rafael J. Wysocki28eb5f22010-08-21 22:02:38 +02002981 auto Ask the BIOS whether or not to use native PCIe services
2982 associated with PCIe ports (PME, hot-plug, AER). Use
2983 them only if that is allowed by the BIOS.
2984 native Use native PCIe services associated with PCIe ports
2985 unconditionally.
Rafael J. Wysocki79dd9182010-08-21 01:51:44 +02002986 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
2987 ports driver.
2988
Mika Westerberg9d26d3a2016-06-02 11:17:12 +03002989 pcie_port_pm= [PCIE] PCIe port power management handling:
2990 off Disable power management of all PCIe ports
2991 force Forcibly enable power management of all PCIe ports
2992
Rafael J. Wysockic7f48652010-02-17 23:39:08 +01002993 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
Rafael J. Wysockic39fae12010-02-17 23:40:07 +01002994 nomsi Do not use MSI for native PCIe PME signaling (this makes
Rafael J. Wysocki28eb5f22010-08-21 22:02:38 +02002995 all PCIe root ports use INTx for all services).
Rafael J. Wysockic7f48652010-02-17 23:39:08 +01002996
Linus Torvalds1da177e2005-04-16 15:20:36 -07002997 pcmv= [HW,PCMCIA] BadgePAD 4
2998
Tushar Behera39ac5ba2014-03-28 10:50:21 +05302999 pd_ignore_unused
3000 [PM]
3001 Keep all power-domains already enabled by bootloader on,
3002 even if no driver has claimed them. This is useful
3003 for debug and development, but should not be
3004 needed on a platform with proper driver support.
3005
Linus Torvalds1da177e2005-04-16 15:20:36 -07003006 pd. [PARIDE]
Randy Dunlap31c00fc2008-11-13 21:33:24 +00003007 See Documentation/blockdev/paride.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003008
3009 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3010 boot time.
3011 Format: { 0 | 1 }
3012 See arch/parisc/kernel/pdc_chassis.c
3013
Tejun Heof58dc012009-08-14 15:00:50 +09003014 percpu_alloc= Select which percpu first chunk allocator to use.
Tejun Heoe933a732009-08-14 15:00:53 +09003015 Currently supported values are "embed" and "page".
3016 Archs may support subset or none of the selections.
3017 See comments in mm/percpu.c for details on each
3018 allocator. This parameter is primarily for debugging
3019 and performance comparison.
Tejun Heofa8a7092009-06-22 11:56:24 +09003020
Linus Torvalds1da177e2005-04-16 15:20:36 -07003021 pf. [PARIDE]
Randy Dunlap31c00fc2008-11-13 21:33:24 +00003022 See Documentation/blockdev/paride.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003023
3024 pg. [PARIDE]
Randy Dunlap31c00fc2008-11-13 21:33:24 +00003025 See Documentation/blockdev/paride.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003026
3027 pirq= [SMP,APIC] Manual mp-table setup
Uwe Hermann71cced62008-10-20 09:32:21 -07003028 See Documentation/x86/i386/IO-APIC.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003029
3030 plip= [PPT,NET] Parallel port network link
3031 Format: { parport<nr> | timid | 0 }
3032 See also Documentation/parport.txt.
3033
Randy Dunlap16290242011-08-13 12:34:52 -07003034 pmtmr= [X86] Manual setup of pmtmr I/O Port.
Thomas Gleixnerde32a242008-07-12 05:33:30 +02003035 Override pmtimer IOPort with a hex value.
3036 e.g. pmtmr=0x508
3037
Bjorn Helgaas96242112011-08-11 12:14:05 -06003038 pnp.debug=1 [PNP]
3039 Enable PNP debug messages (depends on the
3040 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3041 via /sys/module/pnp/parameters/debug. We always show
3042 current resource usage; turning this on also shows
3043 possible settings and some assignment information.
Bjorn Helgaas97ef0622008-08-19 16:53:41 -06003044
Linus Torvalds1da177e2005-04-16 15:20:36 -07003045 pnpacpi= [ACPI]
3046 { off }
3047
3048 pnpbios= [ISAPNP]
3049 { on | off | curr | res | no-curr | no-res }
3050
3051 pnp_reserve_irq=
3052 [ISAPNP] Exclude IRQs for the autoconfiguration
3053
3054 pnp_reserve_dma=
3055 [ISAPNP] Exclude DMAs for the autoconfiguration
3056
3057 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
Randy Dunlapa9913042005-10-23 12:57:11 -07003058 Ranges are in pairs (I/O port base and size).
Linus Torvalds1da177e2005-04-16 15:20:36 -07003059
3060 pnp_reserve_mem=
Randy Dunlapa9913042005-10-23 12:57:11 -07003061 [ISAPNP] Exclude memory regions for the
3062 autoconfiguration.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003063 Ranges are in pairs (memory base and size).
3064
Randy Dunlap4af94f32009-04-17 18:30:28 -07003065 ports= [IP_VS_FTP] IPVS ftp helper module
3066 Default is 21.
3067 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3068 may be specified.
3069 Format: <port>,<port>....
3070
Balbir Singhc3cbd072016-12-02 00:08:26 +11003071 powersave=off [PPC] This option disables power saving features.
3072 It specifically disables cpuidle and sets the
3073 platform machine description specific power_save
3074 function to NULL. On Idle the CPU just reduces
3075 execution priority.
3076
Anton Blanchard3eb5d582015-10-29 11:44:06 +11003077 ppc_strict_facility_enable
3078 [PPC] This option catches any kernel floating point,
3079 Altivec, VSX and SPE outside of regions specifically
3080 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3081 There is some performance impact when enabling this.
3082
Ingo Molnar45807a12007-07-15 23:40:10 -07003083 print-fatal-signals=
3084 [KNL] debug: print fatal signals
Naohiro Ooiwaf84d49b2009-11-09 00:46:42 +09003085
3086 If enabled, warn about various signal handling
3087 related application anomalies: too many signals,
3088 too many POSIX.1 timers, fatal signals causing a
3089 coredump - etc.
3090
3091 If you hit the warning due to signal overflow,
3092 you might want to try "ulimit -i unlimited".
3093
Ingo Molnar45807a12007-07-15 23:40:10 -07003094 default: off.
3095
Matthew Garrettc22ab3322012-03-05 14:59:10 -08003096 printk.always_kmsg_dump=
3097 Trigger kmsg_dump for cases other than kernel oops or
3098 panics
3099 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3100 default: disabled
3101
Borislav Petkov750afe72016-08-02 14:04:07 -07003102 printk.devkmsg={on,off,ratelimit}
3103 Control writing to /dev/kmsg.
3104 on - unlimited logging to /dev/kmsg from userspace
3105 off - logging to /dev/kmsg disabled
3106 ratelimit - ratelimit the logging
3107 Default: ratelimit
3108
Randy Dunlape84845c2007-07-15 23:40:25 -07003109 printk.time= Show timing data prefixed to each printk message line
3110 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3111
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07003112 processor.max_cstate= [HW,ACPI]
3113 Limit processor to maximum C-state
3114 max_cstate=9 overrides any DMI blacklist limit.
3115
3116 processor.nocst [HW,ACPI]
3117 Ignore the _CST method to determine C-states,
3118 instead using the legacy FADT method
3119
Linus Torvalds1da177e2005-04-16 15:20:36 -07003120 profile= [KNL] Enable kernel profiling via /proc/profile
Randy Dunlapa9913042005-10-23 12:57:11 -07003121 Format: [schedule,]<number>
3122 Param: "schedule" - profile schedule points.
3123 Param: <number> - step/bucket size as a power of 2 for
3124 statistical time based profiling.
Mel Gormanb3da2a72007-10-24 18:23:50 +02003125 Param: "sleep" - profile D-state sleeping (millisecs).
3126 Requires CONFIG_SCHEDSTATS
Dave Jonesc0fe2e62007-10-20 03:08:22 +02003127 Param: "kvm" - profile VM exits.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003128
Linus Torvalds1da177e2005-04-16 15:20:36 -07003129 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3130 before loading.
Randy Dunlap31c00fc2008-11-13 21:33:24 +00003131 See Documentation/blockdev/ramdisk.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003132
Randy Dunlapa9913042005-10-23 12:57:11 -07003133 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3134 probe for; one of (bare|imps|exps|lifebook|any).
Linus Torvalds1da177e2005-04-16 15:20:36 -07003135 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3136 per second.
Randy Dunlapa9913042005-10-23 12:57:11 -07003137 psmouse.resetafter= [HW,MOUSE]
3138 Try to reset the device after so many bad packets
Linus Torvalds1da177e2005-04-16 15:20:36 -07003139 (0 = never).
3140 psmouse.resolution=
3141 [HW,MOUSE] Set desired mouse resolution, in dpi.
3142 psmouse.smartscroll=
Randy Dunlapa9913042005-10-23 12:57:11 -07003143 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003144 0 = disabled, 1 = enabled (default).
3145
Matthew Garrettdee28e72011-07-21 16:57:55 -04003146 pstore.backend= Specify the name of the pstore backend to use
3147
Linus Torvalds1da177e2005-04-16 15:20:36 -07003148 pt. [PARIDE]
Randy Dunlap31c00fc2008-11-13 21:33:24 +00003149 See Documentation/blockdev/paride.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003150
Kay Sieversdc8c8582007-08-15 12:25:38 +02003151 pty.legacy_count=
3152 [KNL] Number of legacy pty's. Overwrites compiled-in
3153 default number.
3154
Randy Dunlap7d2c5022006-09-29 02:01:02 -07003155 quiet [KNL] Disable most log messages
Randy Dunlapa9913042005-10-23 12:57:11 -07003156
Linus Torvalds1da177e2005-04-16 15:20:36 -07003157 r128= [HW,DRM]
3158
3159 raid= [HW,RAID]
Jani Nikulae52347b2016-11-03 12:10:10 +02003160 See Documentation/admin-guide/md.rst.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003161
Linus Torvalds1da177e2005-04-16 15:20:36 -07003162 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
Randy Dunlap31c00fc2008-11-13 21:33:24 +00003163 See Documentation/blockdev/ramdisk.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003164
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003165 rcu_nocbs= [KNL]
Noam Camus2d13e6c2016-10-11 13:51:35 -07003166 The argument is a cpu list, as described above.
3167
Paul E. McKenney3fbfbf72012-08-19 21:35:53 -07003168 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3169 the specified list of CPUs to be no-callback CPUs.
3170 Invocation of these CPUs' RCU callbacks will
Paul E. McKenneya4889852012-12-03 08:16:28 -08003171 be offloaded to "rcuox/N" kthreads created for
3172 that purpose, where "x" is "b" for RCU-bh, "p"
3173 for RCU-preempt, and "s" for RCU-sched, and "N"
3174 is the CPU number. This reduces OS jitter on the
Paul E. McKenney3fbfbf72012-08-19 21:35:53 -07003175 offloaded CPUs, which can be useful for HPC and
3176 real-time workloads. It can also improve energy
3177 efficiency for asymmetric multiprocessors.
3178
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003179 rcu_nocb_poll [KNL]
Paul E. McKenney3fbfbf72012-08-19 21:35:53 -07003180 Rather than requiring that offloaded CPUs
3181 (specified by rcu_nocbs= above) explicitly
3182 awaken the corresponding "rcuoN" kthreads,
3183 make these kthreads poll for callbacks.
3184 This improves the real-time response for the
3185 offloaded CPUs by relieving them of the need to
3186 wake up the corresponding kthread, but degrades
3187 energy efficiency by requiring that the kthreads
3188 periodically wake up to do the polling.
3189
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003190 rcutree.blimit= [KNL]
Paul E. McKenney97e63f02013-10-27 09:44:03 -07003191 Set maximum number of finished RCU callbacks to
3192 process in one batch.
Dipankar Sarma21a1ea92006-03-07 21:55:33 -08003193
Paul E. McKenneya3dc2942015-04-20 11:40:50 -07003194 rcutree.dump_tree= [KNL]
3195 Dump the structure of the rcu_node combining tree
3196 out at early boot. This is used for diagnostic
3197 purposes, to verify correct tree setup.
3198
Paul E. McKenney0f41c0d2015-03-10 18:33:20 -07003199 rcutree.gp_cleanup_delay= [KNL]
3200 Set the number of jiffies to delay each step of
3201 RCU grace-period cleanup. This only has effect
3202 when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3203
Paul E. McKenney37745d22015-01-22 18:24:08 -08003204 rcutree.gp_init_delay= [KNL]
3205 Set the number of jiffies to delay each step of
3206 RCU grace-period initialization. This only has
Paul E. McKenney0f41c0d2015-03-10 18:33:20 -07003207 effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3208 is set.
3209
3210 rcutree.gp_preinit_delay= [KNL]
3211 Set the number of jiffies to delay each step of
3212 RCU grace-period pre-initialization, that is,
3213 the propagation of recent CPU-hotplug changes up
3214 the rcu_node combining tree. This only has effect
3215 when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
Paul E. McKenney37745d22015-01-22 18:24:08 -08003216
Paul E. McKenney7fa27002015-04-20 10:27:15 -07003217 rcutree.rcu_fanout_exact= [KNL]
3218 Disable autobalancing of the rcu_node combining
3219 tree. This is used by rcutorture, and might
3220 possibly be useful for architectures having high
3221 cache-to-cache transfer latencies.
Dipankar Sarma21a1ea92006-03-07 21:55:33 -08003222
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003223 rcutree.rcu_fanout_leaf= [KNL]
Paul E. McKenneyee968ac2015-07-31 08:28:35 -07003224 Change the number of CPUs assigned to each
3225 leaf rcu_node structure. Useful for very
3226 large systems, which will choose the value 64,
3227 and for NUMA systems with large remote-access
3228 latencies, which will choose a value aligned
3229 with the appropriate hardware boundaries.
Paul E. McKenneyf885b7f2012-04-23 15:52:53 -07003230
Paul E. McKenney4a81e832014-06-20 16:49:01 -07003231 rcutree.jiffies_till_sched_qs= [KNL]
3232 Set required age in jiffies for a
3233 given grace period before RCU starts
3234 soliciting quiescent-state help from
3235 rcu_note_context_switch().
3236
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003237 rcutree.jiffies_till_first_fqs= [KNL]
Paul E. McKenneyc0f4dfd2012-12-28 11:30:36 -08003238 Set delay from grace-period initialization to
3239 first attempt to force quiescent states.
3240 Units are jiffies, minimum value is zero,
3241 and maximum value is HZ.
3242
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003243 rcutree.jiffies_till_next_fqs= [KNL]
Paul E. McKenneyc0f4dfd2012-12-28 11:30:36 -08003244 Set delay between subsequent attempts to force
3245 quiescent states. Units are jiffies, minimum
3246 value is one, and maximum value is HZ.
3247
Clark Williams21871d72014-09-12 21:21:09 -05003248 rcutree.kthread_prio= [KNL,BOOT]
Paul E. McKenneyd2af1ad2015-01-20 23:54:59 -08003249 Set the SCHED_FIFO priority of the RCU per-CPU
3250 kthreads (rcuc/N). This value is also used for
3251 the priority of the RCU boost threads (rcub/N)
3252 and for the RCU grace-period kthreads (rcu_bh,
3253 rcu_preempt, and rcu_sched). If RCU_BOOST is
3254 set, valid values are 1-99 and the default is 1
3255 (the least-favored priority). Otherwise, when
3256 RCU_BOOST is not set, valid values are 0-99 and
3257 the default is zero (non-realtime operation).
Clark Williams21871d72014-09-12 21:21:09 -05003258
Paul E. McKenneyfbce7492014-06-24 09:26:11 -07003259 rcutree.rcu_nocb_leader_stride= [KNL]
3260 Set the number of NOCB kthread groups, which
3261 defaults to the square root of the number of
3262 CPUs. Larger numbers reduces the wakeup overhead
3263 on the per-CPU grace-period kthreads, but increases
3264 that same overhead on each group's leader.
3265
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003266 rcutree.qhimark= [KNL]
Paul E. McKenney97e63f02013-10-27 09:44:03 -07003267 Set threshold of queued RCU callbacks beyond which
3268 batch limiting is disabled.
Dipankar Sarma21a1ea92006-03-07 21:55:33 -08003269
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003270 rcutree.qlowmark= [KNL]
Robert P. J. Day24aaef82008-02-03 15:20:26 +02003271 Set threshold of queued RCU callbacks below which
3272 batch limiting is re-enabled.
Dipankar Sarma21a1ea92006-03-07 21:55:33 -08003273
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003274 rcutree.rcu_idle_gp_delay= [KNL]
Paul E. McKenneyc0f4dfd2012-12-28 11:30:36 -08003275 Set wakeup interval for idle CPUs that have
3276 RCU callbacks (RCU_FAST_NO_HZ=y).
Paul E. McKenneyd40011f2012-06-26 20:45:57 -07003277
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003278 rcutree.rcu_idle_lazy_gp_delay= [KNL]
Paul E. McKenneyc0f4dfd2012-12-28 11:30:36 -08003279 Set wakeup interval for idle CPUs that have
3280 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3281 Lazy RCU callbacks are those which RCU can
3282 prove do nothing more than free memory.
Paul E. McKenneyd40011f2012-06-26 20:45:57 -07003283
Paul E. McKenneybdea9e32016-01-01 13:47:19 -08003284 rcuperf.gp_exp= [KNL]
3285 Measure performance of expedited synchronous
3286 grace-period primitives.
3287
Paul E. McKenneydf37e662016-01-30 20:56:38 -08003288 rcuperf.holdoff= [KNL]
3289 Set test-start holdoff period. The purpose of
3290 this parameter is to delay the start of the
3291 test until boot completes in order to avoid
3292 interference.
3293
Paul E. McKenneybdea9e32016-01-01 13:47:19 -08003294 rcuperf.nreaders= [KNL]
3295 Set number of RCU readers. The value -1 selects
3296 N, where N is the number of CPUs. A value
3297 "n" less than -1 selects N-n+1, where N is again
3298 the number of CPUs. For example, -2 selects N
3299 (the number of CPUs), -3 selects N+1, and so on.
3300 A value of "n" less than or equal to -N selects
3301 a single reader.
3302
3303 rcuperf.nwriters= [KNL]
3304 Set number of RCU writers. The values operate
3305 the same as for rcuperf.nreaders.
3306 N, where N is the number of CPUs
3307
3308 rcuperf.perf_runnable= [BOOT]
3309 Start rcuperf running at boot time.
3310
3311 rcuperf.shutdown= [KNL]
3312 Shut the system down after performance tests
3313 complete. This is useful for hands-off automated
3314 testing.
3315
3316 rcuperf.perf_type= [KNL]
3317 Specify the RCU implementation to test.
3318
3319 rcuperf.verbose= [KNL]
3320 Enable additional printk() statements.
3321
Paul E. McKenney38706bc2014-08-18 21:12:17 -07003322 rcutorture.cbflood_inter_holdoff= [KNL]
3323 Set holdoff time (jiffies) between successive
3324 callback-flood tests.
3325
3326 rcutorture.cbflood_intra_holdoff= [KNL]
3327 Set holdoff time (jiffies) between successive
3328 bursts of callbacks within a given callback-flood
3329 test.
3330
3331 rcutorture.cbflood_n_burst= [KNL]
3332 Set the number of bursts making up a given
3333 callback-flood test. Set this to zero to
3334 disable callback-flood testing.
3335
3336 rcutorture.cbflood_n_per_burst= [KNL]
3337 Set the number of callbacks to be registered
3338 in a given burst of a callback-flood test.
3339
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003340 rcutorture.fqs_duration= [KNL]
Paul E. McKenney21b05de2015-05-14 17:29:51 -07003341 Set duration of force_quiescent_state bursts
3342 in microseconds.
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003343
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003344 rcutorture.fqs_holdoff= [KNL]
Paul E. McKenney21b05de2015-05-14 17:29:51 -07003345 Set holdoff time within force_quiescent_state bursts
3346 in microseconds.
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003347
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003348 rcutorture.fqs_stutter= [KNL]
Paul E. McKenney21b05de2015-05-14 17:29:51 -07003349 Set wait time between force_quiescent_state bursts
3350 in seconds.
3351
3352 rcutorture.gp_cond= [KNL]
3353 Use conditional/asynchronous update-side
3354 primitives, if available.
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003355
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003356 rcutorture.gp_exp= [KNL]
Paul E. McKenney21b05de2015-05-14 17:29:51 -07003357 Use expedited update-side primitives, if available.
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003358
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003359 rcutorture.gp_normal= [KNL]
Paul E. McKenney21b05de2015-05-14 17:29:51 -07003360 Use normal (non-expedited) asynchronous
3361 update-side primitives, if available.
3362
3363 rcutorture.gp_sync= [KNL]
3364 Use normal (non-expedited) synchronous
3365 update-side primitives, if available. If all
3366 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3367 rcutorture.gp_normal=, and rcutorture.gp_sync=
3368 are zero, rcutorture acts as if is interpreted
3369 they are all non-zero.
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003370
3371 rcutorture.n_barrier_cbs= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003372 Set callbacks/threads for rcu_barrier() testing.
3373
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003374 rcutorture.nfakewriters= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003375 Set number of concurrent RCU writers. These just
3376 stress RCU, they don't participate in the actual
3377 test, hence the "fake".
3378
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003379 rcutorture.nreaders= [KNL]
Paul E. McKenney3838cc12015-03-12 13:55:48 -07003380 Set number of RCU readers. The value -1 selects
3381 N-1, where N is the number of CPUs. A value
3382 "n" less than -1 selects N-n-2, where N is again
3383 the number of CPUs. For example, -2 selects N
3384 (the number of CPUs), -3 selects N+1, and so on.
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003385
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003386 rcutorture.object_debug= [KNL]
3387 Enable debug-object double-call_rcu() testing.
3388
3389 rcutorture.onoff_holdoff= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003390 Set time (s) after boot for CPU-hotplug testing.
3391
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003392 rcutorture.onoff_interval= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003393 Set time (s) between CPU-hotplug operations, or
3394 zero to disable CPU-hotplug testing.
3395
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003396 rcutorture.shuffle_interval= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003397 Set task-shuffle interval (s). Shuffling tasks
3398 allows some CPUs to go into dyntick-idle mode
3399 during the rcutorture test.
3400
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003401 rcutorture.shutdown_secs= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003402 Set time (s) after boot system shutdown. This
3403 is useful for hands-off automated testing.
3404
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003405 rcutorture.stall_cpu= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003406 Duration of CPU stall (s) to test RCU CPU stall
3407 warnings, zero to disable.
3408
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003409 rcutorture.stall_cpu_holdoff= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003410 Time to wait (s) after boot before inducing stall.
3411
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003412 rcutorture.stat_interval= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003413 Time (s) between statistics printk()s.
3414
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003415 rcutorture.stutter= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003416 Time (s) to stutter testing, for example, specifying
3417 five seconds causes the test to run for five seconds,
3418 wait for five seconds, and so on. This tests RCU's
3419 ability to transition abruptly to and from idle.
3420
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003421 rcutorture.test_boost= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003422 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3423 "Maybe" means test if the RCU implementation
3424 under test support RCU priority boosting.
3425
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003426 rcutorture.test_boost_duration= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003427 Duration (s) of each individual boost test.
3428
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003429 rcutorture.test_boost_interval= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003430 Interval (s) between each boost test.
3431
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003432 rcutorture.test_no_idle_hz= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003433 Test RCU's dyntick-idle handling. See also the
3434 rcutorture.shuffle_interval parameter.
3435
Paul E. McKenney21b05de2015-05-14 17:29:51 -07003436 rcutorture.torture_runnable= [BOOT]
3437 Start rcutorture running at boot time.
3438
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003439 rcutorture.torture_type= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003440 Specify the RCU implementation to test.
3441
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003442 rcutorture.verbose= [KNL]
Paul E. McKenneydabb8aa2012-04-23 10:54:45 -07003443 Enable additional printk() statements.
3444
Paul E. McKenney5a9be7c2015-11-24 15:44:06 -08003445 rcupdate.rcu_cpu_stall_suppress= [KNL]
3446 Suppress RCU CPU stall warning messages.
3447
3448 rcupdate.rcu_cpu_stall_timeout= [KNL]
3449 Set timeout for RCU CPU stall warning messages.
3450
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003451 rcupdate.rcu_expedited= [KNL]
3452 Use expedited grace-period primitives, for
3453 example, synchronize_rcu_expedited() instead
3454 of synchronize_rcu(). This reduces latency,
3455 but can increase CPU utilization, degrade
3456 real-time latency, and degrade energy efficiency.
Paul E. McKenney79cfea022015-12-07 13:09:52 -08003457 No effect on CONFIG_TINY_RCU kernels.
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003458
Paul E. McKenney5a9be7c2015-11-24 15:44:06 -08003459 rcupdate.rcu_normal= [KNL]
3460 Use only normal grace-period primitives,
3461 for example, synchronize_rcu() instead of
3462 synchronize_rcu_expedited(). This improves
Paul E. McKenney79cfea022015-12-07 13:09:52 -08003463 real-time latency, CPU utilization, and
3464 energy efficiency, but can expose users to
3465 increased grace-period latency. This parameter
3466 overrides rcupdate.rcu_expedited. No effect on
3467 CONFIG_TINY_RCU kernels.
Paul E. McKenney4102ada2013-10-08 20:23:47 -07003468
Paul E. McKenney3e42ec12015-11-25 18:56:00 -08003469 rcupdate.rcu_normal_after_boot= [KNL]
3470 Once boot has completed (that is, after
3471 rcu_end_inkernel_boot() has been invoked), use
Paul E. McKenney79cfea022015-12-07 13:09:52 -08003472 only normal grace-period primitives. No effect
3473 on CONFIG_TINY_RCU kernels.
Paul E. McKenney3e42ec12015-11-25 18:56:00 -08003474
Paul E. McKenney52db30a2014-07-01 18:16:30 -07003475 rcupdate.rcu_task_stall_timeout= [KNL]
3476 Set timeout in jiffies for RCU task stall warning
3477 messages. Disable with a value less than or equal
3478 to zero.
3479
Pranith Kumar74860fe2014-09-19 11:34:09 -04003480 rcupdate.rcu_self_test= [KNL]
3481 Run the RCU early boot self tests
3482
3483 rcupdate.rcu_self_test_bh= [KNL]
3484 Run the RCU bh early boot self tests
3485
3486 rcupdate.rcu_self_test_sched= [KNL]
3487 Run the RCU sched early boot self tests
3488
Olof Johanssonffdfc402005-09-06 15:17:19 -07003489 rdinit= [KNL]
3490 Format: <full_path>
3491 Run specified binary instead of /init from the ramdisk,
3492 used for early userspace startup. See initrd.
3493
Robin Holt1b3a5d02013-07-08 16:01:42 -07003494 reboot= [KNL]
3495 Format (x86 or x86_64):
3496 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3497 [[,]s[mp]#### \
3498 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3499 [[,]f[orce]
3500 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3501 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3502 reboot_force is either force or not specified,
3503 reboot_cpu is s[mp]#### with #### being the processor
3504 to be used for rebooting.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003505
Paul Jackson46b6d942008-07-04 10:00:09 -07003506 relax_domain_level=
3507 [KNL, SMP] Set scheduler's default relax_domain_level.
seokhoon.yoon09c3bcc2016-08-02 23:23:57 +09003508 See Documentation/cgroup-v1/cpusets.txt.
Paul Jackson46b6d942008-07-04 10:00:09 -07003509
Linus Torvalds1da177e2005-04-16 15:20:36 -07003510 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
3511
Alan Coxcd4f0ef2007-07-31 00:37:59 -07003512 reservetop= [X86-32]
Zachary Amsden461a9aff2006-09-25 23:32:25 -07003513 Format: nn[KMG]
3514 Reserves a hole at the top of the kernel virtual
3515 address space.
3516
H. Peter Anvin9ea77bd2010-08-25 16:38:20 -07003517 reservelow= [X86]
3518 Format: nn[K]
3519 Set the amount of memory to reserve for BIOS at
3520 the bottom of the address space.
3521
Vivek Goyal7e962872006-09-27 01:50:44 -07003522 reset_devices [KNL] Force drivers to reset the underlying device
3523 during initialization.
3524
Randy Dunlapa9913042005-10-23 12:57:11 -07003525 resume= [SWSUSP]
3526 Specify the partition device for software suspend
Minho Ban2df83fa2012-05-14 21:45:31 +02003527 Format:
3528 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
Linus Torvalds1da177e2005-04-16 15:20:36 -07003529
Rafael J. Wysockiecbd0da2006-12-06 20:34:13 -08003530 resume_offset= [SWSUSP]
3531 Specify the offset from the beginning of the partition
3532 given by "resume=" at which the swap header is located,
3533 in <PAGE_SIZE> units (needed only for swap files).
3534 See Documentation/power/swsusp-and-swap-files.txt
3535
Barry Songf126f732011-10-10 23:38:41 +02003536 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3537 read the resume files
3538
Barry Song6f8d7022011-10-06 20:34:46 +02003539 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3540 Useful for devices that are detected asynchronously
3541 (e.g. USB and MMC devices).
3542
Bojan Smojverf996fc92010-09-09 23:06:23 +02003543 hibernate= [HIBERNATION]
3544 noresume Don't check if there's a hibernation image
3545 present during boot.
3546 nocompress Don't compress/decompress hibernation images.
Kees Cooka6e15a32014-06-13 13:30:35 -07003547 no Disable hibernation and resume.
Rafael J. Wysocki4c0b6c12016-07-10 02:12:10 +02003548 protect_image Turn on image protection during restoration
3549 (that will set all pages holding image data
3550 during restoration read-only).
Bojan Smojverf996fc92010-09-09 23:06:23 +02003551
Michael Neuling0a7b35cb2007-02-10 01:44:33 -08003552 retain_initrd [RAM] Keep initrd memory after extraction
3553
Andrew Clausen0efbb782015-01-09 20:24:55 +00003554 rfkill.default_state=
3555 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3556 etc. communication is blocked by default.
3557 1 Unblocked.
3558
3559 rfkill.master_switch_mode=
3560 0 The "airplane mode" button does nothing.
3561 1 The "airplane mode" button toggles between everything
3562 blocked and the previous configuration.
3563 2 The "airplane mode" button toggles between everything
3564 blocked and everything unblocked.
3565
Linus Torvalds1da177e2005-04-16 15:20:36 -07003566 rhash_entries= [KNL,NET]
3567 Set number of hash buckets for route cache
3568
Linus Torvalds1da177e2005-04-16 15:20:36 -07003569 ro [KNL] Mount root device read-only on boot
3570
Kees Cookd2aa1ac2016-02-17 14:41:13 -08003571 rodata= [KNL]
3572 on Mark read-only kernel memory as read-only (default).
3573 off Leave read-only kernel memory writable for debugging.
3574
Heiko Stuebner605df8a2016-02-22 12:55:01 +01003575 rockchip.usb_uart
3576 Enable the uart passthrough on the designated usb port
3577 on Rockchip SoCs. When active, the signals of the
3578 debug-uart get routed to the D+ and D- pins of the usb
3579 port and the regular usb controller gets disabled.
3580
Linus Torvalds1da177e2005-04-16 15:20:36 -07003581 root= [KNL] Root filesystem
Will Drewryf2d34fd92011-08-03 16:21:08 -07003582 See name_to_dev_t comment in init/do_mounts.c.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003583
3584 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
3585 mount the root filesystem
3586
3587 rootflags= [KNL] Set root filesystem mount option string
3588
3589 rootfstype= [KNL] Set root filesystem type
3590
Pierre Ossmancc1ed752007-07-15 23:40:35 -07003591 rootwait [KNL] Wait (indefinitely) for root device to show up.
3592 Useful for devices that are detected asynchronously
3593 (e.g. USB and MMC devices).
3594
Robert Tivy5c71d612013-03-28 18:41:46 -07003595 rproc_mem=nn[KMG][@address]
3596 [KNL,ARM,CMA] Remoteproc physical memory block.
3597 Memory area to be used by remote processor image,
3598 managed by CMA.
3599
Linus Torvalds1da177e2005-04-16 15:20:36 -07003600 rw [KNL] Mount root device read-write on boot
3601
3602 S [KNL] Run init in single mode
3603
Gerald Schaeferc60d1ae2014-07-18 17:37:08 +02003604 s390_iommu= [HW,S390]
3605 Set s390 IOTLB flushing mode
3606 strict
3607 With strict flushing every unmap operation will result in
3608 an IOTLB flush. Default is lazy flushing before reuse,
3609 which is faster.
3610
Linus Torvalds1da177e2005-04-16 15:20:36 -07003611 sa1100ir [NET]
3612 See drivers/net/irda/sa1100_ir.c.
3613
Linus Torvalds1da177e2005-04-16 15:20:36 -07003614 sbni= [NET] Granch SBNI12 leased line adapter
Randy Dunlapa9913042005-10-23 12:57:11 -07003615
Mike Travisf6630112009-11-17 18:22:15 -06003616 sched_debug [KNL] Enables verbose scheduler debug messages.
3617
Mel Gormancb251762016-02-05 09:08:36 +00003618 schedstats= [KNL,X86] Enable or disable scheduled statistics.
3619 Allowed values are enable and disable. This feature
3620 incurs a small amount of overhead in the scheduler
3621 but is useful for debugging and performance tuning.
3622
Mike Galbraith5307c952012-05-08 12:20:58 +02003623 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
3624 xtime_lock contention on larger systems, and/or RCU lock
3625 contention on all systems with CONFIG_MAXSMP set.
3626 Format: { "0" | "1" }
3627 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3628 1 -- enable.
3629 Note: increases power consumption, thus should only be
3630 enabled if running jitter sensitive (HPC/RT) workloads.
3631
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07003632 security= [SECURITY] Choose a security module to enable at boot.
3633 If this boot parameter is not specified, only the first
3634 security module asking for security registration will be
3635 loaded. An invalid security module name will be treated
3636 as if no module has been chosen.
3637
3638 selinux= [SELINUX] Disable or enable SELinux at boot time.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003639 Format: { "0" | "1" }
3640 See security/selinux/Kconfig help text.
3641 0 -- disable.
3642 1 -- enable.
3643 Default value is set via kernel config option.
3644 If enabled at boot time, /selinux/disable can be used
3645 later to disable prior to initial policy load.
3646
John Johansenc1c124e2010-07-29 14:48:09 -07003647 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3648 Format: { "0" | "1" }
3649 See security/apparmor/Kconfig help text
3650 0 -- disable.
3651 1 -- enable.
3652 Default value is set via kernel config option.
3653
Alan Coxcd4f0ef2007-07-31 00:37:59 -07003654 serialnumber [BUGS=X86-32]
Linus Torvalds1da177e2005-04-16 15:20:36 -07003655
Linus Torvalds1da177e2005-04-16 15:20:36 -07003656 shapers= [NET]
3657 Maximal number of shapers.
Randy Dunlapa9913042005-10-23 12:57:11 -07003658
Linus Torvalds1da177e2005-04-16 15:20:36 -07003659 simeth= [IA-64]
3660 simscsi=
Randy Dunlapa9913042005-10-23 12:57:11 -07003661
Linus Torvalds1da177e2005-04-16 15:20:36 -07003662 slram= [HW,MTD]
3663
Joonsoo Kim423c9292014-10-09 15:26:22 -07003664 slab_nomerge [MM]
3665 Disable merging of slabs with similar size. May be
3666 necessary if there is some reason to distinguish
3667 allocs to different slabs. Debug options disable
3668 merging on their own.
3669 For more information see Documentation/vm/slub.txt.
3670
David Rientjes3df1ccc2011-10-18 22:09:28 -07003671 slab_max_order= [MM, SLAB]
3672 Determines the maximum allowed order for slabs.
3673 A high setting may cause OOMs due to memory
3674 fragmentation. Defaults to 1 for systems with
3675 more than 32MB of RAM, 0 otherwise.
3676
Christoph Lameterf0630ff2007-07-15 23:38:14 -07003677 slub_debug[=options[,slabs]] [MM, SLUB]
3678 Enabling slub_debug allows one to determine the
3679 culprit if slab objects become corrupted. Enabling
3680 slub_debug can create guard zones around objects and
3681 may poison objects when not in use. Also tracks the
3682 last alloc / free. For more information see
3683 Documentation/vm/slub.txt.
Christoph Lameterc1aee212007-05-31 00:40:47 -07003684
3685 slub_max_order= [MM, SLUB]
Christoph Lameterf0630ff2007-07-15 23:38:14 -07003686 Determines the maximum allowed order for slabs.
3687 A high setting may cause OOMs due to memory
3688 fragmentation. For more information see
3689 Documentation/vm/slub.txt.
Christoph Lameterc1aee212007-05-31 00:40:47 -07003690
3691 slub_min_objects= [MM, SLUB]
Christoph Lameterf0630ff2007-07-15 23:38:14 -07003692 The minimum number of objects per slab. SLUB will
3693 increase the slab order up to slub_max_order to
3694 generate a sufficiently large slab able to contain
3695 the number of objects indicated. The higher the number
3696 of objects the smaller the overhead of tracking slabs
3697 and the less frequently locks need to be acquired.
Christoph Lameterc1aee212007-05-31 00:40:47 -07003698 For more information see Documentation/vm/slub.txt.
3699
3700 slub_min_order= [MM, SLUB]
Masanari Iida24775d62012-02-15 00:26:42 +09003701 Determines the minimum page order for slabs. Must be
Christoph Lameterf0630ff2007-07-15 23:38:14 -07003702 lower than slub_max_order.
Christoph Lameterc1aee212007-05-31 00:40:47 -07003703 For more information see Documentation/vm/slub.txt.
3704
3705 slub_nomerge [MM, SLUB]
Joonsoo Kim423c9292014-10-09 15:26:22 -07003706 Same with slab_nomerge. This is supported for legacy.
3707 See slab_nomerge for more information.
Christoph Lameterc1aee212007-05-31 00:40:47 -07003708
Linus Torvalds1da177e2005-04-16 15:20:36 -07003709 smart2= [HW]
3710 Format: <io1>[,<io2>[,...,<io8>]]
3711
Bjorn Helgaasd0d4f692007-05-08 00:36:05 -07003712 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3713 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3714 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3715 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3716 smsc-ircc2.ircc_irq= [HW] IRQ line
3717 smsc-ircc2.ircc_dma= [HW] DMA channel
3718 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3719 0: Toshiba Satellite 1800 (GP data pin select)
3720 1: Fast pin select (default)
3721 2: ATC IRMode
3722
Sascha Silbe52c48c52016-04-05 12:53:38 +02003723 smt [KNL,S390] Set the maximum number of threads (logical
3724 CPUs) to use per physical CPU on systems capable of
3725 symmetric multithreading (SMT). Will be capped to the
3726 actual hardware limit.
3727 Format: <integer>
3728 Default: -1 (no limit)
3729
Ingo Molnar9c44bc02008-05-12 21:21:04 +02003730 softlockup_panic=
3731 [KNL] Should the soft-lockup detector generate panics.
Randy Dunlap44a4dcf2011-04-04 15:02:24 -07003732 Format: <integer>
Ingo Molnar9c44bc02008-05-12 21:21:04 +02003733
Aaron Tomlined235872014-06-23 13:22:05 -07003734 softlockup_all_cpu_backtrace=
3735 [KNL] Should the soft-lockup detector generate
3736 backtraces on all cpus.
3737 Format: <integer>
3738
Linus Torvalds1da177e2005-04-16 15:20:36 -07003739 sonypi.*= [HW] Sony Programmable I/O Control Device driver
Paul Bolle395cf962011-08-15 02:02:26 +02003740 See Documentation/laptops/sonypi.txt
Linus Torvalds1da177e2005-04-16 15:20:36 -07003741
Linus Torvalds1da177e2005-04-16 15:20:36 -07003742 spia_io_base= [HW,MTD]
3743 spia_fio_base=
3744 spia_pedr=
3745 spia_peddr=
3746
Steven Rostedtf38f1d22008-12-16 23:06:40 -05003747 stacktrace [FTRACE]
3748 Enabled the stack tracer on boot up.
3749
Steven Rostedt762e1202011-12-19 22:01:00 -05003750 stacktrace_filter=[function-list]
3751 [FTRACE] Limit the functions that the stack tracer
3752 will trace at boot up. function-list is a comma separated
3753 list of functions. This list can be changed at run
3754 time by the stack_trace_filter file in the debugfs
3755 tracing directory. Note, this enables stack tracing
3756 and the stacktrace above is not needed.
3757
Linus Torvalds1da177e2005-04-16 15:20:36 -07003758 sti= [PARISC,HW]
3759 Format: <num>
3760 Set the STI (builtin display/keyboard on the HP-PARISC
3761 machines) console (graphic card) which should be used
3762 as the initial boot-console.
3763 See also comment in drivers/video/console/sticore.c.
3764
3765 sti_font= [HW]
3766 See comment in drivers/video/console/sticore.c.
3767
3768 stifb= [HW]
3769 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
3770
Trond Myklebustcbf11072009-08-09 15:06:19 -04003771 sunrpc.min_resvport=
3772 sunrpc.max_resvport=
3773 [NFS,SUNRPC]
3774 SunRPC servers often require that client requests
3775 originate from a privileged port (i.e. a port in the
3776 range 0 < portnr < 1024).
3777 An administrator who wishes to reserve some of these
3778 ports for other uses may adjust the range that the
3779 kernel's sunrpc client considers to be privileged
3780 using these two parameters to set the minimum and
3781 maximum port values.
3782
Trond Myklebustff3ac5c2016-06-24 10:55:50 -04003783 sunrpc.svc_rpc_per_connection_limit=
3784 [NFS,SUNRPC]
3785 Limit the number of requests that the server will
3786 process in parallel from a single connection.
3787 The default value is 0 (no limit).
3788
Greg Banks42a7fc42007-03-06 01:42:23 -08003789 sunrpc.pool_mode=
3790 [NFS]
3791 Control how the NFS server code allocates CPUs to
3792 service thread pools. Depending on how many NICs
3793 you have and where their interrupts are bound, this
3794 option will affect which CPUs will do NFS serving.
3795 Note: this parameter cannot be changed while the
3796 NFS server is running.
3797
3798 auto the server chooses an appropriate mode
3799 automatically using heuristics
3800 global a single global pool contains all CPUs
3801 percpu one pool for each CPU
3802 pernode one pool for each NUMA node (equivalent
3803 to global on non-NUMA machines)
3804
Trond Myklebustcbf11072009-08-09 15:06:19 -04003805 sunrpc.tcp_slot_table_entries=
3806 sunrpc.udp_slot_table_entries=
3807 [NFS,SUNRPC]
3808 Sets the upper limit on the number of simultaneous
3809 RPC calls that can be sent from the client to a
3810 server. Increasing these values may allow you to
3811 improve throughput, but will also increase the
3812 amount of memory reserved for use by the client.
3813
Brian Norris1d4a9c12015-02-22 21:16:49 -08003814 suspend.pm_test_delay=
3815 [SUSPEND]
3816 Sets the number of seconds to remain in a suspend test
3817 mode before resuming the system (see
3818 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
3819 is set. Default value is 5.
3820
Michal Hocko07555ac2013-08-22 16:35:46 -07003821 swapaccount=[0|1]
Michal Hockoa42c3902010-11-24 12:57:08 -08003822 [KNL] Enable accounting of swap in memory resource
3823 controller if no parameter or 1 is given or disable
seokhoon.yoon09c3bcc2016-08-02 23:23:57 +09003824 it if 0 is given (See Documentation/cgroup-v1/memory.txt)
Michal Hockoa42c3902010-11-24 12:57:08 -08003825
Jiri Kosina91fec0f2013-11-27 13:48:09 +01003826 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
Geert Uytterhoevenfff5d992016-12-16 14:28:42 +01003827 Format: { <int> | force | noforce }
Jiri Kosina91fec0f2013-11-27 13:48:09 +01003828 <int> -- Number of I/O TLB slabs
3829 force -- force using of bounce buffers even if they
3830 wouldn't be automatically used by the kernel
Geert Uytterhoevenfff5d992016-12-16 14:28:42 +01003831 noforce -- Never use bounce buffers (for debugging)
Randy Dunlapa9913042005-10-23 12:57:11 -07003832
Linus Torvalds1da177e2005-04-16 15:20:36 -07003833 switches= [HW,M68k]
3834
Andi Kleene52eec12010-09-08 16:54:17 +02003835 sysfs.deprecated=0|1 [KNL]
3836 Enable/disable old style sysfs layout for old udev
3837 on older distributions. When this option is enabled
3838 very new udev will not work anymore. When this option
3839 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
3840 in older udev will not work anymore.
3841 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
3842 the kernel configuration.
3843
Ingo Molnar5d6f6472006-12-13 00:34:36 -08003844 sysrq_always_enabled
3845 [KNL]
3846 Ignore sysrq setting - this boot parameter will
3847 neutralize any effect of /proc/sys/kernel/sysrq.
3848 Useful for debugging.
3849
Fabian Frederick747029a2014-11-06 19:46:50 +01003850 tcpmhash_entries= [KNL,NET]
3851 Set the number of tcp_metrics_hash slots.
3852 Default value is 8192 or 16384 depending on total
3853 ram pages. This is used to specify the TCP metrics
3854 cache size. See Documentation/networking/ip-sysctl.txt
3855 "tcp_no_metrics_save" section for more details.
3856
Linus Torvalds1da177e2005-04-16 15:20:36 -07003857 tdfx= [HW,DRM]
3858
Srinivas Pandruvadaacc82342014-09-02 11:54:41 -07003859 test_suspend= [SUSPEND][,N]
David Brownell77437fd2008-07-23 21:28:33 -07003860 Specify "mem" (for Suspend-to-RAM) or "standby" (for
Srinivas Pandruvadaacc82342014-09-02 11:54:41 -07003861 standby suspend) or "freeze" (for suspend type freeze)
3862 as the system sleep state during system startup with
3863 the optional capability to repeat N number of times.
3864 The system is woken from this state using a
3865 wakeup-capable RTC alarm.
David Brownell77437fd2008-07-23 21:28:33 -07003866
Linus Torvalds1da177e2005-04-16 15:20:36 -07003867 thash_entries= [KNL,NET]
3868 Set number of hash buckets for TCP connection
3869
Len Brownf8707ec2007-08-12 00:12:54 -04003870 thermal.act= [HW,ACPI]
3871 -1: disable all active trip points in all thermal zones
3872 <degrees C>: override all lowest active trip points
3873
Len Brownc52a7412007-08-14 15:49:32 -04003874 thermal.crt= [HW,ACPI]
3875 -1: disable all critical trip points in all thermal zones
Zhang Rui22a94d72008-10-17 02:41:20 -04003876 <degrees C>: override all critical trip points
Len Brownc52a7412007-08-14 15:49:32 -04003877
Len Brownf5487142007-08-12 00:12:44 -04003878 thermal.nocrt= [HW,ACPI]
3879 Set to disable actions on ACPI thermal zone
3880 critical and hot trip points.
3881
Len Brown72b33ef2007-08-12 00:12:17 -04003882 thermal.off= [HW,ACPI]
3883 1: disable ACPI thermal control
3884
Len Browna70cdc52007-08-12 00:12:35 -04003885 thermal.psv= [HW,ACPI]
3886 -1: disable all passive trip points
Randy Dunlapada9cfd2008-12-19 10:57:32 -08003887 <degrees C>: override all passive trip points to this
3888 value
Len Browna70cdc52007-08-12 00:12:35 -04003889
Len Brown730ff342007-08-12 00:12:26 -04003890 thermal.tzp= [HW,ACPI]
3891 Specify global default ACPI thermal zone polling rate
3892 <deci-seconds>: poll all this frequency
3893 0: no polling (default)
3894
Thomas Gleixner8d32a302011-02-23 23:52:23 +00003895 threadirqs [KNL]
3896 Force threading of all interrupt handlers except those
Masanari Iida24775d62012-02-15 00:26:42 +09003897 marked explicitly IRQF_NO_THREAD.
Thomas Gleixner8d32a302011-02-23 23:52:23 +00003898
Konrad Rzeszutek Wilk2ca62b02013-05-08 17:12:44 -04003899 tmem [KNL,XEN]
3900 Enable the Transcendent memory driver if built-in.
3901
3902 tmem.cleancache=0|1 [KNL, XEN]
3903 Default is on (1). Disable the usage of the cleancache
3904 API to send anonymous pages to the hypervisor.
3905
3906 tmem.frontswap=0|1 [KNL, XEN]
3907 Default is on (1). Disable the usage of the frontswap
Konrad Rzeszutek Wilk37d46e12013-05-14 13:56:42 -04003908 API to send swap pages to the hypervisor. If disabled
3909 the selfballooning and selfshrinking are force disabled.
Konrad Rzeszutek Wilk2ca62b02013-05-08 17:12:44 -04003910
3911 tmem.selfballooning=0|1 [KNL, XEN]
3912 Default is on (1). Disable the driving of swap pages
3913 to the hypervisor.
3914
3915 tmem.selfshrinking=0|1 [KNL, XEN]
3916 Default is on (1). Partial swapoff that immediately
3917 transfers pages from Xen hypervisor back to the
3918 kernel based on different criteria.
3919
Heiko Carstens2b1a61f2008-12-25 13:39:23 +01003920 topology= [S390]
3921 Format: {off | on}
3922 Specify if the kernel should make use of the cpu
Sylvestre Ledruf65e51d2011-04-04 15:04:46 -07003923 topology information if the hardware supports this.
3924 The scheduler will make use of this information and
Heiko Carstens2b1a61f2008-12-25 13:39:23 +01003925 e.g. base its process migration decisions on it.
Heiko Carstensc9af3fa2010-10-25 16:10:43 +02003926 Default is on.
Heiko Carstens2b1a61f2008-12-25 13:39:23 +01003927
Nishanth Aravamudan2d73bae2014-10-10 09:04:49 -07003928 topology_updates= [KNL, PPC, NUMA]
3929 Format: {off}
3930 Specify if the kernel should ignore (off)
3931 topology updates sent by the hypervisor to this
3932 LPAR.
3933
Linus Torvalds1da177e2005-04-16 15:20:36 -07003934 tp720= [HW,PS2]
3935
Rajiv Andrade225a9be2010-03-25 00:55:32 -03003936 tpm_suspend_pcr=[HW,TPM]
3937 Format: integer pcr id
3938 Specify that at suspend time, the tpm driver
3939 should extend the specified pcr with zeros,
3940 as a workaround for some chips which fail to
3941 flush the last written pcr on TPM_SaveState.
3942 This will guarantee that all the other pcrs
3943 are saved.
3944
Li Zefan9d612be2009-06-24 17:33:15 +08003945 trace_buf_size=nn[KMG]
Joonsoo Kim3e6fb8e2014-12-03 10:39:20 +09003946 [FTRACE] will set tracing buffer size on each cpu.
KOSAKI Motohiro631595f2009-03-10 13:57:10 +09003947
Li Zefan020e5f82009-07-01 10:47:05 +08003948 trace_event=[event-list]
3949 [FTRACE] Set and start specified trace events in order
Brian Norrisd81749e2016-05-23 13:37:58 -07003950 to facilitate early boot debugging. The event-list is a
3951 comma separated list of trace events to enable. See
3952 also Documentation/trace/events.txt
Li Zefan020e5f82009-07-01 10:47:05 +08003953
Steven Rostedt7bcfaf52012-11-01 22:56:07 -04003954 trace_options=[option-list]
3955 [FTRACE] Enable or disable tracer options at boot.
3956 The option-list is a comma delimited list of options
3957 that can be enabled or disabled just as if you were
3958 to echo the option name into
3959
3960 /sys/kernel/debug/tracing/trace_options
3961
3962 For example, to enable stacktrace option (to dump the
3963 stack trace of each event), add to the command line:
3964
3965 trace_options=stacktrace
3966
3967 See also Documentation/trace/ftrace.txt "trace options"
3968 section.
3969
Steven Rostedt (Red Hat)0daa23022014-12-12 22:27:10 -05003970 tp_printk[FTRACE]
3971 Have the tracepoints sent to printk as well as the
3972 tracing ring buffer. This is useful for early boot up
3973 where the system hangs or reboots and does not give the
3974 option for reading the tracing buffer or performing a
3975 ftrace_dump_on_oops.
3976
3977 To turn off having tracepoints sent to printk,
3978 echo 0 > /proc/sys/kernel/tracepoint_printk
3979 Note, echoing 1 into this file without the
3980 tracepoint_printk kernel cmdline option has no effect.
3981
3982 ** CAUTION **
3983
3984 Having tracepoints sent to printk() and activating high
3985 frequency tracepoints such as irq or sched, can cause
3986 the system to live lock.
3987
Steven Rostedt (Red Hat)de7edd32013-06-14 16:21:43 -04003988 traceoff_on_warning
3989 [FTRACE] enable this option to disable tracing when a
3990 warning is hit. This turns off "tracing_on". Tracing can
3991 be enabled again by echoing '1' into the "tracing_on"
3992 file located in /sys/kernel/debug/tracing/
3993
3994 This option is useful, as it disables the trace before
3995 the WARNING dump is called, which prevents the trace to
3996 be filled with content caused by the warning output.
3997
3998 This option can also be set at run time via the sysctl
3999 option: kernel/traceoff_on_warning
4000
Jiri Kosinafcf4d822012-03-21 16:34:02 -07004001 transparent_hugepage=
4002 [KNL]
4003 Format: [always|madvise|never]
4004 Can be used to control the default behavior of the system
4005 with respect to transparent hugepages.
4006 See Documentation/vm/transhuge.txt for more details.
4007
john stultzd3b8f882009-08-17 16:40:47 -07004008 tsc= Disable clocksource stability checks for TSC.
Alok Kataria395628e2008-10-24 17:22:01 -07004009 Format: <string>
4010 [x86] reliable: mark tsc clocksource as reliable, this
john stultzd3b8f882009-08-17 16:40:47 -07004011 disables clocksource verification at runtime, as well
4012 as the stability checks done at bootup. Used to enable
4013 high-resolution timer mode on older hardware, and in
4014 virtualized environment.
Venkatesh Pallipadie82b8e42010-10-04 17:03:20 -07004015 [x86] noirqtime: Do not use TSC to do irq accounting.
4016 Used to run time disable IRQ_TIME_ACCOUNTING on any
4017 platforms where RDTSC is slow and this accounting
4018 can add overhead.
Alok Kataria395628e2008-10-24 17:22:01 -07004019
Randy Dunlapa9913042005-10-23 12:57:11 -07004020 turbografx.map[2|3]= [HW,JOY]
4021 TurboGraFX parallel port interface
4022 Format:
4023 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
Linus Torvalds1da177e2005-04-16 15:20:36 -07004024 See also Documentation/input/joystick-parport.txt
4025
Christian Kujaub6935f82011-05-31 15:22:05 +00004026 udbg-immortal [PPC] When debugging early kernel crashes that
Jani Nikulae52347b2016-11-03 12:10:10 +02004027 happen after console_init() and before a proper
Christian Kujaub6935f82011-05-31 15:22:05 +00004028 console driver takes over, this boot options might
4029 help "seeing" what's going on.
4030
Eric Dumazetf86dcc52009-10-07 00:37:59 +00004031 uhash_entries= [KNL,NET]
4032 Set number of hash buckets for UDP/UDP-Lite connections
4033
Alan Stern5f8364b2006-12-05 16:29:55 -05004034 uhci-hcd.ignore_oc=
4035 [USB] Ignore overcurrent events (default N).
4036 Some badly-designed motherboards generate lots of
4037 bogus events, for ports that aren't wired to
4038 anything. Set this parameter to avoid log spamming.
4039 Note that genuine overcurrent events won't be
4040 reported either.
4041
Simon Arlotte3a61b02008-07-19 23:32:54 +01004042 unknown_nmi_panic
Randy Dunlap44a4dcf2011-04-04 15:02:24 -07004043 [X86] Cause panic on unknown NMI.
Simon Arlotte3a61b02008-07-19 23:32:54 +01004044
Carl-Daniel Hailfingerc4fc2342011-05-31 21:31:08 +02004045 usbcore.authorized_default=
4046 [USB] Default USB device authorization:
4047 (default -1 = authorized except for wireless USB,
4048 0 = not authorized, 1 = authorized)
4049
Alan Sternb5e795f2007-02-20 15:00:53 -05004050 usbcore.autosuspend=
4051 [USB] The autosuspend time delay (in seconds) used
4052 for newly-detected USB devices (default 2). This
4053 is the time required before an idle device will be
4054 autosuspended. Devices for which the delay is set
Alan Sterneaafbc32007-03-13 16:39:15 -04004055 to a negative value won't be autosuspended at all.
Alan Sternb5e795f2007-02-20 15:00:53 -05004056
Jaroslav Kyselafd7c5192008-10-10 16:24:45 +02004057 usbcore.usbfs_snoop=
4058 [USB] Set to log all usbfs traffic (default 0 = off).
4059
Alan Stern0290cc92015-11-20 13:53:22 -05004060 usbcore.usbfs_snoop_max=
4061 [USB] Maximum number of bytes to snoop in each URB
4062 (default = 65536).
4063
Jaroslav Kyselafd7c5192008-10-10 16:24:45 +02004064 usbcore.blinkenlights=
4065 [USB] Set to cycle leds on hubs (default 0 = off).
4066
4067 usbcore.old_scheme_first=
4068 [USB] Start with the old device initialization
4069 scheme (default 0 = off).
4070
Alan Stern3f5eb8d2011-11-17 16:41:35 -05004071 usbcore.usbfs_memory_mb=
4072 [USB] Memory limit (in MB) for buffers allocated by
4073 usbfs (default = 16, 0 = max = 2047).
4074
Jaroslav Kyselafd7c5192008-10-10 16:24:45 +02004075 usbcore.use_both_schemes=
4076 [USB] Try the other device initialization scheme
4077 if the first one fails (default 1 = enabled).
4078
4079 usbcore.initial_descriptor_timeout=
4080 [USB] Specifies timeout for the initial 64-byte
4081 USB_REQ_GET_DESCRIPTOR request in milliseconds
4082 (default 5000 = 5.0 seconds).
4083
Oliver Neukum40d58142015-12-03 15:03:32 +01004084 usbcore.nousb [USB] Disable the USB subsystem
4085
Linus Torvalds1da177e2005-04-16 15:20:36 -07004086 usbhid.mousepoll=
4087 [USBHID] The interval which mice are to be polled at.
Randy Dunlapa9913042005-10-23 12:57:11 -07004088
Alan Sternd4f373e2008-11-10 14:07:45 -05004089 usb-storage.delay_use=
4090 [UMS] The delay in seconds before a new device is
Mark Knibbs19101952014-11-04 13:00:15 +00004091 scanned for Logical Units (default 1).
Alan Sternd4f373e2008-11-10 14:07:45 -05004092
4093 usb-storage.quirks=
4094 [UMS] A list of quirks entries to supplement or
4095 override the built-in unusual_devs list. List
4096 entries are separated by commas. Each entry has
4097 the form VID:PID:Flags where VID and PID are Vendor
4098 and Product ID values (4-digit hex numbers) and
4099 Flags is a set of characters, each corresponding
4100 to a common usb-storage quirk flag as follows:
Alan Sternc838ea42008-12-15 10:40:06 -05004101 a = SANE_SENSE (collect more than 18 bytes
4102 of sense data);
Alan Sterna0bb1082009-12-07 16:39:16 -05004103 b = BAD_SENSE (don't collect more than 18
4104 bytes of sense data);
Alan Sternd4f373e2008-11-10 14:07:45 -05004105 c = FIX_CAPACITY (decrease the reported
4106 device capacity by one sector);
Karl Relton51169012011-05-18 21:42:34 +01004107 d = NO_READ_DISC_INFO (don't use
4108 READ_DISC_INFO command);
4109 e = NO_READ_CAPACITY_16 (don't use
4110 READ_CAPACITY_16 command);
Hans de Goede734016b2014-09-16 18:36:52 +02004111 f = NO_REPORT_OPCODES (don't use report opcodes
4112 command, uas only);
Hans de Goedeee136af2015-04-21 11:20:31 +02004113 g = MAX_SECTORS_240 (don't transfer more than
4114 240 sectors at a time, uas only);
Alan Sternc838ea42008-12-15 10:40:06 -05004115 h = CAPACITY_HEURISTICS (decrease the
4116 reported device capacity by one
4117 sector if the number is odd);
Alan Sternd4f373e2008-11-10 14:07:45 -05004118 i = IGNORE_DEVICE (don't bind to this
4119 device);
Hans de Goede13630742016-04-12 12:27:09 +02004120 j = NO_REPORT_LUNS (don't use report luns
4121 command, uas only);
Alan Sternd4f373e2008-11-10 14:07:45 -05004122 l = NOT_LOCKABLE (don't try to lock and
4123 unlock ejectable media);
4124 m = MAX_SECTORS_64 (don't transfer more
4125 than 64 sectors = 32 KB at a time);
Alan Stern21c13a42011-06-07 11:35:52 -04004126 n = INITIAL_READ10 (force a retry of the
4127 initial READ(10) command);
Alan Sternc838ea42008-12-15 10:40:06 -05004128 o = CAPACITY_OK (accept the capacity
4129 reported by the device);
Namjae Jeoneaa05df2012-07-07 23:05:28 -04004130 p = WRITE_CACHE (the device cache is ON
4131 by default);
Alan Sternd4f373e2008-11-10 14:07:45 -05004132 r = IGNORE_RESIDUE (the device reports
4133 bogus residue values);
4134 s = SINGLE_LUN (the device has only one
4135 Logical Unit);
Hans de Goede59307852014-09-15 16:04:12 +02004136 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4137 commands, uas only);
Alan Sternb6089f12014-09-02 15:42:18 -04004138 u = IGNORE_UAS (don't bind to the uas driver);
Alan Sternd4f373e2008-11-10 14:07:45 -05004139 w = NO_WP_DETECT (don't test whether the
4140 medium is write-protected).
Oliver Neukum050bc4e2016-09-12 15:19:41 +02004141 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4142 even if the device claims no cache)
Alan Sternd4f373e2008-11-10 14:07:45 -05004143 Example: quirks=0419:aaf5:rl,0421:0433:rc
4144
Stephen Boydac1667d2011-08-13 12:34:50 -07004145 user_debug= [KNL,ARM]
4146 Format: <int>
4147 See arch/arm/Kconfig.debug help text.
4148 1 - undefined instruction events
4149 2 - system calls
4150 4 - invalid data aborts
4151 8 - SIGSEGV faults
4152 16 - SIGBUS faults
4153 Example: user_debug=31
4154
Ian Campbell14315592010-02-17 10:38:10 +00004155 userpte=
4156 [X86] Flags controlling user PTE allocations.
4157
4158 nohigh = do not allocate PTE pages in
4159 HIGHMEM regardless of setting
4160 of CONFIG_HIGHPTE.
4161
Jaswinder Singh Rajput6cececf2009-04-14 14:03:43 +05304162 vdso= [X86,SH]
Andy Lutomirskib0b49f22014-03-13 16:01:26 -07004163 On X86_32, this is an alias for vdso32=. Otherwise:
4164
4165 vdso=1: enable VDSO (the default)
Ingo Molnare6e54942006-06-27 02:53:50 -07004166 vdso=0: disable VDSO mapping
4167
Andy Lutomirskib0b49f22014-03-13 16:01:26 -07004168 vdso32= [X86] Control the 32-bit vDSO
4169 vdso32=1: enable 32-bit VDSO
4170 vdso32=0 or vdso32=2: disable 32-bit VDSO
4171
4172 See the help text for CONFIG_COMPAT_VDSO for more
4173 details. If CONFIG_COMPAT_VDSO is set, the default is
4174 vdso32=0; otherwise, the default is vdso32=1.
4175
4176 For compatibility with older kernels, vdso32=2 is an
4177 alias for vdso32=0.
4178
4179 Try vdso32=0 if you encounter an error that says:
4180 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
Roland McGrathaf65d642008-01-30 13:30:43 +01004181
Yasuaki Ishimatsud080d392007-07-17 21:22:55 +09004182 vector= [IA-64,SMP]
4183 vector=percpu: enable percpu vector domain
4184
Linus Torvalds1da177e2005-04-16 15:20:36 -07004185 video= [FB] Frame buffer configuration
4186 See Documentation/fb/modedb.txt.
4187
Aaron Lu3afe6da2013-06-20 15:08:55 +08004188 video.brightness_switch_enabled= [0,1]
4189 If set to 1, on receiving an ACPI notify event
4190 generated by hotkey, video driver will adjust brightness
4191 level and then send out the event to user space through
4192 the allocated input device; If set to 0, video driver
4193 will only send out the event without touching backlight
4194 brightness level.
Rafael J. Wysocki28437682014-07-14 19:35:45 +02004195 default: 1
Aaron Lu3afe6da2013-06-20 15:08:55 +08004196
Pawel Moll81a054c2012-05-09 18:30:16 +01004197 virtio_mmio.device=
4198 [VMMIO] Memory mapped virtio (platform) device.
4199
4200 <size>@<baseaddr>:<irq>[:<id>]
4201 where:
4202 <size> := size (can use standard suffixes
4203 like K, M and G)
4204 <baseaddr> := physical base address
4205 <irq> := interrupt number (as passed to
4206 request_irq())
4207 <id> := (optional) platform device id
4208 example:
4209 virtio_mmio.device=1K@0x100b0000:48:7
4210
4211 Can be used multiple times for multiple devices.
4212
Alan Coxcd4f0ef2007-07-31 00:37:59 -07004213 vga= [BOOT,X86-32] Select a particular video mode
Kyle McMartin954a8b82009-02-19 16:48:14 -05004214 See Documentation/x86/boot.txt and
Randy Dunlapa9913042005-10-23 12:57:11 -07004215 Documentation/svga.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -07004216 Use vga=ask for menu.
4217 This is actually a boot loader parameter; the value is
4218 passed to the kernel using a special protocol.
4219
Randy Dunlapa9913042005-10-23 12:57:11 -07004220 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
Linus Torvalds1da177e2005-04-16 15:20:36 -07004221 size of <nn>. This can be used to increase the
4222 minimum size (128MB on x86). It can also be used to
4223 decrease the size and leave more room for directly
4224 mapped kernel RAM.
4225
Peter Oberparleiter585c3042006-06-29 15:08:25 +02004226 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
4227 Format: <command>
Linus Torvalds1da177e2005-04-16 15:20:36 -07004228
Peter Oberparleiter585c3042006-06-29 15:08:25 +02004229 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
4230 Format: <command>
4231
4232 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4233 Format: <command>
Randy Dunlapa9913042005-10-23 12:57:11 -07004234
Andy Lutomirski3ae36652011-08-10 11:15:32 -04004235 vsyscall= [X86-64]
4236 Controls the behavior of vsyscalls (i.e. calls to
4237 fixed addresses of 0xffffffffff600x00 from legacy
4238 code). Most statically-linked binaries and older
4239 versions of glibc use these calls. Because these
4240 functions are at fixed addresses, they make nice
4241 targets for exploits that can control RIP.
4242
Andy Lutomirski2e57ae02011-11-07 16:33:41 -08004243 emulate [default] Vsyscalls turn into traps and are
4244 emulated reasonably safely.
Andy Lutomirski3ae36652011-08-10 11:15:32 -04004245
Andy Lutomirski2e57ae02011-11-07 16:33:41 -08004246 native Vsyscalls are native syscall instructions.
Andy Lutomirski3ae36652011-08-10 11:15:32 -04004247 This is a little bit faster than trapping
4248 and makes a few dynamic recompilers work
4249 better than they would in emulation mode.
4250 It also makes exploits much easier to write.
4251
4252 none Vsyscalls don't work at all. This makes
4253 them quite hard to use for exploits but
4254 might break your system.
4255
Clemens Ladisch3855ae12013-08-04 13:09:50 +02004256 vt.color= [VT] Default text color.
4257 Format: 0xYX, X = foreground, Y = background.
4258 Default: 0x07 = light gray on black.
4259
Clemens Ladisch9ea9a882009-12-15 16:45:39 -08004260 vt.cur_default= [VT] Default cursor shape.
4261 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4262 the parameters of the <Esc>[?A;B;Cc escape sequence;
4263 see VGA-softcursor.txt. Default: 2 = underline.
4264
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07004265 vt.default_blu= [VT]
4266 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4267 Change the default blue palette of the console.
4268 This is a 16-member array composed of values
4269 ranging from 0-255.
4270
4271 vt.default_grn= [VT]
4272 Format: <green0>,<green1>,<green2>,...,<green15>
4273 Change the default green palette of the console.
4274 This is a 16-member array composed of values
4275 ranging from 0-255.
4276
4277 vt.default_red= [VT]
4278 Format: <red0>,<red1>,<red2>,...,<red15>
4279 Change the default red palette of the console.
4280 This is a 16-member array composed of values
4281 ranging from 0-255.
4282
4283 vt.default_utf8=
4284 [VT]
4285 Format=<0|1>
4286 Set system-wide default UTF-8 mode for all tty's.
4287 Default is 1, i.e. UTF-8 mode is enabled for all
4288 newly opened terminals.
4289
Matthew Garrettf6c06b62009-11-13 15:14:11 -05004290 vt.global_cursor_default=
4291 [VT]
4292 Format=<-1|0|1>
4293 Set system-wide default for whether a cursor
4294 is shown on new VTs. Default is -1,
4295 i.e. cursors will be created by default unless
4296 overridden by individual drivers. 0 will hide
4297 cursors, 1 will display them.
4298
Clemens Ladisch3855ae12013-08-04 13:09:50 +02004299 vt.italic= [VT] Default color for italic text; 0-15.
4300 Default: 2 = green.
4301
4302 vt.underline= [VT] Default color for underlined text; 0-15.
4303 Default: 3 = cyan.
4304
Randy Dunlap4724ba572010-05-03 11:42:52 -07004305 watchdog timers [HW,WDT] For information on watchdog timers,
4306 see Documentation/watchdog/watchdog-parameters.txt
4307 or other driver-specific files in the
4308 Documentation/watchdog/ directory.
Linus Torvalds1da177e2005-04-16 15:20:36 -07004309
Tejun Heo82607adc2015-12-08 11:28:04 -05004310 workqueue.watchdog_thresh=
4311 If CONFIG_WQ_WATCHDOG is configured, workqueue can
4312 warn stall conditions and dump internal state to
4313 help debugging. 0 disables workqueue stall
4314 detection; otherwise, it's the stall threshold
4315 duration in seconds. The default value is 30 and
4316 it can be updated at runtime by writing to the
4317 corresponding sysfs file.
4318
Tejun Heod55262c2013-04-01 11:23:38 -07004319 workqueue.disable_numa
4320 By default, all work items queued to unbound
4321 workqueues are affine to the NUMA nodes they're
4322 issued on, which results in better behavior in
4323 general. If NUMA affinity needs to be disabled for
4324 whatever reason, this option can be used. Note
4325 that this also can be controlled per-workqueue for
4326 workqueues visible under /sys/bus/workqueue/.
4327
Viresh Kumarcee22a12013-04-08 16:45:40 +05304328 workqueue.power_efficient
4329 Per-cpu workqueues are generally preferred because
4330 they show better performance thanks to cache
4331 locality; unfortunately, per-cpu workqueues tend to
4332 be more power hungry than unbound workqueues.
4333
4334 Enabling this makes the per-cpu workqueues which
4335 were observed to contribute significantly to power
4336 consumption unbound, leading to measurably lower
4337 power usage at the cost of small performance
4338 overhead.
4339
4340 The default value of this parameter is determined by
4341 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4342
Tejun Heof303fccb2016-02-09 17:59:38 -05004343 workqueue.debug_force_rr_cpu
4344 Workqueue used to implicitly guarantee that work
4345 items queued without explicit CPU specified are put
4346 on the local CPU. This guarantee is no longer true
4347 and while local CPU is still preferred work items
4348 may be put on foreign CPUs. This debug option
4349 forces round-robin CPU selection to flush out
4350 usages which depend on the now broken guarantee.
4351 When enabled, memory and cache locality will be
4352 impacted.
4353
Randy Dunlap0cb55ad2009-04-05 15:55:22 -07004354 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
4355 default x2apic cluster mode on platforms
4356 supporting x2apic.
4357
Kuppuswamy Sathyanarayanan712b6aa2013-10-17 15:35:29 -07004358 x86_intel_mid_timer= [X86-32,APBT]
4359 Choose timer option for x86 Intel MID platform.
Jacob Panbb24c472009-09-02 07:37:17 -07004360 Two valid options are apbt timer only and lapic timer
4361 plus one apbt timer for broadcast timer.
Kuppuswamy Sathyanarayanan712b6aa2013-10-17 15:35:29 -07004362 x86_intel_mid_timer=apbt_only | lapic_and_apbt
Jacob Panbb24c472009-09-02 07:37:17 -07004363
Juergen Grossc70727a2015-07-17 06:51:36 +02004364 xen_512gb_limit [KNL,X86-64,XEN]
4365 Restricts the kernel running paravirtualized under Xen
4366 to use only up to 512 GB of RAM. The reason to do so is
4367 crash analysis tools and Xen tools for doing domain
4368 save/restore/migration must be enabled to handle larger
4369 domains.
4370
Stefano Stabellinic1c54132010-05-14 12:44:30 +01004371 xen_emul_unplug= [HW,X86,XEN]
4372 Unplug Xen emulated devices
4373 Format: [unplug0,][unplug1]
4374 ide-disks -- unplug primary master IDE devices
4375 aux-ide-disks -- unplug non-primary-master IDE devices
4376 nics -- unplug network devices
4377 all -- unplug all emulated devices (NICs and IDE disks)
Ian Campbell1dc7ce92010-08-23 11:59:29 +01004378 unnecessary -- unplugging emulated devices is
4379 unnecessary even if the host did not respond to
4380 the unplug protocol
Ian Campbellc93a4df2010-08-23 11:59:28 +01004381 never -- do not unplug even if version check succeeds
Stefano Stabellinic1c54132010-05-14 12:44:30 +01004382
Konrad Rzeszutek Wilk15a3eac2013-09-25 10:07:20 -04004383 xen_nopvspin [X86,XEN]
4384 Disables the ticketlock slowpath using Xen PV
4385 optimizations.
4386
Konrad Rzeszutek Wilk8d693b92014-07-11 11:51:35 -04004387 xen_nopv [X86]
4388 Disables the PV optimizations forcing the HVM guest to
4389 run as generic HVM guest with no PV drivers.
4390
Linus Torvalds1da177e2005-04-16 15:20:36 -07004391 xirc2ps_cs= [NET,PCMCIA]
Randy Dunlapa9913042005-10-23 12:57:11 -07004392 Format:
4393 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]