• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1                          Kernel Parameters
2                          ~~~~~~~~~~~~~~~~~
3
4The following is a consolidated list of the kernel parameters as
5implemented by the __setup(), core_param() and module_param() macros
6and sorted into English Dictionary order (defined as ignoring all
7punctuation and sorting digits before letters in a case insensitive
8manner), and with descriptions where known.
9
10The kernel parses parameters from the kernel command line up to "--";
11if it doesn't recognize a parameter and it doesn't contain a '.', the
12parameter gets passed to init: parameters with '=' go into init's
13environment, others are passed as command line arguments to init.
14Everything after "--" is passed as an argument to init.
15
16Module parameters can be specified in two ways: via the kernel command
17line with a module name prefix, or via modprobe, e.g.:
18
19	(kernel command line) usbcore.blinkenlights=1
20	(modprobe command line) modprobe usbcore blinkenlights=1
21
22Parameters for modules which are built into the kernel need to be
23specified on the kernel command line.  modprobe looks through the
24kernel command line (/proc/cmdline) and collects module parameters
25when it loads a module, so the kernel command line can be used for
26loadable modules too.
27
28Hyphens (dashes) and underscores are equivalent in parameter names, so
29	log_buf_len=1M print-fatal-signals=1
30can also be entered as
31	log-buf-len=1M print_fatal_signals=1
32
33Double-quotes can be used to protect spaces in values, e.g.:
34	param="spaces in here"
35
36cpu lists:
37----------
38
39Some kernel parameters take a list of CPUs as a value, e.g.  isolcpus,
40nohz_full, irqaffinity, rcu_nocbs.  The format of this list is:
41
42	<cpu number>,...,<cpu number>
43
44or
45
46	<cpu number>-<cpu number>
47	(must be a positive range in ascending order)
48
49or a mixture
50
51<cpu number>,...,<cpu number>-<cpu number>
52
53Note that for the special case of a range one can split the range into equal
54sized groups and for each group use some amount from the beginning of that
55group:
56
57	<cpu number>-cpu number>:<used size>/<group size>
58
59For example one can add to the command line following parameter:
60
61	isolcpus=1,2,10-20,100-2000:2/25
62
63where the final item represents CPUs 100,101,125,126,150,151,...
64
65
66
67This document may not be entirely up to date and comprehensive. The command
68"modinfo -p ${modulename}" shows a current list of all parameters of a loadable
69module. Loadable modules, after being loaded into the running kernel, also
70reveal their parameters in /sys/module/${modulename}/parameters/. Some of these
71parameters may be changed at runtime by the command
72"echo -n ${value} > /sys/module/${modulename}/parameters/${parm}".
73
74The parameters listed below are only valid if certain kernel build options were
75enabled and if respective hardware is present. The text in square brackets at
76the beginning of each description states the restrictions within which a
77parameter is applicable:
78
79	ACPI	ACPI support is enabled.
80	AGP	AGP (Accelerated Graphics Port) is enabled.
81	ALSA	ALSA sound support is enabled.
82	APIC	APIC support is enabled.
83	APM	Advanced Power Management support is enabled.
84	ARM	ARM architecture is enabled.
85	AVR32	AVR32 architecture is enabled.
86	AX25	Appropriate AX.25 support is enabled.
87	BLACKFIN Blackfin architecture is enabled.
88	CLK	Common clock infrastructure is enabled.
89	CMA	Contiguous Memory Area support is enabled.
90	DM	Device mapper support is enabled.
91	DRM	Direct Rendering Management support is enabled.
92	DYNAMIC_DEBUG Build in debug messages and enable them at runtime
93	EDD	BIOS Enhanced Disk Drive Services (EDD) is enabled
94	EFI	EFI Partitioning (GPT) is enabled
95	EIDE	EIDE/ATAPI support is enabled.
96	EVM	Extended Verification Module
97	FB	The frame buffer device is enabled.
98	FTRACE	Function tracing enabled.
99	GCOV	GCOV profiling is enabled.
100	HW	Appropriate hardware is enabled.
101	IA-64	IA-64 architecture is enabled.
102	IMA     Integrity measurement architecture is enabled.
103	IOSCHED	More than one I/O scheduler is enabled.
104	IP_PNP	IP DHCP, BOOTP, or RARP is enabled.
105	IPV6	IPv6 support is enabled.
106	ISAPNP	ISA PnP code is enabled.
107	ISDN	Appropriate ISDN support is enabled.
108	JOY	Appropriate joystick support is enabled.
109	KGDB	Kernel debugger support is enabled.
110	KVM	Kernel Virtual Machine support is enabled.
111	LIBATA  Libata driver is enabled
112	LP	Printer support is enabled.
113	LOOP	Loopback device support is enabled.
114	M68k	M68k architecture is enabled.
115			These options have more detailed description inside of
116			Documentation/m68k/kernel-options.txt.
117	MDA	MDA console support is enabled.
118	MIPS	MIPS architecture is enabled.
119	MOUSE	Appropriate mouse support is enabled.
120	MSI	Message Signaled Interrupts (PCI).
121	MTD	MTD (Memory Technology Device) support is enabled.
122	NET	Appropriate network support is enabled.
123	NUMA	NUMA support is enabled.
124	NFS	Appropriate NFS support is enabled.
125	OSS	OSS sound support is enabled.
126	PV_OPS	A paravirtualized kernel is enabled.
127	PARIDE	The ParIDE (parallel port IDE) subsystem is enabled.
128	PARISC	The PA-RISC architecture is enabled.
129	PCI	PCI bus support is enabled.
130	PCIE	PCI Express support is enabled.
131	PCMCIA	The PCMCIA subsystem is enabled.
132	PNP	Plug & Play support is enabled.
133	PPC	PowerPC architecture is enabled.
134	PPT	Parallel port support is enabled.
135	PS2	Appropriate PS/2 support is enabled.
136	RAM	RAM disk support is enabled.
137	S390	S390 architecture is enabled.
138	SCSI	Appropriate SCSI support is enabled.
139			A lot of drivers have their options described inside
140			the Documentation/scsi/ sub-directory.
141	SECURITY Different security models are enabled.
142	SELINUX SELinux support is enabled.
143	APPARMOR AppArmor support is enabled.
144	SERIAL	Serial support is enabled.
145	SH	SuperH architecture is enabled.
146	SMP	The kernel is an SMP kernel.
147	SPARC	Sparc architecture is enabled.
148	SWSUSP	Software suspend (hibernation) is enabled.
149	SUSPEND	System suspend states are enabled.
150	TPM	TPM drivers are enabled.
151	TS	Appropriate touchscreen support is enabled.
152	UMS	USB Mass Storage support is enabled.
153	USB	USB support is enabled.
154	USBHID	USB Human Interface Device support is enabled.
155	V4L	Video For Linux support is enabled.
156	VMMIO   Driver for memory mapped virtio devices is enabled.
157	VGA	The VGA console has been enabled.
158	VT	Virtual terminal support is enabled.
159	WDT	Watchdog support is enabled.
160	XT	IBM PC/XT MFM hard disk support is enabled.
161	X86-32	X86-32, aka i386 architecture is enabled.
162	X86-64	X86-64 architecture is enabled.
163			More X86-64 boot options can be found in
164			Documentation/x86/x86_64/boot-options.txt .
165	X86	Either 32-bit or 64-bit x86 (same as X86-32+X86-64)
166	X86_UV	SGI UV support is enabled.
167	XEN	Xen support is enabled
168
169In addition, the following text indicates that the option:
170
171	BUGS=	Relates to possible processor bugs on the said processor.
172	KNL	Is a kernel start-up parameter.
173	BOOT	Is a boot loader parameter.
174
175Parameters denoted with BOOT are actually interpreted by the boot
176loader, and have no meaning to the kernel directly.
177Do not modify the syntax of boot loader parameters without extreme
178need or coordination with <Documentation/x86/boot.txt>.
179
180There are also arch-specific kernel-parameters not documented here.
181See for example <Documentation/x86/x86_64/boot-options.txt>.
182
183Note that ALL kernel parameters listed below are CASE SENSITIVE, and that
184a trailing = on the name of any parameter states that that parameter will
185be entered as an environment variable, whereas its absence indicates that
186it will appear as a kernel argument readable via /proc/cmdline by programs
187running once the system is up.
188
189The number of kernel parameters is not limited, but the length of the
190complete command line (parameters including spaces etc.) is limited to
191a fixed number of characters. This limit depends on the architecture
192and is between 256 and 4096 characters. It is defined in the file
193./include/asm/setup.h as COMMAND_LINE_SIZE.
194
195Finally, the [KMG] suffix is commonly described after a number of kernel
196parameter values. These 'K', 'M', and 'G' letters represent the _binary_
197multipliers 'Kilo', 'Mega', and 'Giga', equalling 2^10, 2^20, and 2^30
198bytes respectively. Such letter suffixes can also be entirely omitted.
199
200
201	acpi=		[HW,ACPI,X86,ARM64]
202			Advanced Configuration and Power Interface
203			Format: { force | on | off | strict | noirq | rsdt |
204				  copy_dsdt }
205			force -- enable ACPI if default was off
206			on -- enable ACPI but allow fallback to DT [arm64]
207			off -- disable ACPI if default was on
208			noirq -- do not use ACPI for IRQ routing
209			strict -- Be less tolerant of platforms that are not
210				strictly ACPI specification compliant.
211			rsdt -- prefer RSDT over (default) XSDT
212			copy_dsdt -- copy DSDT to memory
213			For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
214			are available
215
216			See also Documentation/power/runtime_pm.txt, pci=noacpi
217
218	acpi_apic_instance=	[ACPI, IOAPIC]
219			Format: <int>
220			2: use 2nd APIC table, if available
221			1,0: use 1st APIC table
222			default: 0
223
224	acpi_backlight=	[HW,ACPI]
225			acpi_backlight=vendor
226			acpi_backlight=video
227			If set to vendor, prefer vendor specific driver
228			(e.g. thinkpad_acpi, sony_acpi, etc.) instead
229			of the ACPI video.ko driver.
230
231	acpi_force_32bit_fadt_addr
232			force FADT to use 32 bit addresses rather than the
233			64 bit X_* addresses. Some firmware have broken 64
234			bit addresses for force ACPI ignore these and use
235			the older legacy 32 bit addresses.
236
237	acpica_no_return_repair [HW, ACPI]
238			Disable AML predefined validation mechanism
239			This mechanism can repair the evaluation result to make
240			the return objects more ACPI specification compliant.
241			This option is useful for developers to identify the
242			root cause of an AML interpreter issue when the issue
243			has something to do with the repair mechanism.
244
245	acpi.debug_layer=	[HW,ACPI,ACPI_DEBUG]
246	acpi.debug_level=	[HW,ACPI,ACPI_DEBUG]
247			Format: <int>
248			CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
249			debug output.  Bits in debug_layer correspond to a
250			_COMPONENT in an ACPI source file, e.g.,
251			    #define _COMPONENT ACPI_PCI_COMPONENT
252			Bits in debug_level correspond to a level in
253			ACPI_DEBUG_PRINT statements, e.g.,
254			    ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
255			The debug_level mask defaults to "info".  See
256			Documentation/acpi/debug.txt for more information about
257			debug layers and levels.
258
259			Enable processor driver info messages:
260			    acpi.debug_layer=0x20000000
261			Enable PCI/PCI interrupt routing info messages:
262			    acpi.debug_layer=0x400000
263			Enable AML "Debug" output, i.e., stores to the Debug
264			object while interpreting AML:
265			    acpi.debug_layer=0xffffffff acpi.debug_level=0x2
266			Enable all messages related to ACPI hardware:
267			    acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
268
269			Some values produce so much output that the system is
270			unusable.  The "log_buf_len" parameter may be useful
271			if you need to capture more output.
272
273	acpi_enforce_resources=	[ACPI]
274			{ strict | lax | no }
275			Check for resource conflicts between native drivers
276			and ACPI OperationRegions (SystemIO and SystemMemory
277			only). IO ports and memory declared in ACPI might be
278			used by the ACPI subsystem in arbitrary AML code and
279			can interfere with legacy drivers.
280			strict (default): access to resources claimed by ACPI
281			is denied; legacy drivers trying to access reserved
282			resources will fail to bind to device using them.
283			lax: access to resources claimed by ACPI is allowed;
284			legacy drivers trying to access reserved resources
285			will bind successfully but a warning message is logged.
286			no: ACPI OperationRegions are not marked as reserved,
287			no further checks are performed.
288
289	acpi_force_table_verification	[HW,ACPI]
290			Enable table checksum verification during early stage.
291			By default, this is disabled due to x86 early mapping
292			size limitation.
293
294	acpi_irq_balance [HW,ACPI]
295			ACPI will balance active IRQs
296			default in APIC mode
297
298	acpi_irq_nobalance [HW,ACPI]
299			ACPI will not move active IRQs (default)
300			default in PIC mode
301
302	acpi_irq_isa=	[HW,ACPI] If irq_balance, mark listed IRQs used by ISA
303			Format: <irq>,<irq>...
304
305	acpi_irq_pci=	[HW,ACPI] If irq_balance, clear listed IRQs for
306			use by PCI
307			Format: <irq>,<irq>...
308
309	acpi_mask_gpe=  [HW,ACPI]
310			Due to the existence of _Lxx/_Exx, some GPEs triggered
311			by unsupported hardware/firmware features can result in
312                        GPE floodings that cannot be automatically disabled by
313                        the GPE dispatcher.
314			This facility can be used to prevent such uncontrolled
315			GPE floodings.
316			Format: <int>
317			Support masking of GPEs numbered from 0x00 to 0x7f.
318
319	acpi_no_auto_serialize	[HW,ACPI]
320			Disable auto-serialization of AML methods
321			AML control methods that contain the opcodes to create
322			named objects will be marked as "Serialized" by the
323			auto-serialization feature.
324			This feature is enabled by default.
325			This option allows to turn off the feature.
326
327	acpi_no_memhotplug [ACPI] Disable memory hotplug.  Useful for kdump
328			   kernels.
329
330	acpi_no_static_ssdt	[HW,ACPI]
331			Disable installation of static SSDTs at early boot time
332			By default, SSDTs contained in the RSDT/XSDT will be
333			installed automatically and they will appear under
334			/sys/firmware/acpi/tables.
335			This option turns off this feature.
336			Note that specifying this option does not affect
337			dynamic table installation which will install SSDT
338			tables to /sys/firmware/acpi/tables/dynamic.
339
340	acpi_rsdp=	[ACPI,EFI,KEXEC]
341			Pass the RSDP address to the kernel, mostly used
342			on machines running EFI runtime service to boot the
343			second kernel for kdump.
344
345	acpi_os_name=	[HW,ACPI] Tell ACPI BIOS the name of the OS
346			Format: To spoof as Windows 98: ="Microsoft Windows"
347
348	acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
349			of 2 which is mandated by ACPI 6) as the supported ACPI
350			specification revision (when using this switch, it may
351			be necessary to carry out a cold reboot _twice_ in a
352			row to make it take effect on the platform firmware).
353
354	acpi_osi=	[HW,ACPI] Modify list of supported OS interface strings
355			acpi_osi="string1"	# add string1
356			acpi_osi="!string2"	# remove string2
357			acpi_osi=!*		# remove all strings
358			acpi_osi=!		# disable all built-in OS vendor
359						  strings
360			acpi_osi=!!		# enable all built-in OS vendor
361						  strings
362			acpi_osi=		# disable all strings
363
364			'acpi_osi=!' can be used in combination with single or
365			multiple 'acpi_osi="string1"' to support specific OS
366			vendor string(s).  Note that such command can only
367			affect the default state of the OS vendor strings, thus
368			it cannot affect the default state of the feature group
369			strings and the current state of the OS vendor strings,
370			specifying it multiple times through kernel command line
371			is meaningless.  This command is useful when one do not
372			care about the state of the feature group strings which
373			should be controlled by the OSPM.
374			Examples:
375			  1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
376			     to 'acpi_osi="Windows 2000" acpi_osi=!', they all
377			     can make '_OSI("Windows 2000")' TRUE.
378
379			'acpi_osi=' cannot be used in combination with other
380			'acpi_osi=' command lines, the _OSI method will not
381			exist in the ACPI namespace.  NOTE that such command can
382			only affect the _OSI support state, thus specifying it
383			multiple times through kernel command line is also
384			meaningless.
385			Examples:
386			  1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
387			     FALSE.
388
389			'acpi_osi=!*' can be used in combination with single or
390			multiple 'acpi_osi="string1"' to support specific
391			string(s).  Note that such command can affect the
392			current state of both the OS vendor strings and the
393			feature group strings, thus specifying it multiple times
394			through kernel command line is meaningful.  But it may
395			still not able to affect the final state of a string if
396			there are quirks related to this string.  This command
397			is useful when one want to control the state of the
398			feature group strings to debug BIOS issues related to
399			the OSPM features.
400			Examples:
401			  1. 'acpi_osi="Module Device" acpi_osi=!*' can make
402			     '_OSI("Module Device")' FALSE.
403			  2. 'acpi_osi=!* acpi_osi="Module Device"' can make
404			     '_OSI("Module Device")' TRUE.
405			  3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
406			     equivalent to
407			     'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
408			     and
409			     'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
410			     they all will make '_OSI("Windows 2000")' TRUE.
411
412	acpi_pm_good	[X86]
413			Override the pmtimer bug detection: force the kernel
414			to assume that this machine's pmtimer latches its value
415			and always returns good values.
416
417	acpi_sci=	[HW,ACPI] ACPI System Control Interrupt trigger mode
418			Format: { level | edge | high | low }
419
420	acpi_skip_timer_override [HW,ACPI]
421			Recognize and ignore IRQ0/pin2 Interrupt Override.
422			For broken nForce2 BIOS resulting in XT-PIC timer.
423
424	acpi_sleep=	[HW,ACPI] Sleep options
425			Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
426				  old_ordering, nonvs, sci_force_enable }
427			See Documentation/power/video.txt for information on
428			s3_bios and s3_mode.
429			s3_beep is for debugging; it makes the PC's speaker beep
430			as soon as the kernel's real-mode entry point is called.
431			s4_nohwsig prevents ACPI hardware signature from being
432			used during resume from hibernation.
433			old_ordering causes the ACPI 1.0 ordering of the _PTS
434			control method, with respect to putting devices into
435			low power states, to be enforced (the ACPI 2.0 ordering
436			of _PTS is used by default).
437			nonvs prevents the kernel from saving/restoring the
438			ACPI NVS memory during suspend/hibernation and resume.
439			sci_force_enable causes the kernel to set SCI_EN directly
440			on resume from S1/S3 (which is against the ACPI spec,
441			but some broken systems don't work without it).
442
443	acpi_use_timer_override [HW,ACPI]
444			Use timer override. For some broken Nvidia NF5 boards
445			that require a timer override, but don't have HPET
446
447	add_efi_memmap	[EFI; X86] Include EFI memory map in
448			kernel's map of available physical RAM.
449
450	agp=		[AGP]
451			{ off | try_unsupported }
452			off: disable AGP support
453			try_unsupported: try to drive unsupported chipsets
454				(may crash computer or cause data corruption)
455
456	ALSA		[HW,ALSA]
457			See Documentation/sound/alsa/alsa-parameters.txt
458
459	alignment=	[KNL,ARM]
460			Allow the default userspace alignment fault handler
461			behaviour to be specified.  Bit 0 enables warnings,
462			bit 1 enables fixups, and bit 2 sends a segfault.
463
464	align_va_addr=	[X86-64]
465			Align virtual addresses by clearing slice [14:12] when
466			allocating a VMA at process creation time. This option
467			gives you up to 3% performance improvement on AMD F15h
468			machines (where it is enabled by default) for a
469			CPU-intensive style benchmark, and it can vary highly in
470			a microbenchmark depending on workload and compiler.
471
472			32: only for 32-bit processes
473			64: only for 64-bit processes
474			on: enable for both 32- and 64-bit processes
475			off: disable for both 32- and 64-bit processes
476
477	alloc_snapshot	[FTRACE]
478			Allocate the ftrace snapshot buffer on boot up when the
479			main buffer is allocated. This is handy if debugging
480			and you need to use tracing_snapshot() on boot up, and
481			do not want to use tracing_snapshot_alloc() as it needs
482			to be done where GFP_KERNEL allocations are allowed.
483
484	amd_iommu=	[HW,X86-64]
485			Pass parameters to the AMD IOMMU driver in the system.
486			Possible values are:
487			fullflush - enable flushing of IO/TLB entries when
488				    they are unmapped. Otherwise they are
489				    flushed before they will be reused, which
490				    is a lot of faster
491			off	  - do not initialize any AMD IOMMU found in
492				    the system
493			force_isolation - Force device isolation for all
494					  devices. The IOMMU driver is not
495					  allowed anymore to lift isolation
496					  requirements as needed. This option
497					  does not override iommu=pt
498
499	amd_iommu_dump=	[HW,X86-64]
500			Enable AMD IOMMU driver option to dump the ACPI table
501			for AMD IOMMU. With this option enabled, AMD IOMMU
502			driver will print ACPI tables for AMD IOMMU during
503			IOMMU initialization.
504
505	amd_iommu_intr=	[HW,X86-64]
506			Specifies one of the following AMD IOMMU interrupt
507			remapping modes:
508			legacy     - Use legacy interrupt remapping mode.
509			vapic      - Use virtual APIC mode, which allows IOMMU
510			             to inject interrupts directly into guest.
511			             This mode requires kvm-amd.avic=1.
512			             (Default when IOMMU HW support is present.)
513
514	amijoy.map=	[HW,JOY] Amiga joystick support
515			Map of devices attached to JOY0DAT and JOY1DAT
516			Format: <a>,<b>
517			See also Documentation/input/joystick.txt
518
519	analog.map=	[HW,JOY] Analog joystick and gamepad support
520			Specifies type or capabilities of an analog joystick
521			connected to one of 16 gameports
522			Format: <type1>,<type2>,..<type16>
523
524	apc=		[HW,SPARC]
525			Power management functions (SPARCstation-4/5 + deriv.)
526			Format: noidle
527			Disable APC CPU standby support. SPARCstation-Fox does
528			not play well with APC CPU idle - disable it if you have
529			APC and your system crashes randomly.
530
531	apic=		[APIC,X86-32] Advanced Programmable Interrupt Controller
532			Change the output verbosity whilst booting
533			Format: { quiet (default) | verbose | debug }
534			Change the amount of debugging information output
535			when initialising the APIC and IO-APIC components.
536
537	apic_extnmi=	[APIC,X86] External NMI delivery setting
538			Format: { bsp (default) | all | none }
539			bsp:  External NMI is delivered only to CPU 0
540			all:  External NMIs are broadcast to all CPUs as a
541			      backup of CPU 0
542			none: External NMI is masked for all CPUs. This is
543			      useful so that a dump capture kernel won't be
544			      shot down by NMI
545
546	autoconf=	[IPV6]
547			See Documentation/networking/ipv6.txt.
548
549	show_lapic=	[APIC,X86] Advanced Programmable Interrupt Controller
550			Limit apic dumping. The parameter defines the maximal
551			number of local apics being dumped. Also it is possible
552			to set it to "all" by meaning -- no limit here.
553			Format: { 1 (default) | 2 | ... | all }.
554			The parameter valid if only apic=debug or
555			apic=verbose is specified.
556			Example: apic=debug show_lapic=all
557
558	apm=		[APM] Advanced Power Management
559			See header of arch/x86/kernel/apm_32.c.
560
561	arcrimi=	[HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
562			Format: <io>,<irq>,<nodeID>
563
564	ataflop=	[HW,M68k]
565
566	atarimouse=	[HW,MOUSE] Atari Mouse
567
568	atkbd.extra=	[HW] Enable extra LEDs and keys on IBM RapidAccess,
569			EzKey and similar keyboards
570
571	atkbd.reset=	[HW] Reset keyboard during initialization
572
573	atkbd.set=	[HW] Select keyboard code set
574			Format: <int> (2 = AT (default), 3 = PS/2)
575
576	atkbd.scroll=	[HW] Enable scroll wheel on MS Office and similar
577			keyboards
578
579	atkbd.softraw=	[HW] Choose between synthetic and real raw mode
580			Format: <bool> (0 = real, 1 = synthetic (default))
581
582	atkbd.softrepeat= [HW]
583			Use software keyboard repeat
584
585	audit=		[KNL] Enable the audit sub-system
586			Format: { "0" | "1" } (0 = disabled, 1 = enabled)
587			0 - kernel audit is disabled and can not be enabled
588			    until the next reboot
589			unset - kernel audit is initialized but disabled and
590			    will be fully enabled by the userspace auditd.
591			1 - kernel audit is initialized and partially enabled,
592			    storing at most audit_backlog_limit messages in
593			    RAM until it is fully enabled by the userspace
594			    auditd.
595			Default: unset
596
597	audit_backlog_limit= [KNL] Set the audit queue size limit.
598			Format: <int> (must be >=0)
599			Default: 64
600
601	bau=		[X86_UV] Enable the BAU on SGI UV.  The default
602			behavior is to disable the BAU (i.e. bau=0).
603			Format: { "0" | "1" }
604			0 - Disable the BAU.
605			1 - Enable the BAU.
606			unset - Disable the BAU.
607
608	baycom_epp=	[HW,AX25]
609			Format: <io>,<mode>
610
611	baycom_par=	[HW,AX25] BayCom Parallel Port AX.25 Modem
612			Format: <io>,<mode>
613			See header of drivers/net/hamradio/baycom_par.c.
614
615	baycom_ser_fdx=	[HW,AX25]
616			BayCom Serial Port AX.25 Modem (Full Duplex Mode)
617			Format: <io>,<irq>,<mode>[,<baud>]
618			See header of drivers/net/hamradio/baycom_ser_fdx.c.
619
620	baycom_ser_hdx=	[HW,AX25]
621			BayCom Serial Port AX.25 Modem (Half Duplex Mode)
622			Format: <io>,<irq>,<mode>
623			See header of drivers/net/hamradio/baycom_ser_hdx.c.
624
625	blkdevparts=	Manual partition parsing of block device(s) for
626			embedded devices based on command line input.
627			See Documentation/block/cmdline-partition.txt
628
629	boot_delay=	Milliseconds to delay each printk during boot.
630			Values larger than 10 seconds (10000) are changed to
631			no delay (0).
632			Format: integer
633
634	bootmem_debug	[KNL] Enable bootmem allocator debug messages.
635
636	bert_disable	[ACPI]
637			Disable BERT OS support on buggy BIOSes.
638
639	bttv.card=	[HW,V4L] bttv (bt848 + bt878 based grabber cards)
640	bttv.radio=	Most important insmod options are available as
641			kernel args too.
642	bttv.pll=	See Documentation/video4linux/bttv/Insmod-options
643	bttv.tuner=
644
645	bulk_remove=off	[PPC]  This parameter disables the use of the pSeries
646			firmware feature for flushing multiple hpte entries
647			at a time.
648
649	c101=		[NET] Moxa C101 synchronous serial card
650
651	cachesize=	[BUGS=X86-32] Override level 2 CPU cache size detection.
652			Sometimes CPU hardware bugs make them report the cache
653			size incorrectly. The kernel will attempt work arounds
654			to fix known problems, but for some CPUs it is not
655			possible to determine what the correct size should be.
656			This option provides an override for these situations.
657
658	ca_keys=	[KEYS] This parameter identifies a specific key(s) on
659			the system trusted keyring to be used for certificate
660			trust validation.
661			format: { id:<keyid> | builtin }
662
663	cca=		[MIPS] Override the kernel pages' cache coherency
664			algorithm.  Accepted values range from 0 to 7
665			inclusive. See arch/mips/include/asm/pgtable-bits.h
666			for platform specific values (SB1, Loongson3 and
667			others).
668
669	ccw_timeout_log [S390]
670			See Documentation/s390/CommonIO for details.
671
672	cgroup_disable= [KNL] Disable a particular controller
673			Format: {name of the controller(s) to disable}
674			The effects of cgroup_disable=foo are:
675			- foo isn't auto-mounted if you mount all cgroups in
676			  a single hierarchy
677			- foo isn't visible as an individually mountable
678			  subsystem
679			{Currently only "memory" controller deal with this and
680			cut the overhead, others just disable the usage. So
681			only cgroup_disable=memory is actually worthy}
682
683	cgroup_no_v1=	[KNL] Disable one, multiple, all cgroup controllers in v1
684			Format: { controller[,controller...] | "all" }
685			Like cgroup_disable, but only applies to cgroup v1;
686			the blacklisted controllers remain available in cgroup2.
687
688	cgroup.memory=	[KNL] Pass options to the cgroup memory controller.
689			Format: <string>
690			nosocket -- Disable socket memory accounting.
691			nokmem -- Disable kernel memory accounting.
692
693	checkreqprot	[SELINUX] Set initial checkreqprot flag value.
694			Format: { "0" | "1" }
695			See security/selinux/Kconfig help text.
696			0 -- check protection applied by kernel (includes
697				any implied execute protection).
698			1 -- check protection requested by application.
699			Default value is set via a kernel config option.
700			Value can be changed at runtime via
701				/selinux/checkreqprot.
702
703	cio_ignore=	[S390]
704			See Documentation/s390/CommonIO for details.
705	clk_ignore_unused
706			[CLK]
707			Prevents the clock framework from automatically gating
708			clocks that have not been explicitly enabled by a Linux
709			device driver but are enabled in hardware at reset or
710			by the bootloader/firmware. Note that this does not
711			force such clocks to be always-on nor does it reserve
712			those clocks in any way. This parameter is useful for
713			debug and development, but should not be needed on a
714			platform with proper driver support.  For more
715			information, see Documentation/clk.txt.
716
717	clock=		[BUGS=X86-32, HW] gettimeofday clocksource override.
718			[Deprecated]
719			Forces specified clocksource (if available) to be used
720			when calculating gettimeofday(). If specified
721			clocksource is not available, it defaults to PIT.
722			Format: { pit | tsc | cyclone | pmtmr }
723
724	clocksource=	Override the default clocksource
725			Format: <string>
726			Override the default clocksource and use the clocksource
727			with the name specified.
728			Some clocksource names to choose from, depending on
729			the platform:
730			[all] jiffies (this is the base, fallback clocksource)
731			[ACPI] acpi_pm
732			[ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
733				pxa_timer,timer3,32k_counter,timer0_1
734			[AVR32] avr32
735			[X86-32] pit,hpet,tsc;
736				scx200_hrt on Geode; cyclone on IBM x440
737			[MIPS] MIPS
738			[PARISC] cr16
739			[S390] tod
740			[SH] SuperH
741			[SPARC64] tick
742			[X86-64] hpet,tsc
743
744	clocksource.arm_arch_timer.evtstrm=
745			[ARM,ARM64]
746			Format: <bool>
747			Enable/disable the eventstream feature of the ARM
748			architected timer so that code using WFE-based polling
749			loops can be debugged more effectively on production
750			systems.
751
752	clocksource.arm_arch_timer.fsl-a008585=
753			[ARM64]
754			Format: <bool>
755			Enable/disable the workaround of Freescale/NXP
756			erratum A-008585.  This can be useful for KVM
757			guests, if the guest device tree doesn't show the
758			erratum.  If unspecified, the workaround is
759			enabled based on the device tree.
760
761	clearcpuid=BITNUM [X86]
762			Disable CPUID feature X for the kernel. See
763			arch/x86/include/asm/cpufeatures.h for the valid bit
764			numbers. Note the Linux specific bits are not necessarily
765			stable over kernel options, but the vendor specific
766			ones should be.
767			Also note that user programs calling CPUID directly
768			or using the feature without checking anything
769			will still see it. This just prevents it from
770			being used by the kernel or shown in /proc/cpuinfo.
771			Also note the kernel might malfunction if you disable
772			some critical bits.
773
774	cma=nn[MG]@[start[MG][-end[MG]]]
775			[ARM,X86,KNL]
776			Sets the size of kernel global memory area for
777			contiguous memory allocations and optionally the
778			placement constraint by the physical address range of
779			memory allocations. A value of 0 disables CMA
780			altogether. For more information, see
781			include/linux/dma-contiguous.h
782
783	cmo_free_hint=	[PPC] Format: { yes | no }
784			Specify whether pages are marked as being inactive
785			when they are freed.  This is used in CMO environments
786			to determine OS memory pressure for page stealing by
787			a hypervisor.
788			Default: yes
789
790	coherent_pool=nn[KMG]	[ARM,KNL]
791			Sets the size of memory pool for coherent, atomic dma
792			allocations, by default set to 256K.
793
794	code_bytes	[X86] How many bytes of object code to print
795			in an oops report.
796			Range: 0 - 8192
797			Default: 64
798
799	com20020=	[HW,NET] ARCnet - COM20020 chipset
800			Format:
801			<io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
802
803	com90io=	[HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
804			Format: <io>[,<irq>]
805
806	com90xx=	[HW,NET]
807			ARCnet - COM90xx chipset (memory-mapped buffers)
808			Format: <io>[,<irq>[,<memstart>]]
809
810	condev=		[HW,S390] console device
811	conmode=
812
813	console=	[KNL] Output console device and options.
814
815		tty<n>	Use the virtual console device <n>.
816
817		ttyS<n>[,options]
818		ttyUSB0[,options]
819			Use the specified serial port.  The options are of
820			the form "bbbbpnf", where "bbbb" is the baud rate,
821			"p" is parity ("n", "o", or "e"), "n" is number of
822			bits, and "f" is flow control ("r" for RTS or
823			omit it).  Default is "9600n8".
824
825			See Documentation/serial-console.txt for more
826			information.  See
827			Documentation/networking/netconsole.txt for an
828			alternative.
829
830		uart[8250],io,<addr>[,options]
831		uart[8250],mmio,<addr>[,options]
832		uart[8250],mmio16,<addr>[,options]
833		uart[8250],mmio32,<addr>[,options]
834		uart[8250],0x<addr>[,options]
835			Start an early, polled-mode console on the 8250/16550
836			UART at the specified I/O port or MMIO address,
837			switching to the matching ttyS device later.
838			MMIO inter-register address stride is either 8-bit
839			(mmio), 16-bit (mmio16), or 32-bit (mmio32).
840			If none of [io|mmio|mmio16|mmio32], <addr> is assumed
841			to be equivalent to 'mmio'. 'options' are specified in
842			the same format described for ttyS above; if unspecified,
843			the h/w is not re-initialized.
844
845		hvc<n>	Use the hypervisor console device <n>. This is for
846			both Xen and PowerPC hypervisors.
847
848                If the device connected to the port is not a TTY but a braille
849                device, prepend "brl," before the device type, for instance
850			console=brl,ttyS0
851		For now, only VisioBraille is supported.
852
853	consoleblank=	[KNL] The console blank (screen saver) timeout in
854			seconds. Defaults to 10*60 = 10mins. A value of 0
855			disables the blank timer.
856
857	coredump_filter=
858			[KNL] Change the default value for
859			/proc/<pid>/coredump_filter.
860			See also Documentation/filesystems/proc.txt.
861
862	cpuidle.off=1	[CPU_IDLE]
863			disable the cpuidle sub-system
864
865	cpu_init_udelay=N
866			[X86] Delay for N microsec between assert and de-assert
867			of APIC INIT to start processors.  This delay occurs
868			on every CPU online, such as boot, and resume from suspend.
869			Default: 10000
870
871	cpcihp_generic=	[HW,PCI] Generic port I/O CompactPCI driver
872			Format:
873			<first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
874
875	crashkernel=size[KMG][@offset[KMG]]
876			[KNL] Using kexec, Linux can switch to a 'crash kernel'
877			upon panic. This parameter reserves the physical
878			memory region [offset, offset + size] for that kernel
879			image. If '@offset' is omitted, then a suitable offset
880			is selected automatically. Check
881			Documentation/kdump/kdump.txt for further details.
882
883	crashkernel=range1:size1[,range2:size2,...][@offset]
884			[KNL] Same as above, but depends on the memory
885			in the running system. The syntax of range is
886			start-[end] where start and end are both
887			a memory unit (amount[KMG]). See also
888			Documentation/kdump/kdump.txt for an example.
889
890	crashkernel=size[KMG],high
891			[KNL, x86_64] range could be above 4G. Allow kernel
892			to allocate physical memory region from top, so could
893			be above 4G if system have more than 4G ram installed.
894			Otherwise memory region will be allocated below 4G, if
895			available.
896			It will be ignored if crashkernel=X is specified.
897	crashkernel=size[KMG],low
898			[KNL, x86_64] range under 4G. When crashkernel=X,high
899			is passed, kernel could allocate physical memory region
900			above 4G, that cause second kernel crash on system
901			that require some amount of low memory, e.g. swiotlb
902			requires at least 64M+32K low memory, also enough extra
903			low memory is needed to make sure DMA buffers for 32-bit
904			devices won't run out. Kernel would try to allocate at
905			at least 256M below 4G automatically.
906			This one let user to specify own low range under 4G
907			for second kernel instead.
908			0: to disable low allocation.
909			It will be ignored when crashkernel=X,high is not used
910			or memory reserved is below 4G.
911
912	cryptomgr.notests
913                        [KNL] Disable crypto self-tests
914
915	cs89x0_dma=	[HW,NET]
916			Format: <dma>
917
918	cs89x0_media=	[HW,NET]
919			Format: { rj45 | aui | bnc }
920
921	dasd=		[HW,NET]
922			See header of drivers/s390/block/dasd_devmap.c.
923
924	db9.dev[2|3]=	[HW,JOY] Multisystem joystick support via parallel port
925			(one device per port)
926			Format: <port#>,<type>
927			See also Documentation/input/joystick-parport.txt
928
929	ddebug_query=   [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
930			time. See Documentation/dynamic-debug-howto.txt for
931			details.  Deprecated, see dyndbg.
932
933	debug		[KNL] Enable kernel debugging (events log level).
934
935	debug_locks_verbose=
936			[KNL] verbose self-tests
937			Format=<0|1>
938			Print debugging info while doing the locking API
939			self-tests.
940			We default to 0 (no extra messages), setting it to
941			1 will print _a lot_ more information - normally
942			only useful to kernel developers.
943
944	debug_objects	[KNL] Enable object debugging
945
946	no_debug_objects
947			[KNL] Disable object debugging
948
949	debug_guardpage_minorder=
950			[KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
951			parameter allows control of the order of pages that will
952			be intentionally kept free (and hence protected) by the
953			buddy allocator. Bigger value increase the probability
954			of catching random memory corruption, but reduce the
955			amount of memory for normal system use. The maximum
956			possible value is MAX_ORDER/2.  Setting this parameter
957			to 1 or 2 should be enough to identify most random
958			memory corruption problems caused by bugs in kernel or
959			driver code when a CPU writes to (or reads from) a
960			random memory location. Note that there exists a class
961			of memory corruptions problems caused by buggy H/W or
962			F/W or by drivers badly programing DMA (basically when
963			memory is written at bus level and the CPU MMU is
964			bypassed) which are not detectable by
965			CONFIG_DEBUG_PAGEALLOC, hence this option will not help
966			tracking down these problems.
967
968	debug_pagealloc=
969			[KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
970			parameter enables the feature at boot time. In
971			default, it is disabled. We can avoid allocating huge
972			chunk of memory for debug pagealloc if we don't enable
973			it at boot time and the system will work mostly same
974			with the kernel built without CONFIG_DEBUG_PAGEALLOC.
975			on: enable the feature
976
977	debugpat	[X86] Enable PAT debugging
978
979	decnet.addr=	[HW,NET]
980			Format: <area>[,<node>]
981			See also Documentation/networking/decnet.txt.
982
983	default_hugepagesz=
984			[same as hugepagesz=] The size of the default
985			HugeTLB page size. This is the size represented by
986			the legacy /proc/ hugepages APIs, used for SHM, and
987			default size when mounting hugetlbfs filesystems.
988			Defaults to the default architecture's huge page size
989			if not specified.
990
991	dhash_entries=	[KNL]
992			Set number of hash buckets for dentry cache.
993
994	disable_1tb_segments [PPC]
995			Disables the use of 1TB hash page table segments. This
996			causes the kernel to fall back to 256MB segments which
997			can be useful when debugging issues that require an SLB
998			miss to occur.
999
1000	disable=	[IPV6]
1001			See Documentation/networking/ipv6.txt.
1002
1003	disable_radix	[PPC]
1004			Disable RADIX MMU mode on POWER9
1005
1006	disable_cpu_apicid= [X86,APIC,SMP]
1007			Format: <int>
1008			The number of initial APIC ID for the
1009			corresponding CPU to be disabled at boot,
1010			mostly used for the kdump 2nd kernel to
1011			disable BSP to wake up multiple CPUs without
1012			causing system reset or hang due to sending
1013			INIT from AP to BSP.
1014
1015	disable_ddw     [PPC/PSERIES]
1016			Disable Dynamic DMA Window support. Use this if
1017			to workaround buggy firmware.
1018
1019	disable_ipv6=	[IPV6]
1020			See Documentation/networking/ipv6.txt.
1021
1022	disable_mtrr_cleanup [X86]
1023			The kernel tries to adjust MTRR layout from continuous
1024			to discrete, to make X server driver able to add WB
1025			entry later. This parameter disables that.
1026
1027	disable_mtrr_trim [X86, Intel and AMD only]
1028			By default the kernel will trim any uncacheable
1029			memory out of your available memory pool based on
1030			MTRR settings.  This parameter disables that behavior,
1031			possibly causing your machine to run very slowly.
1032
1033	disable_timer_pin_1 [X86]
1034			Disable PIN 1 of APIC timer
1035			Can be useful to work around chipset bugs.
1036
1037	dis_ucode_ldr	[X86] Disable the microcode loader.
1038
1039	dm=		[DM] Allows early creation of a device-mapper device.
1040			See Documentation/device-mapper/boot.txt.
1041
1042	dmasound=	[HW,OSS] Sound subsystem buff
1043
1044	dma_debug=off	If the kernel is compiled with DMA_API_DEBUG support,
1045			this option disables the debugging code at boot.
1046
1047	dma_debug_entries=<number>
1048			This option allows to tune the number of preallocated
1049			entries for DMA-API debugging code. One entry is
1050			required per DMA-API allocation. Use this if the
1051			DMA-API debugging code disables itself because the
1052			architectural default is too low.
1053
1054	dma_debug_driver=<driver_name>
1055			With this option the DMA-API debugging driver
1056			filter feature can be enabled at boot time. Just
1057			pass the driver to filter for as the parameter.
1058			The filter can be disabled or changed to another
1059			driver later using sysfs.
1060
1061	drm_kms_helper.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
1062			Broken monitors, graphic adapters, KVMs and EDIDless
1063			panels may send no or incorrect EDID data sets.
1064			This parameter allows to specify an EDID data sets
1065			in the /lib/firmware directory that are used instead.
1066			Generic built-in EDID data sets are used, if one of
1067			edid/1024x768.bin, edid/1280x1024.bin,
1068			edid/1680x1050.bin, or edid/1920x1080.bin is given
1069			and no file with the same name exists. Details and
1070			instructions how to build your own EDID data are
1071			available in Documentation/EDID/HOWTO.txt. An EDID
1072			data set will only be used for a particular connector,
1073			if its name and a colon are prepended to the EDID
1074			name. Each connector may use a unique EDID data
1075			set by separating the files with a comma.  An EDID
1076			data set with no connector name will be used for
1077			any connectors not explicitly specified.
1078
1079	dscc4.setup=	[NET]
1080
1081	dyndbg[="val"]		[KNL,DYNAMIC_DEBUG]
1082	module.dyndbg[="val"]
1083			Enable debug messages at boot time.  See
1084			Documentation/dynamic-debug-howto.txt for details.
1085
1086	nompx		[X86] Disables Intel Memory Protection Extensions.
1087			See Documentation/x86/intel_mpx.txt for more
1088			information about the feature.
1089
1090	nopku		[X86] Disable Memory Protection Keys CPU feature found
1091			in some Intel CPUs.
1092
1093	eagerfpu=	[X86]
1094			on	enable eager fpu restore
1095			off	disable eager fpu restore
1096			auto	selects the default scheme, which automatically
1097				enables eagerfpu restore for xsaveopt.
1098
1099	module.async_probe [KNL]
1100			Enable asynchronous probe on this module.
1101
1102	early_ioremap_debug [KNL]
1103			Enable debug messages in early_ioremap support. This
1104			is useful for tracking down temporary early mappings
1105			which are not unmapped.
1106
1107	earlycon=	[KNL] Output early console device and options.
1108
1109			When used with no options, the early console is
1110			determined by the stdout-path property in device
1111			tree's chosen node.
1112
1113		cdns,<addr>[,options]
1114			Start an early, polled-mode console on a Cadence
1115			(xuartps) serial port at the specified address. Only
1116			supported option is baud rate. If baud rate is not
1117			specified, the serial port must already be setup and
1118			configured.
1119
1120		uart[8250],io,<addr>[,options]
1121		uart[8250],mmio,<addr>[,options]
1122		uart[8250],mmio32,<addr>[,options]
1123		uart[8250],mmio32be,<addr>[,options]
1124		uart[8250],0x<addr>[,options]
1125			Start an early, polled-mode console on the 8250/16550
1126			UART at the specified I/O port or MMIO address.
1127			MMIO inter-register address stride is either 8-bit
1128			(mmio) or 32-bit (mmio32 or mmio32be).
1129			If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1130			to be equivalent to 'mmio'. 'options' are specified
1131			in the same format described for "console=ttyS<n>"; if
1132			unspecified, the h/w is not initialized.
1133
1134		pl011,<addr>
1135		pl011,mmio32,<addr>
1136			Start an early, polled-mode console on a pl011 serial
1137			port at the specified address. The pl011 serial port
1138			must already be setup and configured. Options are not
1139			yet supported.  If 'mmio32' is specified, then only
1140			the driver will use only 32-bit accessors to read/write
1141			the device registers.
1142
1143		meson,<addr>
1144			Start an early, polled-mode console on a meson serial
1145			port at the specified address. The serial port must
1146			already be setup and configured. Options are not yet
1147			supported.
1148
1149		msm_serial,<addr>
1150			Start an early, polled-mode console on an msm serial
1151			port at the specified address. The serial port
1152			must already be setup and configured. Options are not
1153			yet supported.
1154
1155		msm_serial_dm,<addr>
1156			Start an early, polled-mode console on an msm serial
1157			dm port at the specified address. The serial port
1158			must already be setup and configured. Options are not
1159			yet supported.
1160
1161		smh	Use ARM semihosting calls for early console.
1162
1163		s3c2410,<addr>
1164		s3c2412,<addr>
1165		s3c2440,<addr>
1166		s3c6400,<addr>
1167		s5pv210,<addr>
1168		exynos4210,<addr>
1169			Use early console provided by serial driver available
1170			on Samsung SoCs, requires selecting proper type and
1171			a correct base address of the selected UART port. The
1172			serial port must already be setup and configured.
1173			Options are not yet supported.
1174
1175		lpuart,<addr>
1176		lpuart32,<addr>
1177			Use early console provided by Freescale LP UART driver
1178			found on Freescale Vybrid and QorIQ LS1021A processors.
1179			A valid base address must be provided, and the serial
1180			port must already be setup and configured.
1181
1182		armada3700_uart,<addr>
1183			Start an early, polled-mode console on the
1184			Armada 3700 serial port at the specified
1185			address. The serial port must already be setup
1186			and configured. Options are not yet supported.
1187
1188	earlyprintk=	[X86,SH,BLACKFIN,ARM,M68k]
1189			earlyprintk=vga
1190			earlyprintk=efi
1191			earlyprintk=xen
1192			earlyprintk=serial[,ttySn[,baudrate]]
1193			earlyprintk=serial[,0x...[,baudrate]]
1194			earlyprintk=ttySn[,baudrate]
1195			earlyprintk=dbgp[debugController#]
1196			earlyprintk=pciserial,bus:device.function[,baudrate]
1197
1198			earlyprintk is useful when the kernel crashes before
1199			the normal console is initialized. It is not enabled by
1200			default because it has some cosmetic problems.
1201
1202			Append ",keep" to not disable it when the real console
1203			takes over.
1204
1205			Only one of vga, efi, serial, or usb debug port can
1206			be used at a time.
1207
1208			Currently only ttyS0 and ttyS1 may be specified by
1209			name.  Other I/O ports may be explicitly specified
1210			on some architectures (x86 and arm at least) by
1211			replacing ttySn with an I/O port address, like this:
1212				earlyprintk=serial,0x1008,115200
1213			You can find the port for a given device in
1214			/proc/tty/driver/serial:
1215				2: uart:ST16650V2 port:00001008 irq:18 ...
1216
1217			Interaction with the standard serial driver is not
1218			very good.
1219
1220			The VGA and EFI output is eventually overwritten by
1221			the real console.
1222
1223			The xen output can only be used by Xen PV guests.
1224
1225	edac_report=	[HW,EDAC] Control how to report EDAC event
1226			Format: {"on" | "off" | "force"}
1227			on: enable EDAC to report H/W event. May be overridden
1228			by other higher priority error reporting module.
1229			off: disable H/W event reporting through EDAC.
1230			force: enforce the use of EDAC to report H/W event.
1231			default: on.
1232
1233	ekgdboc=	[X86,KGDB] Allow early kernel console debugging
1234			ekgdboc=kbd
1235
1236			This is designed to be used in conjunction with
1237			the boot argument: earlyprintk=vga
1238
1239	edd=		[EDD]
1240			Format: {"off" | "on" | "skip[mbr]"}
1241
1242	efi=		[EFI]
1243			Format: { "old_map", "nochunk", "noruntime", "debug" }
1244			old_map [X86-64]: switch to the old ioremap-based EFI
1245			runtime services mapping. 32-bit still uses this one by
1246			default.
1247			nochunk: disable reading files in "chunks" in the EFI
1248			boot stub, as chunking can cause problems with some
1249			firmware implementations.
1250			noruntime : disable EFI runtime services support
1251			debug: enable misc debug output
1252
1253	efi_no_storage_paranoia [EFI; X86]
1254			Using this parameter you can use more than 50% of
1255			your efi variable storage. Use this parameter only if
1256			you are really sure that your UEFI does sane gc and
1257			fulfills the spec otherwise your board may brick.
1258
1259	efi_fake_mem=	nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1260			Add arbitrary attribute to specific memory range by
1261			updating original EFI memory map.
1262			Region of memory which aa attribute is added to is
1263			from ss to ss+nn.
1264			If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1265			is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1266			attribute is added to range 0x100000000-0x180000000 and
1267			0x10a0000000-0x1120000000.
1268
1269			Using this parameter you can do debugging of EFI memmap
1270			related feature. For example, you can do debugging of
1271			Address Range Mirroring feature even if your box
1272			doesn't support it.
1273
1274	efivar_ssdt=	[EFI; X86] Name of an EFI variable that contains an SSDT
1275			that is to be dynamically loaded by Linux. If there are
1276			multiple variables with the same name but with different
1277			vendor GUIDs, all of them will be loaded. See
1278			Documentation/acpi/ssdt-overlays.txt for details.
1279
1280
1281	eisa_irq_edge=	[PARISC,HW]
1282			See header of drivers/parisc/eisa.c.
1283
1284	elanfreq=	[X86-32]
1285			See comment before function elanfreq_setup() in
1286			arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1287
1288	elevator=	[IOSCHED]
1289			Format: {"cfq" | "deadline" | "noop"}
1290			See Documentation/block/cfq-iosched.txt and
1291			Documentation/block/deadline-iosched.txt for details.
1292
1293	elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1294			Specifies physical address of start of kernel core
1295			image elf header and optionally the size. Generally
1296			kexec loader will pass this option to capture kernel.
1297			See Documentation/kdump/kdump.txt for details.
1298
1299	enable_mtrr_cleanup [X86]
1300			The kernel tries to adjust MTRR layout from continuous
1301			to discrete, to make X server driver able to add WB
1302			entry later. This parameter enables that.
1303
1304	enable_timer_pin_1 [X86]
1305			Enable PIN 1 of APIC timer
1306			Can be useful to work around chipset bugs
1307			(in particular on some ATI chipsets).
1308			The kernel tries to set a reasonable default.
1309
1310	enforcing	[SELINUX] Set initial enforcing status.
1311			Format: {"0" | "1"}
1312			See security/selinux/Kconfig help text.
1313			0 -- permissive (log only, no denials).
1314			1 -- enforcing (deny and log).
1315			Default value is 0.
1316			Value can be changed at runtime via /selinux/enforce.
1317
1318	erst_disable	[ACPI]
1319			Disable Error Record Serialization Table (ERST)
1320			support.
1321
1322	ether=		[HW,NET] Ethernet cards parameters
1323			This option is obsoleted by the "netdev=" option, which
1324			has equivalent usage. See its documentation for details.
1325
1326	evm=		[EVM]
1327			Format: { "fix" }
1328			Permit 'security.evm' to be updated regardless of
1329			current integrity status.
1330
1331	failslab=
1332	fail_page_alloc=
1333	fail_make_request=[KNL]
1334			General fault injection mechanism.
1335			Format: <interval>,<probability>,<space>,<times>
1336			See also Documentation/fault-injection/.
1337
1338	floppy=		[HW]
1339			See Documentation/blockdev/floppy.txt.
1340
1341	force_pal_cache_flush
1342			[IA-64] Avoid check_sal_cache_flush which may hang on
1343			buggy SAL_CACHE_FLUSH implementations. Using this
1344			parameter will force ia64_sal_cache_flush to call
1345			ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1346
1347	forcepae [X86-32]
1348			Forcefully enable Physical Address Extension (PAE).
1349			Many Pentium M systems disable PAE but may have a
1350			functionally usable PAE implementation.
1351			Warning: use of this parameter will taint the kernel
1352			and may cause unknown problems.
1353
1354	ftrace=[tracer]
1355			[FTRACE] will set and start the specified tracer
1356			as early as possible in order to facilitate early
1357			boot debugging.
1358
1359	ftrace_dump_on_oops[=orig_cpu]
1360			[FTRACE] will dump the trace buffers on oops.
1361			If no parameter is passed, ftrace will dump
1362			buffers of all CPUs, but if you pass orig_cpu, it will
1363			dump only the buffer of the CPU that triggered the
1364			oops.
1365
1366	ftrace_filter=[function-list]
1367			[FTRACE] Limit the functions traced by the function
1368			tracer at boot up. function-list is a comma separated
1369			list of functions. This list can be changed at run
1370			time by the set_ftrace_filter file in the debugfs
1371			tracing directory.
1372
1373	ftrace_notrace=[function-list]
1374			[FTRACE] Do not trace the functions specified in
1375			function-list. This list can be changed at run time
1376			by the set_ftrace_notrace file in the debugfs
1377			tracing directory.
1378
1379	ftrace_graph_filter=[function-list]
1380			[FTRACE] Limit the top level callers functions traced
1381			by the function graph tracer at boot up.
1382			function-list is a comma separated list of functions
1383			that can be changed at run time by the
1384			set_graph_function file in the debugfs tracing directory.
1385
1386	ftrace_graph_notrace=[function-list]
1387			[FTRACE] Do not trace from the functions specified in
1388			function-list.  This list is a comma separated list of
1389			functions that can be changed at run time by the
1390			set_graph_notrace file in the debugfs tracing directory.
1391
1392	gamecon.map[2|3]=
1393			[HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1394			support via parallel port (up to 5 devices per port)
1395			Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1396			See also Documentation/input/joystick-parport.txt
1397
1398	gamma=		[HW,DRM]
1399
1400	gart_fix_e820=  [X86_64] disable the fix e820 for K8 GART
1401			Format: off | on
1402			default: on
1403
1404	gcov_persist=	[GCOV] When non-zero (default), profiling data for
1405			kernel modules is saved and remains accessible via
1406			debugfs, even when the module is unloaded/reloaded.
1407			When zero, profiling data is discarded and associated
1408			debugfs files are removed at module unload time.
1409
1410	goldfish	[X86] Enable the goldfish android emulator platform.
1411			Don't use this when you are not running on the
1412			android emulator
1413
1414	gpt		[EFI] Forces disk with valid GPT signature but
1415			invalid Protective MBR to be treated as GPT. If the
1416			primary GPT is corrupted, it enables the backup/alternate
1417			GPT to be used instead.
1418
1419	grcan.enable0=	[HW] Configuration of physical interface 0. Determines
1420			the "Enable 0" bit of the configuration register.
1421			Format: 0 | 1
1422			Default: 0
1423	grcan.enable1=	[HW] Configuration of physical interface 1. Determines
1424			the "Enable 0" bit of the configuration register.
1425			Format: 0 | 1
1426			Default: 0
1427	grcan.select=	[HW] Select which physical interface to use.
1428			Format: 0 | 1
1429			Default: 0
1430	grcan.txsize=	[HW] Sets the size of the tx buffer.
1431			Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1432			Default: 1024
1433	grcan.rxsize=	[HW] Sets the size of the rx buffer.
1434			Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1435			Default: 1024
1436
1437	gpio-mockup.gpio_mockup_ranges
1438			[HW] Sets the ranges of gpiochip of for this device.
1439			Format: <start1>,<end1>,<start2>,<end2>...
1440
1441	hardlockup_all_cpu_backtrace=
1442			[KNL] Should the hard-lockup detector generate
1443			backtraces on all cpus.
1444			Format: <integer>
1445
1446	hashdist=	[KNL,NUMA] Large hashes allocated during boot
1447			are distributed across NUMA nodes.  Defaults on
1448			for 64-bit NUMA, off otherwise.
1449			Format: 0 | 1 (for off | on)
1450
1451	hcl=		[IA-64] SGI's Hardware Graph compatibility layer
1452
1453	hd=		[EIDE] (E)IDE hard drive subsystem geometry
1454			Format: <cyl>,<head>,<sect>
1455
1456	hest_disable	[ACPI]
1457			Disable Hardware Error Source Table (HEST) support;
1458			corresponding firmware-first mode error processing
1459			logic will be disabled.
1460
1461	highmem=nn[KMG]	[KNL,BOOT] forces the highmem zone to have an exact
1462			size of <nn>. This works even on boxes that have no
1463			highmem otherwise. This also works to reduce highmem
1464			size on bigger boxes.
1465
1466	highres=	[KNL] Enable/disable high resolution timer mode.
1467			Valid parameters: "on", "off"
1468			Default: "on"
1469
1470	hisax=		[HW,ISDN]
1471			See Documentation/isdn/README.HiSax.
1472
1473	hlt		[BUGS=ARM,SH]
1474
1475	hpet=		[X86-32,HPET] option to control HPET usage
1476			Format: { enable (default) | disable | force |
1477				verbose }
1478			disable: disable HPET and use PIT instead
1479			force: allow force enabled of undocumented chips (ICH4,
1480				VIA, nVidia)
1481			verbose: show contents of HPET registers during setup
1482
1483	hpet_mmap=	[X86, HPET_MMAP] Allow userspace to mmap HPET
1484			registers.  Default set by CONFIG_HPET_MMAP_DEFAULT.
1485
1486	hugepages=	[HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1487	hugepagesz=	[HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1488			On x86-64 and powerpc, this option can be specified
1489			multiple times interleaved with hugepages= to reserve
1490			huge pages of different sizes. Valid pages sizes on
1491			x86-64 are 2M (when the CPU supports "pse") and 1G
1492			(when the CPU supports the "pdpe1gb" cpuinfo flag).
1493
1494	hvc_iucv=	[S390] Number of z/VM IUCV hypervisor console (HVC)
1495			       terminal devices. Valid values: 0..8
1496	hvc_iucv_allow=	[S390] Comma-separated list of z/VM user IDs.
1497			       If specified, z/VM IUCV HVC accepts connections
1498			       from listed z/VM user IDs only.
1499
1500	hwthread_map=	[METAG] Comma-separated list of Linux cpu id to
1501			        hardware thread id mappings.
1502				Format: <cpu>:<hwthread>
1503
1504	keep_bootcon	[KNL]
1505			Do not unregister boot console at start. This is only
1506			useful for debugging when something happens in the window
1507			between unregistering the boot console and initializing
1508			the real console.
1509
1510	i2c_bus=	[HW] Override the default board specific I2C bus speed
1511			     or register an additional I2C bus that is not
1512			     registered from board initialization code.
1513			     Format:
1514			     <bus_id>,<clkrate>
1515
1516	i8042.debug	[HW] Toggle i8042 debug mode
1517	i8042.unmask_kbd_data
1518			[HW] Enable printing of interrupt data from the KBD port
1519			     (disabled by default, and as a pre-condition
1520			     requires that i8042.debug=1 be enabled)
1521	i8042.direct	[HW] Put keyboard port into non-translated mode
1522	i8042.dumbkbd	[HW] Pretend that controller can only read data from
1523			     keyboard and cannot control its state
1524			     (Don't attempt to blink the leds)
1525	i8042.noaux	[HW] Don't check for auxiliary (== mouse) port
1526	i8042.nokbd	[HW] Don't check/create keyboard port
1527	i8042.noloop	[HW] Disable the AUX Loopback command while probing
1528			     for the AUX port
1529	i8042.nomux	[HW] Don't check presence of an active multiplexing
1530			     controller
1531	i8042.nopnp	[HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1532			     controllers
1533	i8042.notimeout	[HW] Ignore timeout condition signalled by controller
1534	i8042.reset	[HW] Reset the controller during init, cleanup and
1535			     suspend-to-ram transitions, only during s2r
1536			     transitions, or never reset
1537			Format: { 1 | Y | y | 0 | N | n }
1538			1, Y, y: always reset controller
1539			0, N, n: don't ever reset controller
1540			Default: only on s2r transitions on x86; most other
1541			architectures force reset to be always executed
1542	i8042.unlock	[HW] Unlock (ignore) the keylock
1543	i8042.kbdreset  [HW] Reset device connected to KBD port
1544
1545	i810=		[HW,DRM]
1546
1547	i8k.ignore_dmi	[HW] Continue probing hardware even if DMI data
1548			indicates that the driver is running on unsupported
1549			hardware.
1550	i8k.force	[HW] Activate i8k driver even if SMM BIOS signature
1551			does not match list of supported models.
1552	i8k.power_status
1553			[HW] Report power status in /proc/i8k
1554			(disabled by default)
1555	i8k.restricted	[HW] Allow controlling fans only if SYS_ADMIN
1556			capability is set.
1557
1558	i915.invert_brightness=
1559			[DRM] Invert the sense of the variable that is used to
1560			set the brightness of the panel backlight. Normally a
1561			brightness value of 0 indicates backlight switched off,
1562			and the maximum of the brightness value sets the backlight
1563			to maximum brightness. If this parameter is set to 0
1564			(default) and the machine requires it, or this parameter
1565			is set to 1, a brightness value of 0 sets the backlight
1566			to maximum brightness, and the maximum of the brightness
1567			value switches the backlight off.
1568			-1 -- never invert brightness
1569			 0 -- machine default
1570			 1 -- force brightness inversion
1571
1572	icn=		[HW,ISDN]
1573			Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1574
1575	ide-core.nodma=	[HW] (E)IDE subsystem
1576			Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1577			.vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1578			.cdrom .chs .ignore_cable are additional options
1579			See Documentation/ide/ide.txt.
1580
1581	ide-generic.probe-mask= [HW] (E)IDE subsystem
1582			Format: <int>
1583			Probe mask for legacy ISA IDE ports.  Depending on
1584			platform up to 6 ports are supported, enabled by
1585			setting corresponding bits in the mask to 1.  The
1586			default value is 0x0, which has a special meaning.
1587			On systems that have PCI, it triggers scanning the
1588			PCI bus for the first and the second port, which
1589			are then probed.  On systems without PCI the value
1590			of 0x0 enables probing the two first ports as if it
1591			was 0x3.
1592
1593	ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1594			Claim all unknown PCI IDE storage controllers.
1595
1596	idle=		[X86]
1597			Format: idle=poll, idle=halt, idle=nomwait
1598			Poll forces a polling idle loop that can slightly
1599			improve the performance of waking up a idle CPU, but
1600			will use a lot of power and make the system run hot.
1601			Not recommended.
1602			idle=halt: Halt is forced to be used for CPU idle.
1603			In such case C2/C3 won't be used again.
1604			idle=nomwait: Disable mwait for CPU C-states
1605
1606	ieee754=	[MIPS] Select IEEE Std 754 conformance mode
1607			Format: { strict | legacy | 2008 | relaxed }
1608			Default: strict
1609
1610			Choose which programs will be accepted for execution
1611			based on the IEEE 754 NaN encoding(s) supported by
1612			the FPU and the NaN encoding requested with the value
1613			of an ELF file header flag individually set by each
1614			binary.  Hardware implementations are permitted to
1615			support either or both of the legacy and the 2008 NaN
1616			encoding mode.
1617
1618			Available settings are as follows:
1619			strict	accept binaries that request a NaN encoding
1620				supported by the FPU
1621			legacy	only accept legacy-NaN binaries, if supported
1622				by the FPU
1623			2008	only accept 2008-NaN binaries, if supported
1624				by the FPU
1625			relaxed	accept any binaries regardless of whether
1626				supported by the FPU
1627
1628			The FPU emulator is always able to support both NaN
1629			encodings, so if no FPU hardware is present or it has
1630			been disabled with 'nofpu', then the settings of
1631			'legacy' and '2008' strap the emulator accordingly,
1632			'relaxed' straps the emulator for both legacy-NaN and
1633			2008-NaN, whereas 'strict' enables legacy-NaN only on
1634			legacy processors and both NaN encodings on MIPS32 or
1635			MIPS64 CPUs.
1636
1637			The setting for ABS.fmt/NEG.fmt instruction execution
1638			mode generally follows that for the NaN encoding,
1639			except where unsupported by hardware.
1640
1641	ignore_loglevel	[KNL]
1642			Ignore loglevel setting - this will print /all/
1643			kernel messages to the console. Useful for debugging.
1644			We also add it as printk module parameter, so users
1645			could change it dynamically, usually by
1646			/sys/module/printk/parameters/ignore_loglevel.
1647
1648	ignore_rlimit_data
1649			Ignore RLIMIT_DATA setting for data mappings,
1650			print warning at first misuse.  Can be changed via
1651			/sys/module/kernel/parameters/ignore_rlimit_data.
1652
1653	ihash_entries=	[KNL]
1654			Set number of hash buckets for inode cache.
1655
1656	ima_appraise=	[IMA] appraise integrity measurements
1657			Format: { "off" | "enforce" | "fix" | "log" }
1658			default: "enforce"
1659
1660	ima_appraise_tcb [IMA]
1661			The builtin appraise policy appraises all files
1662			owned by uid=0.
1663
1664	ima_hash=	[IMA]
1665			Format: { md5 | sha1 | rmd160 | sha256 | sha384
1666				   | sha512 | ... }
1667			default: "sha1"
1668
1669			The list of supported hash algorithms is defined
1670			in crypto/hash_info.h.
1671
1672	ima_policy=	[IMA]
1673			The builtin measurement policy to load during IMA
1674			setup.  Specyfing "tcb" as the value, measures all
1675			programs exec'd, files mmap'd for exec, and all files
1676			opened with the read mode bit set by either the
1677			effective uid (euid=0) or uid=0.
1678			Format: "tcb"
1679
1680	ima_tcb		[IMA] Deprecated.  Use ima_policy= instead.
1681			Load a policy which meets the needs of the Trusted
1682			Computing Base.  This means IMA will measure all
1683			programs exec'd, files mmap'd for exec, and all files
1684			opened for read by uid=0.
1685
1686	ima_template=   [IMA]
1687			Select one of defined IMA measurements template formats.
1688			Formats: { "ima" | "ima-ng" | "ima-sig" }
1689			Default: "ima-ng"
1690
1691	ima_template_fmt=
1692	                [IMA] Define a custom template format.
1693			Format: { "field1|...|fieldN" }
1694
1695	ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1696			Format: <min_file_size>
1697			Set the minimal file size for using asynchronous hash.
1698			If left unspecified, ahash usage is disabled.
1699
1700			ahash performance varies for different data sizes on
1701			different crypto accelerators. This option can be used
1702			to achieve the best performance for a particular HW.
1703
1704	ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1705			Format: <bufsize>
1706			Set hashing buffer size. Default: 4k.
1707
1708			ahash performance varies for different chunk sizes on
1709			different crypto accelerators. This option can be used
1710			to achieve best performance for particular HW.
1711
1712	init=		[KNL]
1713			Format: <full_path>
1714			Run specified binary instead of /sbin/init as init
1715			process.
1716
1717	initcall_debug	[KNL] Trace initcalls as they are executed.  Useful
1718			for working out where the kernel is dying during
1719			startup.
1720
1721	initcall_blacklist=  [KNL] Do not execute a comma-separated list of
1722			initcall functions.  Useful for debugging built-in
1723			modules and initcalls.
1724
1725	initrd=		[BOOT] Specify the location of the initial ramdisk
1726
1727	init_pkru=	[x86] Specify the default memory protection keys rights
1728			register contents for all processes.  0x55555554 by
1729			default (disallow access to all but pkey 0).  Can
1730			override in debugfs after boot.
1731
1732	inport.irq=	[HW] Inport (ATI XL and Microsoft) busmouse driver
1733			Format: <irq>
1734
1735	int_pln_enable  [x86] Enable power limit notification interrupt
1736
1737	integrity_audit=[IMA]
1738			Format: { "0" | "1" }
1739			0 -- basic integrity auditing messages. (Default)
1740			1 -- additional integrity auditing messages.
1741
1742	intel_iommu=	[DMAR] Intel IOMMU driver (DMAR) option
1743		on
1744			Enable intel iommu driver.
1745		off
1746			Disable intel iommu driver.
1747		igfx_off [Default Off]
1748			By default, gfx is mapped as normal device. If a gfx
1749			device has a dedicated DMAR unit, the DMAR unit is
1750			bypassed by not enabling DMAR with this option. In
1751			this case, gfx device will use physical address for
1752			DMA.
1753		forcedac [x86_64]
1754			With this option iommu will not optimize to look
1755			for io virtual address below 32-bit forcing dual
1756			address cycle on pci bus for cards supporting greater
1757			than 32-bit addressing. The default is to look
1758			for translation below 32-bit and if not available
1759			then look in the higher range.
1760		strict [Default Off]
1761			With this option on every unmap_single operation will
1762			result in a hardware IOTLB flush operation as opposed
1763			to batching them for performance.
1764		sp_off [Default Off]
1765			By default, super page will be supported if Intel IOMMU
1766			has the capability. With this option, super page will
1767			not be supported.
1768		ecs_off [Default Off]
1769			By default, extended context tables will be supported if
1770			the hardware advertises that it has support both for the
1771			extended tables themselves, and also PASID support. With
1772			this option set, extended tables will not be used even
1773			on hardware which claims to support them.
1774
1775	intel_idle.max_cstate=	[KNL,HW,ACPI,X86]
1776			0	disables intel_idle and fall back on acpi_idle.
1777			1 to 9	specify maximum depth of C-state.
1778
1779	intel_pstate=  [X86]
1780		       disable
1781		         Do not enable intel_pstate as the default
1782		         scaling driver for the supported processors
1783		       force
1784			 Enable intel_pstate on systems that prohibit it by default
1785			 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1786			 instead of acpi-cpufreq may disable platform features, such
1787			 as thermal controls and power capping, that rely on ACPI
1788			 P-States information being indicated to OSPM and therefore
1789			 should be used with caution. This option does not work with
1790			 processors that aren't supported by the intel_pstate driver
1791			 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1792		       no_hwp
1793		         Do not enable hardware P state control (HWP)
1794			 if available.
1795		hwp_only
1796			Only load intel_pstate on systems which support
1797			hardware P state control (HWP) if available.
1798		support_acpi_ppc
1799			Enforce ACPI _PPC performance limits. If the Fixed ACPI
1800			Description Table, specifies preferred power management
1801			profile as "Enterprise Server" or "Performance Server",
1802			then this feature is turned on by default.
1803
1804	intremap=	[X86-64, Intel-IOMMU]
1805			on	enable Interrupt Remapping (default)
1806			off	disable Interrupt Remapping
1807			nosid	disable Source ID checking
1808			no_x2apic_optout
1809				BIOS x2APIC opt-out request will be ignored
1810			nopost	disable Interrupt Posting
1811
1812	iomem=		Disable strict checking of access to MMIO memory
1813		strict	regions from userspace.
1814		relaxed
1815
1816	iommu=		[x86]
1817		off
1818		force
1819		noforce
1820		biomerge
1821		panic
1822		nopanic
1823		merge
1824		nomerge
1825		forcesac
1826		soft
1827		pt		[x86, IA-64]
1828		nobypass	[PPC/POWERNV]
1829			Disable IOMMU bypass, using IOMMU for PCI devices.
1830
1831
1832	io7=		[HW] IO7 for Marvel based alpha systems
1833			See comment before marvel_specify_io7 in
1834			arch/alpha/kernel/core_marvel.c.
1835
1836	io_delay=	[X86] I/O delay method
1837		0x80
1838			Standard port 0x80 based delay
1839		0xed
1840			Alternate port 0xed based delay (needed on some systems)
1841		udelay
1842			Simple two microseconds delay
1843		none
1844			No delay
1845
1846	ip=		[IP_PNP]
1847			See Documentation/filesystems/nfs/nfsroot.txt.
1848
1849	irqaffinity=	[SMP] Set the default irq affinity mask
1850			The argument is a cpu list, as described above.
1851
1852	irqfixup	[HW]
1853			When an interrupt is not handled search all handlers
1854			for it. Intended to get systems with badly broken
1855			firmware running.
1856
1857	irqpoll		[HW]
1858			When an interrupt is not handled search all handlers
1859			for it. Also check all handlers each timer
1860			interrupt. Intended to get systems with badly broken
1861			firmware running.
1862
1863	isapnp=		[ISAPNP]
1864			Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1865
1866	isolcpus=	[KNL,SMP] Isolate CPUs from the general scheduler.
1867			The argument is a cpu list, as described above.
1868
1869			This option can be used to specify one or more CPUs
1870			to isolate from the general SMP balancing and scheduling
1871			algorithms. You can move a process onto or off an
1872			"isolated" CPU via the CPU affinity syscalls or cpuset.
1873			<cpu number> begins at 0 and the maximum value is
1874			"number of CPUs in system - 1".
1875
1876			This option is the preferred way to isolate CPUs. The
1877			alternative -- manually setting the CPU mask of all
1878			tasks in the system -- can cause problems and
1879			suboptimal load balancer performance.
1880
1881	iucv=		[HW,NET]
1882
1883	ivrs_ioapic	[HW,X86_64]
1884			Provide an override to the IOAPIC-ID<->DEVICE-ID
1885			mapping provided in the IVRS ACPI table. For
1886			example, to map IOAPIC-ID decimal 10 to
1887			PCI device 00:14.0 write the parameter as:
1888				ivrs_ioapic[10]=00:14.0
1889
1890	ivrs_hpet	[HW,X86_64]
1891			Provide an override to the HPET-ID<->DEVICE-ID
1892			mapping provided in the IVRS ACPI table. For
1893			example, to map HPET-ID decimal 0 to
1894			PCI device 00:14.0 write the parameter as:
1895				ivrs_hpet[0]=00:14.0
1896
1897	ivrs_acpihid	[HW,X86_64]
1898			Provide an override to the ACPI-HID:UID<->DEVICE-ID
1899			mapping provided in the IVRS ACPI table. For
1900			example, to map UART-HID:UID AMD0020:0 to
1901			PCI device 00:14.5 write the parameter as:
1902				ivrs_acpihid[00:14.5]=AMD0020:0
1903
1904	js=		[HW,JOY] Analog joystick
1905			See Documentation/input/joystick.txt.
1906
1907	nokaslr		[KNL]
1908			When CONFIG_RANDOMIZE_BASE is set, this disables
1909			kernel and module base offset ASLR (Address Space
1910			Layout Randomization).
1911
1912	kasan_multi_shot
1913			[KNL] Enforce KASAN (Kernel Address Sanitizer) to print
1914			report on every invalid memory access. Without this
1915			parameter KASAN will print report only for the first
1916			invalid access.
1917
1918	keepinitrd	[HW,ARM]
1919
1920	kernelcore=	[KNL,X86,IA-64,PPC]
1921			Format: nn[KMGTPE] | "mirror"
1922			This parameter
1923			specifies the amount of memory usable by the kernel
1924			for non-movable allocations.  The requested amount is
1925			spread evenly throughout all nodes in the system. The
1926			remaining memory in each node is used for Movable
1927			pages. In the event, a node is too small to have both
1928			kernelcore and Movable pages, kernelcore pages will
1929			take priority and other nodes will have a larger number
1930			of Movable pages.  The Movable zone is used for the
1931			allocation of pages that may be reclaimed or moved
1932			by the page migration subsystem.  This means that
1933			HugeTLB pages may not be allocated from this zone.
1934			Note that allocations like PTEs-from-HighMem still
1935			use the HighMem zone if it exists, and the Normal
1936			zone if it does not.
1937
1938			Instead of specifying the amount of memory (nn[KMGTPE]),
1939			you can specify "mirror" option. In case "mirror"
1940			option is specified, mirrored (reliable) memory is used
1941			for non-movable allocations and remaining memory is used
1942			for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1943			so you can NOT specify nn[KMGTPE] and "mirror" at the same
1944			time.
1945
1946	kgdbdbgp=	[KGDB,HW] kgdb over EHCI usb debug port.
1947			Format: <Controller#>[,poll interval]
1948			The controller # is the number of the ehci usb debug
1949			port as it is probed via PCI.  The poll interval is
1950			optional and is the number seconds in between
1951			each poll cycle to the debug port in case you need
1952			the functionality for interrupting the kernel with
1953			gdb or control-c on the dbgp connection.  When
1954			not using this parameter you use sysrq-g to break into
1955			the kernel debugger.
1956
1957	kgdboc=		[KGDB,HW] kgdb over consoles.
1958			Requires a tty driver that supports console polling,
1959			or a supported polling keyboard driver (non-usb).
1960			 Serial only format: <serial_device>[,baud]
1961			 keyboard only format: kbd
1962			 keyboard and serial format: kbd,<serial_device>[,baud]
1963			Optional Kernel mode setting:
1964			 kms, kbd format: kms,kbd
1965			 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1966
1967	kgdbwait	[KGDB] Stop kernel execution and enter the
1968			kernel debugger at the earliest opportunity.
1969
1970	kmac=		[MIPS] korina ethernet MAC address.
1971			Configure the RouterBoard 532 series on-chip
1972			Ethernet adapter MAC address.
1973
1974	kmemleak=	[KNL] Boot-time kmemleak enable/disable
1975			Valid arguments: on, off
1976			Default: on
1977			Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1978			the default is off.
1979
1980	kmemcheck=	[X86] Boot-time kmemcheck enable/disable/one-shot mode
1981			Valid arguments: 0, 1, 2
1982			kmemcheck=0 (disabled)
1983			kmemcheck=1 (enabled)
1984			kmemcheck=2 (one-shot mode)
1985			Default: 2 (one-shot mode)
1986
1987	kstack=N	[X86] Print N words from the kernel stack
1988			in oops dumps.
1989
1990	kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1991			Default is 0 (don't ignore, but inject #GP)
1992
1993	kvm.mmu_audit=	[KVM] This is a R/W parameter which allows audit
1994			KVM MMU at runtime.
1995			Default is 0 (off)
1996
1997	kvm-amd.nested=	[KVM,AMD] Allow nested virtualization in KVM/SVM.
1998			Default is 1 (enabled)
1999
2000	kvm-amd.npt=	[KVM,AMD] Disable nested paging (virtualized MMU)
2001			for all guests.
2002			Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2003
2004	kvm-intel.ept=	[KVM,Intel] Disable extended page tables
2005			(virtualized MMU) support on capable Intel chips.
2006			Default is 1 (enabled)
2007
2008	kvm-intel.emulate_invalid_guest_state=
2009			[KVM,Intel] Enable emulation of invalid guest states
2010			Default is 0 (disabled)
2011
2012	kvm-intel.flexpriority=
2013			[KVM,Intel] Disable FlexPriority feature (TPR shadow).
2014			Default is 1 (enabled)
2015
2016	kvm-intel.nested=
2017			[KVM,Intel] Enable VMX nesting (nVMX).
2018			Default is 0 (disabled)
2019
2020	kvm-intel.unrestricted_guest=
2021			[KVM,Intel] Disable unrestricted guest feature
2022			(virtualized real and unpaged mode) on capable
2023			Intel chips. Default is 1 (enabled)
2024
2025	kvm-intel.vpid=	[KVM,Intel] Disable Virtual Processor Identification
2026			feature (tagged TLBs) on capable Intel chips.
2027			Default is 1 (enabled)
2028
2029	l2cr=		[PPC]
2030
2031	l3cr=		[PPC]
2032
2033	lapic		[X86-32,APIC] Enable the local APIC even if BIOS
2034			disabled it.
2035
2036	lapic=		[x86,APIC] "notscdeadline" Do not use TSC deadline
2037			value for LAPIC timer one-shot implementation. Default
2038			back to the programmable timer unit in the LAPIC.
2039
2040	lapic_timer_c2_ok	[X86,APIC] trust the local apic timer
2041			in C2 power state.
2042
2043	libata.dma=	[LIBATA] DMA control
2044			libata.dma=0	  Disable all PATA and SATA DMA
2045			libata.dma=1	  PATA and SATA Disk DMA only
2046			libata.dma=2	  ATAPI (CDROM) DMA only
2047			libata.dma=4	  Compact Flash DMA only
2048			Combinations also work, so libata.dma=3 enables DMA
2049			for disks and CDROMs, but not CFs.
2050
2051	libata.ignore_hpa=	[LIBATA] Ignore HPA limit
2052			libata.ignore_hpa=0	  keep BIOS limits (default)
2053			libata.ignore_hpa=1	  ignore limits, using full disk
2054
2055	libata.noacpi	[LIBATA] Disables use of ACPI in libata suspend/resume
2056			when set.
2057			Format: <int>
2058
2059	libata.force=	[LIBATA] Force configurations.  The format is comma
2060			separated list of "[ID:]VAL" where ID is
2061			PORT[.DEVICE].  PORT and DEVICE are decimal numbers
2062			matching port, link or device.  Basically, it matches
2063			the ATA ID string printed on console by libata.  If
2064			the whole ID part is omitted, the last PORT and DEVICE
2065			values are used.  If ID hasn't been specified yet, the
2066			configuration applies to all ports, links and devices.
2067
2068			If only DEVICE is omitted, the parameter applies to
2069			the port and all links and devices behind it.  DEVICE
2070			number of 0 either selects the first device or the
2071			first fan-out link behind PMP device.  It does not
2072			select the host link.  DEVICE number of 15 selects the
2073			host link and device attached to it.
2074
2075			The VAL specifies the configuration to force.  As long
2076			as there's no ambiguity shortcut notation is allowed.
2077			For example, both 1.5 and 1.5G would work for 1.5Gbps.
2078			The following configurations can be forced.
2079
2080			* Cable type: 40c, 80c, short40c, unk, ign or sata.
2081			  Any ID with matching PORT is used.
2082
2083			* SATA link speed limit: 1.5Gbps or 3.0Gbps.
2084
2085			* Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2086			  udma[/][16,25,33,44,66,100,133] notation is also
2087			  allowed.
2088
2089			* [no]ncq: Turn on or off NCQ.
2090
2091			* [no]ncqtrim: Turn off queued DSM TRIM.
2092
2093			* nohrst, nosrst, norst: suppress hard, soft
2094                          and both resets.
2095
2096			* rstonce: only attempt one reset during
2097			  hot-unplug link recovery
2098
2099			* dump_id: dump IDENTIFY data.
2100
2101			* atapi_dmadir: Enable ATAPI DMADIR bridge support
2102
2103			* disable: Disable this device.
2104
2105			If there are multiple matching configurations changing
2106			the same attribute, the last one is used.
2107
2108	memblock=debug	[KNL] Enable memblock debug messages.
2109
2110	load_ramdisk=	[RAM] List of ramdisks to load from floppy
2111			See Documentation/blockdev/ramdisk.txt.
2112
2113	lockd.nlm_grace_period=P  [NFS] Assign grace period.
2114			Format: <integer>
2115
2116	lockd.nlm_tcpport=N	[NFS] Assign TCP port.
2117			Format: <integer>
2118
2119	lockd.nlm_timeout=T	[NFS] Assign timeout value.
2120			Format: <integer>
2121
2122	lockd.nlm_udpport=M	[NFS] Assign UDP port.
2123			Format: <integer>
2124
2125	locktorture.nreaders_stress= [KNL]
2126			Set the number of locking read-acquisition kthreads.
2127			Defaults to being automatically set based on the
2128			number of online CPUs.
2129
2130	locktorture.nwriters_stress= [KNL]
2131			Set the number of locking write-acquisition kthreads.
2132
2133	locktorture.onoff_holdoff= [KNL]
2134			Set time (s) after boot for CPU-hotplug testing.
2135
2136	locktorture.onoff_interval= [KNL]
2137			Set time (s) between CPU-hotplug operations, or
2138			zero to disable CPU-hotplug testing.
2139
2140	locktorture.shuffle_interval= [KNL]
2141			Set task-shuffle interval (jiffies).  Shuffling
2142			tasks allows some CPUs to go into dyntick-idle
2143			mode during the locktorture test.
2144
2145	locktorture.shutdown_secs= [KNL]
2146			Set time (s) after boot system shutdown.  This
2147			is useful for hands-off automated testing.
2148
2149	locktorture.stat_interval= [KNL]
2150			Time (s) between statistics printk()s.
2151
2152	locktorture.stutter= [KNL]
2153			Time (s) to stutter testing, for example,
2154			specifying five seconds causes the test to run for
2155			five seconds, wait for five seconds, and so on.
2156			This tests the locking primitive's ability to
2157			transition abruptly to and from idle.
2158
2159	locktorture.torture_runnable= [BOOT]
2160			Start locktorture running at boot time.
2161
2162	locktorture.torture_type= [KNL]
2163			Specify the locking implementation to test.
2164
2165	locktorture.verbose= [KNL]
2166			Enable additional printk() statements.
2167
2168	logibm.irq=	[HW,MOUSE] Logitech Bus Mouse Driver
2169			Format: <irq>
2170
2171	loglevel=	All Kernel Messages with a loglevel smaller than the
2172			console loglevel will be printed to the console. It can
2173			also be changed with klogd or other programs. The
2174			loglevels are defined as follows:
2175
2176			0 (KERN_EMERG)		system is unusable
2177			1 (KERN_ALERT)		action must be taken immediately
2178			2 (KERN_CRIT)		critical conditions
2179			3 (KERN_ERR)		error conditions
2180			4 (KERN_WARNING)	warning conditions
2181			5 (KERN_NOTICE)		normal but significant condition
2182			6 (KERN_INFO)		informational
2183			7 (KERN_DEBUG)		debug-level messages
2184
2185	log_buf_len=n[KMG]	Sets the size of the printk ring buffer,
2186			in bytes.  n must be a power of two and greater
2187			than the minimal size. The minimal size is defined
2188			by LOG_BUF_SHIFT kernel config parameter. There is
2189			also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2190			that allows to increase the default size depending on
2191			the number of CPUs. See init/Kconfig for more details.
2192
2193	logo.nologo	[FB] Disables display of the built-in Linux logo.
2194			This may be used to provide more screen space for
2195			kernel log messages and is useful when debugging
2196			kernel boot problems.
2197
2198	lp=0		[LP]	Specify parallel ports to use, e.g,
2199	lp=port[,port...]	lp=none,parport0 (lp0 not configured, lp1 uses
2200	lp=reset		first parallel port). 'lp=0' disables the
2201	lp=auto			printer driver. 'lp=reset' (which can be
2202				specified in addition to the ports) causes
2203				attached printers to be reset. Using
2204				lp=port1,port2,... specifies the parallel ports
2205				to associate lp devices with, starting with
2206				lp0. A port specification may be 'none' to skip
2207				that lp device, or a parport name such as
2208				'parport0'. Specifying 'lp=auto' instead of a
2209				port specification list means that device IDs
2210				from each port should be examined, to see if
2211				an IEEE 1284-compliant printer is attached; if
2212				so, the driver will manage that printer.
2213				See also header of drivers/char/lp.c.
2214
2215	lpj=n		[KNL]
2216			Sets loops_per_jiffy to given constant, thus avoiding
2217			time-consuming boot-time autodetection (up to 250 ms per
2218			CPU). 0 enables autodetection (default). To determine
2219			the correct value for your kernel, boot with normal
2220			autodetection and see what value is printed. Note that
2221			on SMP systems the preset will be applied to all CPUs,
2222			which is likely to cause problems if your CPUs need
2223			significantly divergent settings. An incorrect value
2224			will cause delays in the kernel to be wrong, leading to
2225			unpredictable I/O errors and other breakage. Although
2226			unlikely, in the extreme case this might damage your
2227			hardware.
2228
2229	ltpc=		[NET]
2230			Format: <io>,<irq>,<dma>
2231
2232	machvec=	[IA-64] Force the use of a particular machine-vector
2233			(machvec) in a generic kernel.
2234			Example: machvec=hpzx1_swiotlb
2235
2236	machtype=	[Loongson] Share the same kernel image file between different
2237			 yeeloong laptop.
2238			Example: machtype=lemote-yeeloong-2f-7inch
2239
2240	max_addr=nn[KMG]	[KNL,BOOT,ia64] All physical memory greater
2241			than or equal to this physical address is ignored.
2242
2243	maxcpus=	[SMP] Maximum number of processors that	an SMP kernel
2244			will bring up during bootup.  maxcpus=n : n >= 0 limits
2245			the kernel to bring up 'n' processors. Surely after
2246			bootup you can bring up the other plugged cpu by executing
2247			"echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2248			only takes effect during system bootup.
2249			While n=0 is a special case, it is equivalent to "nosmp",
2250			which also disables the IO APIC.
2251
2252	max_loop=	[LOOP] The number of loop block devices that get
2253	(loop.max_loop)	unconditionally pre-created at init time. The default
2254			number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2255			of statically allocating a predefined number, loop
2256			devices can be requested on-demand with the
2257			/dev/loop-control interface.
2258
2259	mce		[X86-32] Machine Check Exception
2260
2261	mce=option	[X86-64] See Documentation/x86/x86_64/boot-options.txt
2262
2263	md=		[HW] RAID subsystems devices and level
2264			See Documentation/md.txt.
2265
2266	mdacon=		[MDA]
2267			Format: <first>,<last>
2268			Specifies range of consoles to be captured by the MDA.
2269
2270	mem=nn[KMG]	[KNL,BOOT] Force usage of a specific amount of memory
2271			Amount of memory to be used when the kernel is not able
2272			to see the whole system memory or for test.
2273			[X86] Work as limiting max address. Use together
2274			with memmap= to avoid physical address space collisions.
2275			Without memmap= PCI devices could be placed at addresses
2276			belonging to unused RAM.
2277
2278	mem=nopentium	[BUGS=X86-32] Disable usage of 4MB pages for kernel
2279			memory.
2280
2281	memchunk=nn[KMG]
2282			[KNL,SH] Allow user to override the default size for
2283			per-device physically contiguous DMA buffers.
2284
2285        memhp_default_state=online/offline
2286			[KNL] Set the initial state for the memory hotplug
2287			onlining policy. If not specified, the default value is
2288			set according to the
2289			CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2290			option.
2291			See Documentation/memory-hotplug.txt.
2292
2293	memmap=exactmap	[KNL,X86] Enable setting of an exact
2294			E820 memory map, as specified by the user.
2295			Such memmap=exactmap lines can be constructed based on
2296			BIOS output or other requirements. See the memmap=nn@ss
2297			option description.
2298
2299	memmap=nn[KMG]@ss[KMG]
2300			[KNL] Force usage of a specific region of memory.
2301			Region of memory to be used is from ss to ss+nn.
2302
2303	memmap=nn[KMG]#ss[KMG]
2304			[KNL,ACPI] Mark specific memory as ACPI data.
2305			Region of memory to be marked is from ss to ss+nn.
2306
2307	memmap=nn[KMG]$ss[KMG]
2308			[KNL,ACPI] Mark specific memory as reserved.
2309			Region of memory to be reserved is from ss to ss+nn.
2310			Example: Exclude memory from 0x18690000-0x1869ffff
2311			         memmap=64K$0x18690000
2312			         or
2313			         memmap=0x10000$0x18690000
2314
2315	memmap=nn[KMG]!ss[KMG]
2316			[KNL,X86] Mark specific memory as protected.
2317			Region of memory to be used, from ss to ss+nn.
2318			The memory region may be marked as e820 type 12 (0xc)
2319			and is NVDIMM or ADR memory.
2320
2321	memory_corruption_check=0/1 [X86]
2322			Some BIOSes seem to corrupt the first 64k of
2323			memory when doing things like suspend/resume.
2324			Setting this option will scan the memory
2325			looking for corruption.  Enabling this will
2326			both detect corruption and prevent the kernel
2327			from using the memory being corrupted.
2328			However, its intended as a diagnostic tool; if
2329			repeatable BIOS-originated corruption always
2330			affects the same memory, you can use memmap=
2331			to prevent the kernel from using that memory.
2332
2333	memory_corruption_check_size=size [X86]
2334			By default it checks for corruption in the low
2335			64k, making this memory unavailable for normal
2336			use.  Use this parameter to scan for
2337			corruption in more or less memory.
2338
2339	memory_corruption_check_period=seconds [X86]
2340			By default it checks for corruption every 60
2341			seconds.  Use this parameter to check at some
2342			other rate.  0 disables periodic checking.
2343
2344	memtest=	[KNL,X86,ARM] Enable memtest
2345			Format: <integer>
2346			default : 0 <disable>
2347			Specifies the number of memtest passes to be
2348			performed. Each pass selects another test
2349			pattern from a given set of patterns. Memtest
2350			fills the memory with this pattern, validates
2351			memory contents and reserves bad memory
2352			regions that are detected.
2353
2354	meye.*=		[HW] Set MotionEye Camera parameters
2355			See Documentation/video4linux/meye.txt.
2356
2357	mfgpt_irq=	[IA-32] Specify the IRQ to use for the
2358			Multi-Function General Purpose Timers on AMD Geode
2359			platforms.
2360
2361	mfgptfix	[X86-32] Fix MFGPT timers on AMD Geode platforms when
2362			the BIOS has incorrectly applied a workaround. TinyBIOS
2363			version 0.98 is known to be affected, 0.99 fixes the
2364			problem by letting the user disable the workaround.
2365
2366	mga=		[HW,DRM]
2367
2368	min_addr=nn[KMG]	[KNL,BOOT,ia64] All physical memory below this
2369			physical address is ignored.
2370
2371	mini2440=	[ARM,HW,KNL]
2372			Format:[0..2][b][c][t]
2373			Default: "0tb"
2374			MINI2440 configuration specification:
2375			0 - The attached screen is the 3.5" TFT
2376			1 - The attached screen is the 7" TFT
2377			2 - The VGA Shield is attached (1024x768)
2378			Leaving out the screen size parameter will not load
2379			the TFT driver, and the framebuffer will be left
2380			unconfigured.
2381			b - Enable backlight. The TFT backlight pin will be
2382			linked to the kernel VESA blanking code and a GPIO
2383			LED. This parameter is not necessary when using the
2384			VGA shield.
2385			c - Enable the s3c camera interface.
2386			t - Reserved for enabling touchscreen support. The
2387			touchscreen support is not enabled in the mainstream
2388			kernel as of 2.6.30, a preliminary port can be found
2389			in the "bleeding edge" mini2440 support kernel at
2390			http://repo.or.cz/w/linux-2.6/mini2440.git
2391
2392	mminit_loglevel=
2393			[KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2394			parameter allows control of the logging verbosity for
2395			the additional memory initialisation checks. A value
2396			of 0 disables mminit logging and a level of 4 will
2397			log everything. Information is printed at KERN_DEBUG
2398			so loglevel=8 may also need to be specified.
2399
2400	module.sig_enforce
2401			[KNL] When CONFIG_MODULE_SIG is set, this means that
2402			modules without (valid) signatures will fail to load.
2403			Note that if CONFIG_MODULE_SIG_FORCE is set, that
2404			is always true, so this option does nothing.
2405
2406	module_blacklist=  [KNL] Do not load a comma-separated list of
2407			modules.  Useful for debugging problem modules.
2408
2409	mousedev.tap_time=
2410			[MOUSE] Maximum time between finger touching and
2411			leaving touchpad surface for touch to be considered
2412			a tap and be reported as a left button click (for
2413			touchpads working in absolute mode only).
2414			Format: <msecs>
2415	mousedev.xres=	[MOUSE] Horizontal screen resolution, used for devices
2416			reporting absolute coordinates, such as tablets
2417	mousedev.yres=	[MOUSE] Vertical screen resolution, used for devices
2418			reporting absolute coordinates, such as tablets
2419
2420	movablecore=nn[KMG]	[KNL,X86,IA-64,PPC] This parameter
2421			is similar to kernelcore except it specifies the
2422			amount of memory used for migratable allocations.
2423			If both kernelcore and movablecore is specified,
2424			then kernelcore will be at *least* the specified
2425			value but may be more. If movablecore on its own
2426			is specified, the administrator must be careful
2427			that the amount of memory usable for all allocations
2428			is not too small.
2429
2430	movable_node	[KNL,X86] Boot-time switch to enable the effects
2431			of CONFIG_MOVABLE_NODE=y. See mm/Kconfig for details.
2432
2433	MTD_Partition=	[MTD]
2434			Format: <name>,<region-number>,<size>,<offset>
2435
2436	MTD_Region=	[MTD] Format:
2437			<name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2438
2439	mtdparts=	[MTD]
2440			See drivers/mtd/cmdlinepart.c.
2441
2442	multitce=off	[PPC]  This parameter disables the use of the pSeries
2443			firmware feature for updating multiple TCE entries
2444			at a time.
2445
2446	onenand.bdry=	[HW,MTD] Flex-OneNAND Boundary Configuration
2447
2448			Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2449
2450			boundary - index of last SLC block on Flex-OneNAND.
2451				   The remaining blocks are configured as MLC blocks.
2452			lock	 - Configure if Flex-OneNAND boundary should be locked.
2453				   Once locked, the boundary cannot be changed.
2454				   1 indicates lock status, 0 indicates unlock status.
2455
2456	mtdset=		[ARM]
2457			ARM/S3C2412 JIVE boot control
2458
2459			See arch/arm/mach-s3c2412/mach-jive.c
2460
2461	mtouchusb.raw_coordinates=
2462			[HW] Make the MicroTouch USB driver use raw coordinates
2463			('y', default) or cooked coordinates ('n')
2464
2465	mtrr_chunk_size=nn[KMG] [X86]
2466			used for mtrr cleanup. It is largest continuous chunk
2467			that could hold holes aka. UC entries.
2468
2469	mtrr_gran_size=nn[KMG] [X86]
2470			Used for mtrr cleanup. It is granularity of mtrr block.
2471			Default is 1.
2472			Large value could prevent small alignment from
2473			using up MTRRs.
2474
2475	mtrr_spare_reg_nr=n [X86]
2476			Format: <integer>
2477			Range: 0,7 : spare reg number
2478			Default : 1
2479			Used for mtrr cleanup. It is spare mtrr entries number.
2480			Set to 2 or more if your graphical card needs more.
2481
2482	n2=		[NET] SDL Inc. RISCom/N2 synchronous serial card
2483
2484	netdev=		[NET] Network devices parameters
2485			Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2486			Note that mem_start is often overloaded to mean
2487			something different and driver-specific.
2488			This usage is only documented in each driver source
2489			file if at all.
2490
2491	nf_conntrack.acct=
2492			[NETFILTER] Enable connection tracking flow accounting
2493			0 to disable accounting
2494			1 to enable accounting
2495			Default value is 0.
2496
2497	nfsaddrs=	[NFS] Deprecated.  Use ip= instead.
2498			See Documentation/filesystems/nfs/nfsroot.txt.
2499
2500	nfsroot=	[NFS] nfs root filesystem for disk-less boxes.
2501			See Documentation/filesystems/nfs/nfsroot.txt.
2502
2503	nfsrootdebug	[NFS] enable nfsroot debugging messages.
2504			See Documentation/filesystems/nfs/nfsroot.txt.
2505
2506	nfs.callback_nr_threads=
2507			[NFSv4] set the total number of threads that the
2508			NFS client will assign to service NFSv4 callback
2509			requests.
2510
2511	nfs.callback_tcpport=
2512			[NFS] set the TCP port on which the NFSv4 callback
2513			channel should listen.
2514
2515	nfs.cache_getent=
2516			[NFS] sets the pathname to the program which is used
2517			to update the NFS client cache entries.
2518
2519	nfs.cache_getent_timeout=
2520			[NFS] sets the timeout after which an attempt to
2521			update a cache entry is deemed to have failed.
2522
2523	nfs.idmap_cache_timeout=
2524			[NFS] set the maximum lifetime for idmapper cache
2525			entries.
2526
2527	nfs.enable_ino64=
2528			[NFS] enable 64-bit inode numbers.
2529			If zero, the NFS client will fake up a 32-bit inode
2530			number for the readdir() and stat() syscalls instead
2531			of returning the full 64-bit number.
2532			The default is to return 64-bit inode numbers.
2533
2534	nfs.max_session_cb_slots=
2535			[NFSv4.1] Sets the maximum number of session
2536			slots the client will assign to the callback
2537			channel. This determines the maximum number of
2538			callbacks the client will process in parallel for
2539			a particular server.
2540
2541	nfs.max_session_slots=
2542			[NFSv4.1] Sets the maximum number of session slots
2543			the client will attempt to negotiate with the server.
2544			This limits the number of simultaneous RPC requests
2545			that the client can send to the NFSv4.1 server.
2546			Note that there is little point in setting this
2547			value higher than the max_tcp_slot_table_limit.
2548
2549	nfs.nfs4_disable_idmapping=
2550			[NFSv4] When set to the default of '1', this option
2551			ensures that both the RPC level authentication
2552			scheme and the NFS level operations agree to use
2553			numeric uids/gids if the mount is using the
2554			'sec=sys' security flavour. In effect it is
2555			disabling idmapping, which can make migration from
2556			legacy NFSv2/v3 systems to NFSv4 easier.
2557			Servers that do not support this mode of operation
2558			will be autodetected by the client, and it will fall
2559			back to using the idmapper.
2560			To turn off this behaviour, set the value to '0'.
2561	nfs.nfs4_unique_id=
2562			[NFS4] Specify an additional fixed unique ident-
2563			ification string that NFSv4 clients can insert into
2564			their nfs_client_id4 string.  This is typically a
2565			UUID that is generated at system install time.
2566
2567	nfs.send_implementation_id =
2568			[NFSv4.1] Send client implementation identification
2569			information in exchange_id requests.
2570			If zero, no implementation identification information
2571			will be sent.
2572			The default is to send the implementation identification
2573			information.
2574
2575	nfs.recover_lost_locks =
2576			[NFSv4] Attempt to recover locks that were lost due
2577			to a lease timeout on the server. Please note that
2578			doing this risks data corruption, since there are
2579			no guarantees that the file will remain unchanged
2580			after the locks are lost.
2581			If you want to enable the kernel legacy behaviour of
2582			attempting to recover these locks, then set this
2583			parameter to '1'.
2584			The default parameter value of '0' causes the kernel
2585			not to attempt recovery of lost locks.
2586
2587	nfs4.layoutstats_timer =
2588			[NFSv4.2] Change the rate at which the kernel sends
2589			layoutstats to the pNFS metadata server.
2590
2591			Setting this to value to 0 causes the kernel to use
2592			whatever value is the default set by the layout
2593			driver. A non-zero value sets the minimum interval
2594			in seconds between layoutstats transmissions.
2595
2596	nfsd.nfs4_disable_idmapping=
2597			[NFSv4] When set to the default of '1', the NFSv4
2598			server will return only numeric uids and gids to
2599			clients using auth_sys, and will accept numeric uids
2600			and gids from such clients.  This is intended to ease
2601			migration from NFSv2/v3.
2602
2603	objlayoutdriver.osd_login_prog=
2604			[NFS] [OBJLAYOUT] sets the pathname to the program which
2605			is used to automatically discover and login into new
2606			osd-targets. Please see:
2607			Documentation/filesystems/pnfs.txt for more explanations
2608
2609	nmi_debug=	[KNL,AVR32,SH] Specify one or more actions to take
2610			when a NMI is triggered.
2611			Format: [state][,regs][,debounce][,die]
2612
2613	nmi_watchdog=	[KNL,BUGS=X86] Debugging features for SMP kernels
2614			Format: [panic,][nopanic,][num]
2615			Valid num: 0 or 1
2616			0 - turn hardlockup detector in nmi_watchdog off
2617			1 - turn hardlockup detector in nmi_watchdog on
2618			When panic is specified, panic when an NMI watchdog
2619			timeout occurs (or 'nopanic' to override the opposite
2620			default). To disable both hard and soft lockup detectors,
2621			please see 'nowatchdog'.
2622			This is useful when you use a panic=... timeout and
2623			need the box quickly up again.
2624
2625	netpoll.carrier_timeout=
2626			[NET] Specifies amount of time (in seconds) that
2627			netpoll should wait for a carrier. By default netpoll
2628			waits 4 seconds.
2629
2630	no387		[BUGS=X86-32] Tells the kernel to use the 387 maths
2631			emulation library even if a 387 maths coprocessor
2632			is present.
2633
2634	no_console_suspend
2635			[HW] Never suspend the console
2636			Disable suspending of consoles during suspend and
2637			hibernate operations.  Once disabled, debugging
2638			messages can reach various consoles while the rest
2639			of the system is being put to sleep (ie, while
2640			debugging driver suspend/resume hooks).  This may
2641			not work reliably with all consoles, but is known
2642			to work with serial and VGA consoles.
2643			To facilitate more flexible debugging, we also add
2644			console_suspend, a printk module parameter to control
2645			it. Users could use console_suspend (usually
2646			/sys/module/printk/parameters/console_suspend) to
2647			turn on/off it dynamically.
2648
2649	noaliencache	[MM, NUMA, SLAB] Disables the allocation of alien
2650			caches in the slab allocator.  Saves per-node memory,
2651			but will impact performance.
2652
2653	noalign		[KNL,ARM]
2654
2655	noaltinstr	[S390] Disables alternative instructions patching
2656			(CPU alternatives feature).
2657
2658	noapic		[SMP,APIC] Tells the kernel to not make use of any
2659			IOAPICs that may be present in the system.
2660
2661	noautogroup	Disable scheduler automatic task group creation.
2662
2663	nobats		[PPC] Do not use BATs for mapping kernel lowmem
2664			on "Classic" PPC cores.
2665
2666	nocache		[ARM]
2667
2668	noclflush	[BUGS=X86] Don't use the CLFLUSH instruction
2669
2670	nodelayacct	[KNL] Disable per-task delay accounting
2671
2672	nodsp		[SH] Disable hardware DSP at boot time.
2673
2674	noefi		Disable EFI runtime services support.
2675
2676	noexec		[IA-64]
2677
2678	noexec		[X86]
2679			On X86-32 available only on PAE configured kernels.
2680			noexec=on: enable non-executable mappings (default)
2681			noexec=off: disable non-executable mappings
2682
2683	nosmap		[X86]
2684			Disable SMAP (Supervisor Mode Access Prevention)
2685			even if it is supported by processor.
2686
2687	nosmep		[X86]
2688			Disable SMEP (Supervisor Mode Execution Prevention)
2689			even if it is supported by processor.
2690
2691	noexec32	[X86-64]
2692			This affects only 32-bit executables.
2693			noexec32=on: enable non-executable mappings (default)
2694				read doesn't imply executable mappings
2695			noexec32=off: disable non-executable mappings
2696				read implies executable mappings
2697
2698	nofpu		[MIPS,SH] Disable hardware FPU at boot time.
2699
2700	nofxsr		[BUGS=X86-32] Disables x86 floating point extended
2701			register save and restore. The kernel will only save
2702			legacy floating-point registers on task switch.
2703
2704	nohugeiomap	[KNL,x86] Disable kernel huge I/O mappings.
2705
2706	nosmt		[KNL,S390] Disable symmetric multithreading (SMT).
2707			Equivalent to smt=1.
2708
2709	nospectre_v2	[X86] Disable all mitigations for the Spectre variant 2
2710			(indirect branch prediction) vulnerability. System may
2711			allow data leaks with this option, which is equivalent
2712			to spectre_v2=off.
2713
2714	noxsave		[BUGS=X86] Disables x86 extended register state save
2715			and restore using xsave. The kernel will fallback to
2716			enabling legacy floating-point and sse state.
2717
2718	noxsaveopt	[X86] Disables xsaveopt used in saving x86 extended
2719			register states. The kernel will fall back to use
2720			xsave to save the states. By using this parameter,
2721			performance of saving the states is degraded because
2722			xsave doesn't support modified optimization while
2723			xsaveopt supports it on xsaveopt enabled systems.
2724
2725	noxsaves	[X86] Disables xsaves and xrstors used in saving and
2726			restoring x86 extended register state in compacted
2727			form of xsave area. The kernel will fall back to use
2728			xsaveopt and xrstor to save and restore the states
2729			in standard form of xsave area. By using this
2730			parameter, xsave area per process might occupy more
2731			memory on xsaves enabled systems.
2732
2733	nohlt		[BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2734			wfi(ARM) instruction doesn't work correctly and not to
2735			use it. This is also useful when using JTAG debugger.
2736
2737	no_file_caps	Tells the kernel not to honor file capabilities.  The
2738			only way then for a file to be executed with privilege
2739			is to be setuid root or executed by root.
2740
2741	nohalt		[IA-64] Tells the kernel not to use the power saving
2742			function PAL_HALT_LIGHT when idle. This increases
2743			power-consumption. On the positive side, it reduces
2744			interrupt wake-up latency, which may improve performance
2745			in certain environments such as networked servers or
2746			real-time systems.
2747
2748	nohibernate	[HIBERNATION] Disable hibernation and resume.
2749
2750	nohz=		[KNL] Boottime enable/disable dynamic ticks
2751			Valid arguments: on, off
2752			Default: on
2753
2754	nohz_full=	[KNL,BOOT]
2755			The argument is a cpu list, as described above.
2756			In kernels built with CONFIG_NO_HZ_FULL=y, set
2757			the specified list of CPUs whose tick will be stopped
2758			whenever possible. The boot CPU will be forced outside
2759			the range to maintain the timekeeping.
2760			The CPUs in this range must also be included in the
2761			rcu_nocbs= set.
2762
2763	noiotrap	[SH] Disables trapped I/O port accesses.
2764
2765	noirqdebug	[X86-32] Disables the code which attempts to detect and
2766			disable unhandled interrupt sources.
2767
2768	no_timer_check	[X86,APIC] Disables the code which tests for
2769			broken timer IRQ sources.
2770
2771	noisapnp	[ISAPNP] Disables ISA PnP code.
2772
2773	noinitrd	[RAM] Tells the kernel not to load any configured
2774			initial RAM disk.
2775
2776	nointremap	[X86-64, Intel-IOMMU] Do not enable interrupt
2777			remapping.
2778			[Deprecated - use intremap=off]
2779
2780	nointroute	[IA-64]
2781
2782	noinvpcid	[X86] Disable the INVPCID cpu feature.
2783
2784	nojitter	[IA-64] Disables jitter checking for ITC timers.
2785
2786	no-kvmclock	[X86,KVM] Disable paravirtualized KVM clock driver
2787
2788	no-kvmapf	[X86,KVM] Disable paravirtualized asynchronous page
2789			fault handling.
2790
2791	no-steal-acc    [X86,KVM] Disable paravirtualized steal time accounting.
2792			steal time is computed, but won't influence scheduler
2793			behaviour
2794
2795	nolapic		[X86-32,APIC] Do not enable or use the local APIC.
2796
2797	nolapic_timer	[X86-32,APIC] Do not use the local APIC timer.
2798
2799	noltlbs		[PPC] Do not use large page/tlb entries for kernel
2800			lowmem mapping on PPC40x and PPC8xx
2801
2802	nomca		[IA-64] Disable machine check abort handling
2803
2804	nomce		[X86-32] Disable Machine Check Exception
2805
2806	nomfgpt		[X86-32] Disable Multi-Function General Purpose
2807			Timer usage (for AMD Geode machines).
2808
2809	nonmi_ipi	[X86] Disable using NMI IPIs during panic/reboot to
2810			shutdown the other cpus.  Instead use the REBOOT_VECTOR
2811			irq.
2812
2813	nomodule	Disable module load
2814
2815	nopat		[X86] Disable PAT (page attribute table extension of
2816			pagetables) support.
2817
2818	nopcid		[X86-64] Disable the PCID cpu feature.
2819
2820	norandmaps	Don't use address space randomization.  Equivalent to
2821			echo 0 > /proc/sys/kernel/randomize_va_space
2822
2823	noreplace-smp	[X86-32,SMP] Don't replace SMP instructions
2824			with UP alternatives
2825
2826	nordrand	[X86] Disable kernel use of the RDRAND and
2827			RDSEED instructions even if they are supported
2828			by the processor.  RDRAND and RDSEED are still
2829			available to user space applications.
2830
2831	noresume	[SWSUSP] Disables resume and restores original swap
2832			space.
2833
2834	no-scroll	[VGA] Disables scrollback.
2835			This is required for the Braillex ib80-piezo Braille
2836			reader made by F.H. Papenmeier (Germany).
2837
2838	nosbagart	[IA-64]
2839
2840	nosep		[BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2841
2842	nosmp		[SMP] Tells an SMP kernel to act as a UP kernel,
2843			and disable the IO APIC.  legacy for "maxcpus=0".
2844
2845	nosoftlockup	[KNL] Disable the soft-lockup detector.
2846
2847	nosync		[HW,M68K] Disables sync negotiation for all devices.
2848
2849	notsc		[BUGS=X86-32] Disable Time Stamp Counter
2850
2851	nowatchdog	[KNL] Disable both lockup detectors, i.e.
2852                        soft-lockup and NMI watchdog (hard-lockup).
2853
2854	nowb		[ARM]
2855
2856	nox2apic	[X86-64,APIC] Do not enable x2APIC mode.
2857
2858	cpu0_hotplug	[X86] Turn on CPU0 hotplug feature when
2859			CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2860			Some features depend on CPU0. Known dependencies are:
2861			1. Resume from suspend/hibernate depends on CPU0.
2862			Suspend/hibernate will fail if CPU0 is offline and you
2863			need to online CPU0 before suspend/hibernate.
2864			2. PIC interrupts also depend on CPU0. CPU0 can't be
2865			removed if a PIC interrupt is detected.
2866			It's said poweroff/reboot may depend on CPU0 on some
2867			machines although I haven't seen such issues so far
2868			after CPU0 is offline on a few tested machines.
2869			If the dependencies are under your control, you can
2870			turn on cpu0_hotplug.
2871
2872	nptcg=		[IA-64] Override max number of concurrent global TLB
2873			purges which is reported from either PAL_VM_SUMMARY or
2874			SAL PALO.
2875
2876	nr_cpus=	[SMP] Maximum number of processors that	an SMP kernel
2877			could support.  nr_cpus=n : n >= 1 limits the kernel to
2878			support 'n' processors. It could be larger than the
2879			number of already plugged CPU during bootup, later in
2880			runtime you can physically add extra cpu until it reaches
2881			n. So during boot up some boot time memory for per-cpu
2882			variables need be pre-allocated for later physical cpu
2883			hot plugging.
2884
2885	nr_uarts=	[SERIAL] maximum number of UARTs to be registered.
2886
2887	numa_balancing=	[KNL,X86] Enable or disable automatic NUMA balancing.
2888			Allowed values are enable and disable
2889
2890	numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2891			one of ['zone', 'node', 'default'] can be specified
2892			This can be set from sysctl after boot.
2893			See Documentation/sysctl/vm.txt for details.
2894
2895	ohci1394_dma=early	[HW] enable debugging via the ohci1394 driver.
2896			See Documentation/debugging-via-ohci1394.txt for more
2897			info.
2898
2899	olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2900			Rather than timing out after 20 ms if an EC
2901			command is not properly ACKed, override the length
2902			of the timeout.  We have interrupts disabled while
2903			waiting for the ACK, so if this is set too high
2904			interrupts *may* be lost!
2905
2906	omap_mux=	[OMAP] Override bootloader pin multiplexing.
2907			Format: <mux_mode0.mode_name=value>...
2908			For example, to override I2C bus2:
2909			omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2910
2911	oprofile.timer=	[HW]
2912			Use timer interrupt instead of performance counters
2913
2914	oprofile.cpu_type=	Force an oprofile cpu type
2915			This might be useful if you have an older oprofile
2916			userland or if you want common events.
2917			Format: { arch_perfmon }
2918			arch_perfmon: [X86] Force use of architectural
2919				perfmon on Intel CPUs instead of the
2920				CPU specific event set.
2921			timer: [X86] Force use of architectural NMI
2922				timer mode (see also oprofile.timer
2923				for generic hr timer mode)
2924
2925	oops=panic	Always panic on oopses. Default is to just kill the
2926			process, but there is a small probability of
2927			deadlocking the machine.
2928			This will also cause panics on machine check exceptions.
2929			Useful together with panic=30 to trigger a reboot.
2930
2931	OSS		[HW,OSS]
2932			See Documentation/sound/oss/oss-parameters.txt
2933
2934	page_owner=	[KNL] Boot-time page_owner enabling option.
2935			Storage of the information about who allocated
2936			each page is disabled in default. With this switch,
2937			we can turn it on.
2938			on: enable the feature
2939
2940	page_poison=	[KNL] Boot-time parameter changing the state of
2941			poisoning on the buddy allocator.
2942			off: turn off poisoning
2943			on: turn on poisoning
2944
2945	panic=		[KNL] Kernel behaviour on panic: delay <timeout>
2946			timeout > 0: seconds before rebooting
2947			timeout = 0: wait forever
2948			timeout < 0: reboot immediately
2949			Format: <timeout>
2950
2951	panic_on_warn	panic() instead of WARN().  Useful to cause kdump
2952			on a WARN().
2953
2954	crash_kexec_post_notifiers
2955			Run kdump after running panic-notifiers and dumping
2956			kmsg. This only for the users who doubt kdump always
2957			succeeds in any situation.
2958			Note that this also increases risks of kdump failure,
2959			because some panic notifiers can make the crashed
2960			kernel more unstable.
2961
2962	parkbd.port=	[HW] Parallel port number the keyboard adapter is
2963			connected to, default is 0.
2964			Format: <parport#>
2965	parkbd.mode=	[HW] Parallel port keyboard adapter mode of operation,
2966			0 for XT, 1 for AT (default is AT).
2967			Format: <mode>
2968
2969	parport=	[HW,PPT] Specify parallel ports. 0 disables.
2970			Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2971			Use 'auto' to force the driver to use any
2972			IRQ/DMA settings detected (the default is to
2973			ignore detected IRQ/DMA settings because of
2974			possible conflicts). You can specify the base
2975			address, IRQ, and DMA settings; IRQ and DMA
2976			should be numbers, or 'auto' (for using detected
2977			settings on that particular port), or 'nofifo'
2978			(to avoid using a FIFO even if it is detected).
2979			Parallel ports are assigned in the order they
2980			are specified on the command line, starting
2981			with parport0.
2982
2983	parport_init_mode=	[HW,PPT]
2984			Configure VIA parallel port to operate in
2985			a specific mode. This is necessary on Pegasos
2986			computer where firmware has no options for setting
2987			up parallel port mode and sets it to spp.
2988			Currently this function knows 686a and 8231 chips.
2989			Format: [spp|ps2|epp|ecp|ecpepp]
2990
2991	pause_on_oops=
2992			Halt all CPUs after the first oops has been printed for
2993			the specified number of seconds.  This is to be used if
2994			your oopses keep scrolling off the screen.
2995
2996	pcbit=		[HW,ISDN]
2997
2998	pcd.		[PARIDE]
2999			See header of drivers/block/paride/pcd.c.
3000			See also Documentation/blockdev/paride.txt.
3001
3002	pci=option[,option...]	[PCI] various PCI subsystem options:
3003		earlydump	[X86] dump PCI config space before the kernel
3004			        changes anything
3005		off		[X86] don't probe for the PCI bus
3006		bios		[X86-32] force use of PCI BIOS, don't access
3007				the hardware directly. Use this if your machine
3008				has a non-standard PCI host bridge.
3009		nobios		[X86-32] disallow use of PCI BIOS, only direct
3010				hardware access methods are allowed. Use this
3011				if you experience crashes upon bootup and you
3012				suspect they are caused by the BIOS.
3013		conf1		[X86] Force use of PCI Configuration Access
3014				Mechanism 1 (config address in IO port 0xCF8,
3015				data in IO port 0xCFC, both 32-bit).
3016		conf2		[X86] Force use of PCI Configuration Access
3017				Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3018				the function, IO port 0xCFA, also 8-bit, sets
3019				bus number. The config space is then accessed
3020				through ports 0xC000-0xCFFF).
3021				See http://wiki.osdev.org/PCI for more info
3022				on the configuration access mechanisms.
3023		noaer		[PCIE] If the PCIEAER kernel config parameter is
3024				enabled, this kernel boot option can be used to
3025				disable the use of PCIE advanced error reporting.
3026		nodomains	[PCI] Disable support for multiple PCI
3027				root domains (aka PCI segments, in ACPI-speak).
3028		nommconf	[X86] Disable use of MMCONFIG for PCI
3029				Configuration
3030		check_enable_amd_mmconf [X86] check for and enable
3031				properly configured MMIO access to PCI
3032				config space on AMD family 10h CPU
3033		nomsi		[MSI] If the PCI_MSI kernel config parameter is
3034				enabled, this kernel boot option can be used to
3035				disable the use of MSI interrupts system-wide.
3036		noioapicquirk	[APIC] Disable all boot interrupt quirks.
3037				Safety option to keep boot IRQs enabled. This
3038				should never be necessary.
3039		ioapicreroute	[APIC] Enable rerouting of boot IRQs to the
3040				primary IO-APIC for bridges that cannot disable
3041				boot IRQs. This fixes a source of spurious IRQs
3042				when the system masks IRQs.
3043		noioapicreroute	[APIC] Disable workaround that uses the
3044				boot IRQ equivalent of an IRQ that connects to
3045				a chipset where boot IRQs cannot be disabled.
3046				The opposite of ioapicreroute.
3047		biosirq		[X86-32] Use PCI BIOS calls to get the interrupt
3048				routing table. These calls are known to be buggy
3049				on several machines and they hang the machine
3050				when used, but on other computers it's the only
3051				way to get the interrupt routing table. Try
3052				this option if the kernel is unable to allocate
3053				IRQs or discover secondary PCI buses on your
3054				motherboard.
3055		rom		[X86] Assign address space to expansion ROMs.
3056				Use with caution as certain devices share
3057				address decoders between ROMs and other
3058				resources.
3059		norom		[X86] Do not assign address space to
3060				expansion ROMs that do not already have
3061				BIOS assigned address ranges.
3062		nobar		[X86] Do not assign address space to the
3063				BARs that weren't assigned by the BIOS.
3064		irqmask=0xMMMM	[X86] Set a bit mask of IRQs allowed to be
3065				assigned automatically to PCI devices. You can
3066				make the kernel exclude IRQs of your ISA cards
3067				this way.
3068		pirqaddr=0xAAAAA	[X86] Specify the physical address
3069				of the PIRQ table (normally generated
3070				by the BIOS) if it is outside the
3071				F0000h-100000h range.
3072		lastbus=N	[X86] Scan all buses thru bus #N. Can be
3073				useful if the kernel is unable to find your
3074				secondary buses and you want to tell it
3075				explicitly which ones they are.
3076		assign-busses	[X86] Always assign all PCI bus
3077				numbers ourselves, overriding
3078				whatever the firmware may have done.
3079		usepirqmask	[X86] Honor the possible IRQ mask stored
3080				in the BIOS $PIR table. This is needed on
3081				some systems with broken BIOSes, notably
3082				some HP Pavilion N5400 and Omnibook XE3
3083				notebooks. This will have no effect if ACPI
3084				IRQ routing is enabled.
3085		noacpi		[X86] Do not use ACPI for IRQ routing
3086				or for PCI scanning.
3087		use_crs		[X86] Use PCI host bridge window information
3088				from ACPI.  On BIOSes from 2008 or later, this
3089				is enabled by default.  If you need to use this,
3090				please report a bug.
3091		nocrs		[X86] Ignore PCI host bridge windows from ACPI.
3092			        If you need to use this, please report a bug.
3093		routeirq	Do IRQ routing for all PCI devices.
3094				This is normally done in pci_enable_device(),
3095				so this option is a temporary workaround
3096				for broken drivers that don't call it.
3097		skip_isa_align	[X86] do not align io start addr, so can
3098				handle more pci cards
3099		noearly		[X86] Don't do any early type 1 scanning.
3100				This might help on some broken boards which
3101				machine check when some devices' config space
3102				is read. But various workarounds are disabled
3103				and some IOMMU drivers will not work.
3104		bfsort		Sort PCI devices into breadth-first order.
3105				This sorting is done to get a device
3106				order compatible with older (<= 2.4) kernels.
3107		nobfsort	Don't sort PCI devices into breadth-first order.
3108		pcie_bus_tune_off	Disable PCIe MPS (Max Payload Size)
3109				tuning and use the BIOS-configured MPS defaults.
3110		pcie_bus_safe	Set every device's MPS to the largest value
3111				supported by all devices below the root complex.
3112		pcie_bus_perf	Set device MPS to the largest allowable MPS
3113				based on its parent bus. Also set MRRS (Max
3114				Read Request Size) to the largest supported
3115				value (no larger than the MPS that the device
3116				or bus can support) for best performance.
3117		pcie_bus_peer2peer	Set every device's MPS to 128B, which
3118				every device is guaranteed to support. This
3119				configuration allows peer-to-peer DMA between
3120				any pair of devices, possibly at the cost of
3121				reduced performance.  This also guarantees
3122				that hot-added devices will work.
3123		cbiosize=nn[KMG]	The fixed amount of bus space which is
3124				reserved for the CardBus bridge's IO window.
3125				The default value is 256 bytes.
3126		cbmemsize=nn[KMG]	The fixed amount of bus space which is
3127				reserved for the CardBus bridge's memory
3128				window. The default value is 64 megabytes.
3129		resource_alignment=
3130				Format:
3131				[<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
3132				[<order of align>@]pci:<vendor>:<device>\
3133						[:<subvendor>:<subdevice>][; ...]
3134				Specifies alignment and device to reassign
3135				aligned memory resources.
3136				If <order of align> is not specified,
3137				PAGE_SIZE is used as alignment.
3138				PCI-PCI bridge can be specified, if resource
3139				windows need to be expanded.
3140				To specify the alignment for several
3141				instances of a device, the PCI vendor,
3142				device, subvendor, and subdevice may be
3143				specified, e.g., 4096@pci:8086:9c22:103c:198f
3144		ecrc=		Enable/disable PCIe ECRC (transaction layer
3145				end-to-end CRC checking).
3146				bios: Use BIOS/firmware settings. This is the
3147				the default.
3148				off: Turn ECRC off
3149				on: Turn ECRC on.
3150		hpiosize=nn[KMG]	The fixed amount of bus space which is
3151				reserved for hotplug bridge's IO window.
3152				Default size is 256 bytes.
3153		hpmemsize=nn[KMG]	The fixed amount of bus space which is
3154				reserved for hotplug bridge's memory window.
3155				Default size is 2 megabytes.
3156		hpbussize=nn	The minimum amount of additional bus numbers
3157				reserved for buses below a hotplug bridge.
3158				Default is 1.
3159		realloc=	Enable/disable reallocating PCI bridge resources
3160				if allocations done by BIOS are too small to
3161				accommodate resources required by all child
3162				devices.
3163				off: Turn realloc off
3164				on: Turn realloc on
3165		realloc		same as realloc=on
3166		noari		do not use PCIe ARI.
3167		pcie_scan_all	Scan all possible PCIe devices.  Otherwise we
3168				only look for one device below a PCIe downstream
3169				port.
3170
3171	pcie_aspm=	[PCIE] Forcibly enable or disable PCIe Active State Power
3172			Management.
3173		off	Disable ASPM.
3174		force	Enable ASPM even on devices that claim not to support it.
3175			WARNING: Forcing ASPM on may cause system lockups.
3176
3177	pcie_hp=	[PCIE] PCI Express Hotplug driver options:
3178		nomsi	Do not use MSI for PCI Express Native Hotplug (this
3179			makes all PCIe ports use INTx for hotplug services).
3180
3181	pcie_ports=	[PCIE] PCIe ports handling:
3182		auto	Ask the BIOS whether or not to use native PCIe services
3183			associated with PCIe ports (PME, hot-plug, AER).  Use
3184			them only if that is allowed by the BIOS.
3185		native	Use native PCIe services associated with PCIe ports
3186			unconditionally.
3187		compat	Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
3188			ports driver.
3189
3190	pcie_port_pm=	[PCIE] PCIe port power management handling:
3191		off	Disable power management of all PCIe ports
3192		force	Forcibly enable power management of all PCIe ports
3193
3194	pcie_pme=	[PCIE,PM] Native PCIe PME signaling options:
3195		nomsi	Do not use MSI for native PCIe PME signaling (this makes
3196			all PCIe root ports use INTx for all services).
3197
3198	pcmv=		[HW,PCMCIA] BadgePAD 4
3199
3200	pd_ignore_unused
3201			[PM]
3202			Keep all power-domains already enabled by bootloader on,
3203			even if no driver has claimed them. This is useful
3204			for debug and development, but should not be
3205			needed on a platform with proper driver support.
3206
3207	pd.		[PARIDE]
3208			See Documentation/blockdev/paride.txt.
3209
3210	pdcchassis=	[PARISC,HW] Disable/Enable PDC Chassis Status codes at
3211			boot time.
3212			Format: { 0 | 1 }
3213			See arch/parisc/kernel/pdc_chassis.c
3214
3215	percpu_alloc=	Select which percpu first chunk allocator to use.
3216			Currently supported values are "embed" and "page".
3217			Archs may support subset or none of the	selections.
3218			See comments in mm/percpu.c for details on each
3219			allocator.  This parameter is primarily	for debugging
3220			and performance comparison.
3221
3222	pf.		[PARIDE]
3223			See Documentation/blockdev/paride.txt.
3224
3225	pg.		[PARIDE]
3226			See Documentation/blockdev/paride.txt.
3227
3228	pirq=		[SMP,APIC] Manual mp-table setup
3229			See Documentation/x86/i386/IO-APIC.txt.
3230
3231	plip=		[PPT,NET] Parallel port network link
3232			Format: { parport<nr> | timid | 0 }
3233			See also Documentation/parport.txt.
3234
3235	pmtmr=		[X86] Manual setup of pmtmr I/O Port.
3236			Override pmtimer IOPort with a hex value.
3237			e.g. pmtmr=0x508
3238
3239	pnp.debug=1	[PNP]
3240			Enable PNP debug messages (depends on the
3241			CONFIG_PNP_DEBUG_MESSAGES option).  Change at run-time
3242			via /sys/module/pnp/parameters/debug.  We always show
3243			current resource usage; turning this on also shows
3244			possible settings and some assignment information.
3245
3246	pnpacpi=	[ACPI]
3247			{ off }
3248
3249	pnpbios=	[ISAPNP]
3250			{ on | off | curr | res | no-curr | no-res }
3251
3252	pnp_reserve_irq=
3253			[ISAPNP] Exclude IRQs for the autoconfiguration
3254
3255	pnp_reserve_dma=
3256			[ISAPNP] Exclude DMAs for the autoconfiguration
3257
3258	pnp_reserve_io=	[ISAPNP] Exclude I/O ports for the autoconfiguration
3259			Ranges are in pairs (I/O port base and size).
3260
3261	pnp_reserve_mem=
3262			[ISAPNP] Exclude memory regions for the
3263			autoconfiguration.
3264			Ranges are in pairs (memory base and size).
3265
3266	ports=		[IP_VS_FTP] IPVS ftp helper module
3267			Default is 21.
3268			Up to 8 (IP_VS_APP_MAX_PORTS) ports
3269			may be specified.
3270			Format: <port>,<port>....
3271
3272	ppc_strict_facility_enable
3273			[PPC] This option catches any kernel floating point,
3274			Altivec, VSX and SPE outside of regions specifically
3275			allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3276			There is some performance impact when enabling this.
3277
3278	print-fatal-signals=
3279			[KNL] debug: print fatal signals
3280
3281			If enabled, warn about various signal handling
3282			related application anomalies: too many signals,
3283			too many POSIX.1 timers, fatal signals causing a
3284			coredump - etc.
3285
3286			If you hit the warning due to signal overflow,
3287			you might want to try "ulimit -i unlimited".
3288
3289			default: off.
3290
3291	printk.always_kmsg_dump=
3292			Trigger kmsg_dump for cases other than kernel oops or
3293			panics
3294			Format: <bool>  (1/Y/y=enable, 0/N/n=disable)
3295			default: disabled
3296
3297	printk.devkmsg={on,off,ratelimit}
3298			Control writing to /dev/kmsg.
3299			on - unlimited logging to /dev/kmsg from userspace
3300			off - logging to /dev/kmsg disabled
3301			ratelimit - ratelimit the logging
3302			Default: ratelimit
3303
3304	printk.time=	Show timing data prefixed to each printk message line
3305			Format: <bool>  (1/Y/y=enable, 0/N/n=disable)
3306
3307	processor.max_cstate=	[HW,ACPI]
3308			Limit processor to maximum C-state
3309			max_cstate=9 overrides any DMI blacklist limit.
3310
3311	processor.nocst	[HW,ACPI]
3312			Ignore the _CST method to determine C-states,
3313			instead using the legacy FADT method
3314
3315	profile=	[KNL] Enable kernel profiling via /proc/profile
3316			Format: [schedule,]<number>
3317			Param: "schedule" - profile schedule points.
3318			Param: <number> - step/bucket size as a power of 2 for
3319				statistical time based profiling.
3320			Param: "sleep" - profile D-state sleeping (millisecs).
3321				Requires CONFIG_SCHEDSTATS
3322			Param: "kvm" - profile VM exits.
3323
3324	prompt_ramdisk=	[RAM] List of RAM disks to prompt for floppy disk
3325			before loading.
3326			See Documentation/blockdev/ramdisk.txt.
3327
3328	psmouse.proto=	[HW,MOUSE] Highest PS2 mouse protocol extension to
3329			probe for; one of (bare|imps|exps|lifebook|any).
3330	psmouse.rate=	[HW,MOUSE] Set desired mouse report rate, in reports
3331			per second.
3332	psmouse.resetafter=	[HW,MOUSE]
3333			Try to reset the device after so many bad packets
3334			(0 = never).
3335	psmouse.resolution=
3336			[HW,MOUSE] Set desired mouse resolution, in dpi.
3337	psmouse.smartscroll=
3338			[HW,MOUSE] Controls Logitech smartscroll autorepeat.
3339			0 = disabled, 1 = enabled (default).
3340
3341	pstore.backend=	Specify the name of the pstore backend to use
3342
3343	pt.		[PARIDE]
3344			See Documentation/blockdev/paride.txt.
3345
3346	pti=		[X86_64] Control Page Table Isolation of user and
3347			kernel address spaces.  Disabling this feature
3348			removes hardening, but improves performance of
3349			system calls and interrupts.
3350
3351			on   - unconditionally enable
3352			off  - unconditionally disable
3353			auto - kernel detects whether your CPU model is
3354			       vulnerable to issues that PTI mitigates
3355
3356			Not specifying this option is equivalent to pti=auto.
3357
3358	nopti		[X86_64]
3359			Equivalent to pti=off
3360
3361	pty.legacy_count=
3362			[KNL] Number of legacy pty's. Overwrites compiled-in
3363			default number.
3364
3365	quiet		[KNL] Disable most log messages
3366
3367	r128=		[HW,DRM]
3368
3369	raid=		[HW,RAID]
3370			See Documentation/md.txt.
3371
3372	ramdisk_size=	[RAM] Sizes of RAM disks in kilobytes
3373			See Documentation/blockdev/ramdisk.txt.
3374
3375	rcu_nocbs=	[KNL]
3376			The argument is a cpu list, as described above.
3377
3378			In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3379			the specified list of CPUs to be no-callback CPUs.
3380			Invocation of these CPUs' RCU callbacks will
3381			be offloaded to "rcuox/N" kthreads created for
3382			that purpose, where "x" is "b" for RCU-bh, "p"
3383			for RCU-preempt, and "s" for RCU-sched, and "N"
3384			is the CPU number.  This reduces OS jitter on the
3385			offloaded CPUs, which can be useful for HPC and
3386			real-time workloads.  It can also improve energy
3387			efficiency for asymmetric multiprocessors.
3388
3389	rcu_nocb_poll	[KNL]
3390			Rather than requiring that offloaded CPUs
3391			(specified by rcu_nocbs= above) explicitly
3392			awaken the corresponding "rcuoN" kthreads,
3393			make these kthreads poll for callbacks.
3394			This improves the real-time response for the
3395			offloaded CPUs by relieving them of the need to
3396			wake up the corresponding kthread, but degrades
3397			energy efficiency by requiring that the kthreads
3398			periodically wake up to do the polling.
3399
3400	rcutree.blimit=	[KNL]
3401			Set maximum number of finished RCU callbacks to
3402			process in one batch.
3403
3404	rcutree.dump_tree=	[KNL]
3405			Dump the structure of the rcu_node combining tree
3406			out at early boot.  This is used for diagnostic
3407			purposes, to verify correct tree setup.
3408
3409	rcutree.gp_cleanup_delay=	[KNL]
3410			Set the number of jiffies to delay each step of
3411			RCU grace-period cleanup.  This only has effect
3412			when CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP is set.
3413
3414	rcutree.gp_init_delay=	[KNL]
3415			Set the number of jiffies to delay each step of
3416			RCU grace-period initialization.  This only has
3417			effect when CONFIG_RCU_TORTURE_TEST_SLOW_INIT
3418			is set.
3419
3420	rcutree.gp_preinit_delay=	[KNL]
3421			Set the number of jiffies to delay each step of
3422			RCU grace-period pre-initialization, that is,
3423			the propagation of recent CPU-hotplug changes up
3424			the rcu_node combining tree.  This only has effect
3425			when CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT is set.
3426
3427	rcutree.rcu_fanout_exact= [KNL]
3428			Disable autobalancing of the rcu_node combining
3429			tree.  This is used by rcutorture, and might
3430			possibly be useful for architectures having high
3431			cache-to-cache transfer latencies.
3432
3433	rcutree.rcu_fanout_leaf= [KNL]
3434			Change the number of CPUs assigned to each
3435			leaf rcu_node structure.  Useful for very
3436			large systems, which will choose the value 64,
3437			and for NUMA systems with large remote-access
3438			latencies, which will choose a value aligned
3439			with the appropriate hardware boundaries.
3440
3441	rcutree.jiffies_till_sched_qs= [KNL]
3442			Set required age in jiffies for a
3443			given grace period before RCU starts
3444			soliciting quiescent-state help from
3445			rcu_note_context_switch().
3446
3447	rcutree.jiffies_till_first_fqs= [KNL]
3448			Set delay from grace-period initialization to
3449			first attempt to force quiescent states.
3450			Units are jiffies, minimum value is zero,
3451			and maximum value is HZ.
3452
3453	rcutree.jiffies_till_next_fqs= [KNL]
3454			Set delay between subsequent attempts to force
3455			quiescent states.  Units are jiffies, minimum
3456			value is one, and maximum value is HZ.
3457
3458	rcutree.kthread_prio= 	 [KNL,BOOT]
3459			Set the SCHED_FIFO priority of the RCU per-CPU
3460			kthreads (rcuc/N). This value is also used for
3461			the priority of the RCU boost threads (rcub/N)
3462			and for the RCU grace-period kthreads (rcu_bh,
3463			rcu_preempt, and rcu_sched). If RCU_BOOST is
3464			set, valid values are 1-99 and the default is 1
3465			(the least-favored priority).  Otherwise, when
3466			RCU_BOOST is not set, valid values are 0-99 and
3467			the default is zero (non-realtime operation).
3468
3469	rcutree.rcu_nocb_leader_stride= [KNL]
3470			Set the number of NOCB kthread groups, which
3471			defaults to the square root of the number of
3472			CPUs.  Larger numbers reduces the wakeup overhead
3473			on the per-CPU grace-period kthreads, but increases
3474			that same overhead on each group's leader.
3475
3476	rcutree.qhimark= [KNL]
3477			Set threshold of queued RCU callbacks beyond which
3478			batch limiting is disabled.
3479
3480	rcutree.qlowmark= [KNL]
3481			Set threshold of queued RCU callbacks below which
3482			batch limiting is re-enabled.
3483
3484	rcutree.rcu_idle_gp_delay= [KNL]
3485			Set wakeup interval for idle CPUs that have
3486			RCU callbacks (RCU_FAST_NO_HZ=y).
3487
3488	rcutree.rcu_idle_lazy_gp_delay= [KNL]
3489			Set wakeup interval for idle CPUs that have
3490			only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3491			Lazy RCU callbacks are those which RCU can
3492			prove do nothing more than free memory.
3493
3494	rcuperf.gp_exp= [KNL]
3495			Measure performance of expedited synchronous
3496			grace-period primitives.
3497
3498	rcuperf.holdoff= [KNL]
3499			Set test-start holdoff period.  The purpose of
3500			this parameter is to delay the start of the
3501			test until boot completes in order to avoid
3502			interference.
3503
3504	rcuperf.nreaders= [KNL]
3505			Set number of RCU readers.  The value -1 selects
3506			N, where N is the number of CPUs.  A value
3507			"n" less than -1 selects N-n+1, where N is again
3508			the number of CPUs.  For example, -2 selects N
3509			(the number of CPUs), -3 selects N+1, and so on.
3510			A value of "n" less than or equal to -N selects
3511			a single reader.
3512
3513	rcuperf.nwriters= [KNL]
3514			Set number of RCU writers.  The values operate
3515			the same as for rcuperf.nreaders.
3516			N, where N is the number of CPUs
3517
3518	rcuperf.perf_runnable= [BOOT]
3519			Start rcuperf running at boot time.
3520
3521	rcuperf.shutdown= [KNL]
3522			Shut the system down after performance tests
3523			complete.  This is useful for hands-off automated
3524			testing.
3525
3526	rcuperf.perf_type= [KNL]
3527			Specify the RCU implementation to test.
3528
3529	rcuperf.verbose= [KNL]
3530			Enable additional printk() statements.
3531
3532	rcutorture.cbflood_inter_holdoff= [KNL]
3533			Set holdoff time (jiffies) between successive
3534			callback-flood tests.
3535
3536	rcutorture.cbflood_intra_holdoff= [KNL]
3537			Set holdoff time (jiffies) between successive
3538			bursts of callbacks within a given callback-flood
3539			test.
3540
3541	rcutorture.cbflood_n_burst= [KNL]
3542			Set the number of bursts making up a given
3543			callback-flood test.  Set this to zero to
3544			disable callback-flood testing.
3545
3546	rcutorture.cbflood_n_per_burst= [KNL]
3547			Set the number of callbacks to be registered
3548			in a given burst of a callback-flood test.
3549
3550	rcutorture.fqs_duration= [KNL]
3551			Set duration of force_quiescent_state bursts
3552			in microseconds.
3553
3554	rcutorture.fqs_holdoff= [KNL]
3555			Set holdoff time within force_quiescent_state bursts
3556			in microseconds.
3557
3558	rcutorture.fqs_stutter= [KNL]
3559			Set wait time between force_quiescent_state bursts
3560			in seconds.
3561
3562	rcutorture.gp_cond= [KNL]
3563			Use conditional/asynchronous update-side
3564			primitives, if available.
3565
3566	rcutorture.gp_exp= [KNL]
3567			Use expedited update-side primitives, if available.
3568
3569	rcutorture.gp_normal= [KNL]
3570			Use normal (non-expedited) asynchronous
3571			update-side primitives, if available.
3572
3573	rcutorture.gp_sync= [KNL]
3574			Use normal (non-expedited) synchronous
3575			update-side primitives, if available.  If all
3576			of rcutorture.gp_cond=, rcutorture.gp_exp=,
3577			rcutorture.gp_normal=, and rcutorture.gp_sync=
3578			are zero, rcutorture acts as if is interpreted
3579			they are all non-zero.
3580
3581	rcutorture.n_barrier_cbs= [KNL]
3582			Set callbacks/threads for rcu_barrier() testing.
3583
3584	rcutorture.nfakewriters= [KNL]
3585			Set number of concurrent RCU writers.  These just
3586			stress RCU, they don't participate in the actual
3587			test, hence the "fake".
3588
3589	rcutorture.nreaders= [KNL]
3590			Set number of RCU readers.  The value -1 selects
3591			N-1, where N is the number of CPUs.  A value
3592			"n" less than -1 selects N-n-2, where N is again
3593			the number of CPUs.  For example, -2 selects N
3594			(the number of CPUs), -3 selects N+1, and so on.
3595
3596	rcutorture.object_debug= [KNL]
3597			Enable debug-object double-call_rcu() testing.
3598
3599	rcutorture.onoff_holdoff= [KNL]
3600			Set time (s) after boot for CPU-hotplug testing.
3601
3602	rcutorture.onoff_interval= [KNL]
3603			Set time (s) between CPU-hotplug operations, or
3604			zero to disable CPU-hotplug testing.
3605
3606	rcutorture.shuffle_interval= [KNL]
3607			Set task-shuffle interval (s).  Shuffling tasks
3608			allows some CPUs to go into dyntick-idle mode
3609			during the rcutorture test.
3610
3611	rcutorture.shutdown_secs= [KNL]
3612			Set time (s) after boot system shutdown.  This
3613			is useful for hands-off automated testing.
3614
3615	rcutorture.stall_cpu= [KNL]
3616			Duration of CPU stall (s) to test RCU CPU stall
3617			warnings, zero to disable.
3618
3619	rcutorture.stall_cpu_holdoff= [KNL]
3620			Time to wait (s) after boot before inducing stall.
3621
3622	rcutorture.stat_interval= [KNL]
3623			Time (s) between statistics printk()s.
3624
3625	rcutorture.stutter= [KNL]
3626			Time (s) to stutter testing, for example, specifying
3627			five seconds causes the test to run for five seconds,
3628			wait for five seconds, and so on.  This tests RCU's
3629			ability to transition abruptly to and from idle.
3630
3631	rcutorture.test_boost= [KNL]
3632			Test RCU priority boosting?  0=no, 1=maybe, 2=yes.
3633			"Maybe" means test if the RCU implementation
3634			under test support RCU priority boosting.
3635
3636	rcutorture.test_boost_duration= [KNL]
3637			Duration (s) of each individual boost test.
3638
3639	rcutorture.test_boost_interval= [KNL]
3640			Interval (s) between each boost test.
3641
3642	rcutorture.test_no_idle_hz= [KNL]
3643			Test RCU's dyntick-idle handling.  See also the
3644			rcutorture.shuffle_interval parameter.
3645
3646	rcutorture.torture_runnable= [BOOT]
3647			Start rcutorture running at boot time.
3648
3649	rcutorture.torture_type= [KNL]
3650			Specify the RCU implementation to test.
3651
3652	rcutorture.verbose= [KNL]
3653			Enable additional printk() statements.
3654
3655	rcupdate.rcu_cpu_stall_suppress= [KNL]
3656			Suppress RCU CPU stall warning messages.
3657
3658	rcupdate.rcu_cpu_stall_timeout= [KNL]
3659			Set timeout for RCU CPU stall warning messages.
3660
3661	rcupdate.rcu_expedited= [KNL]
3662			Use expedited grace-period primitives, for
3663			example, synchronize_rcu_expedited() instead
3664			of synchronize_rcu().  This reduces latency,
3665			but can increase CPU utilization, degrade
3666			real-time latency, and degrade energy efficiency.
3667			No effect on CONFIG_TINY_RCU kernels.
3668
3669	rcupdate.rcu_normal= [KNL]
3670			Use only normal grace-period primitives,
3671			for example, synchronize_rcu() instead of
3672			synchronize_rcu_expedited().  This improves
3673			real-time latency, CPU utilization, and
3674			energy efficiency, but can expose users to
3675			increased grace-period latency.  This parameter
3676			overrides rcupdate.rcu_expedited.  No effect on
3677			CONFIG_TINY_RCU kernels.
3678
3679	rcupdate.rcu_normal_after_boot= [KNL]
3680			Once boot has completed (that is, after
3681			rcu_end_inkernel_boot() has been invoked), use
3682			only normal grace-period primitives.  No effect
3683			on CONFIG_TINY_RCU kernels.
3684
3685	rcupdate.rcu_task_stall_timeout= [KNL]
3686			Set timeout in jiffies for RCU task stall warning
3687			messages.  Disable with a value less than or equal
3688			to zero.
3689
3690	rcupdate.rcu_self_test= [KNL]
3691			Run the RCU early boot self tests
3692
3693	rcupdate.rcu_self_test_bh= [KNL]
3694			Run the RCU bh early boot self tests
3695
3696	rcupdate.rcu_self_test_sched= [KNL]
3697			Run the RCU sched early boot self tests
3698
3699	rdinit=		[KNL]
3700			Format: <full_path>
3701			Run specified binary instead of /init from the ramdisk,
3702			used for early userspace startup. See initrd.
3703
3704	reboot=		[KNL]
3705			Format (x86 or x86_64):
3706				[w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3707				[[,]s[mp]#### \
3708				[[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3709				[[,]f[orce]
3710			Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3711			      reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3712			      reboot_force is either force or not specified,
3713			      reboot_cpu is s[mp]#### with #### being the processor
3714					to be used for rebooting.
3715
3716	relax_domain_level=
3717			[KNL, SMP] Set scheduler's default relax_domain_level.
3718			See Documentation/cgroup-v1/cpusets.txt.
3719
3720	relative_sleep_states=
3721			[SUSPEND] Use sleep state labeling where the deepest
3722			state available other than hibernation is always "mem".
3723			Format: { "0" | "1" }
3724			0 -- Traditional sleep state labels.
3725			1 -- Relative sleep state labels.
3726
3727	reserve=	[KNL,BUGS] Force the kernel to ignore some iomem area
3728
3729	reservetop=	[X86-32]
3730			Format: nn[KMG]
3731			Reserves a hole at the top of the kernel virtual
3732			address space.
3733
3734	reservelow=	[X86]
3735			Format: nn[K]
3736			Set the amount of memory to reserve for BIOS at
3737			the bottom of the address space.
3738
3739	reset_devices	[KNL] Force drivers to reset the underlying device
3740			during initialization.
3741
3742	resume=		[SWSUSP]
3743			Specify the partition device for software suspend
3744			Format:
3745			{/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3746
3747	resume_offset=	[SWSUSP]
3748			Specify the offset from the beginning of the partition
3749			given by "resume=" at which the swap header is located,
3750			in <PAGE_SIZE> units (needed only for swap files).
3751			See  Documentation/power/swsusp-and-swap-files.txt
3752
3753	resumedelay=	[HIBERNATION] Delay (in seconds) to pause before attempting to
3754			read the resume files
3755
3756	resumewait	[HIBERNATION] Wait (indefinitely) for resume device to show up.
3757			Useful for devices that are detected asynchronously
3758			(e.g. USB and MMC devices).
3759
3760	hibernate=	[HIBERNATION]
3761		noresume	Don't check if there's a hibernation image
3762				present during boot.
3763		nocompress	Don't compress/decompress hibernation images.
3764		no		Disable hibernation and resume.
3765		protect_image	Turn on image protection during restoration
3766				(that will set all pages holding image data
3767				during restoration read-only).
3768
3769	retain_initrd	[RAM] Keep initrd memory after extraction
3770
3771	rfkill.default_state=
3772		0	"airplane mode".  All wifi, bluetooth, wimax, gps, fm,
3773			etc. communication is blocked by default.
3774		1	Unblocked.
3775
3776	rfkill.master_switch_mode=
3777		0	The "airplane mode" button does nothing.
3778		1	The "airplane mode" button toggles between everything
3779			blocked and the previous configuration.
3780		2	The "airplane mode" button toggles between everything
3781			blocked and everything unblocked.
3782
3783	rhash_entries=	[KNL,NET]
3784			Set number of hash buckets for route cache
3785
3786	ro		[KNL] Mount root device read-only on boot
3787
3788	rodata=		[KNL]
3789		on	Mark read-only kernel memory as read-only (default).
3790		off	Leave read-only kernel memory writable for debugging.
3791
3792	rockchip.usb_uart
3793			Enable the uart passthrough on the designated usb port
3794			on Rockchip SoCs. When active, the signals of the
3795			debug-uart get routed to the D+ and D- pins of the usb
3796			port and the regular usb controller gets disabled.
3797
3798	root=		[KNL] Root filesystem
3799			See name_to_dev_t comment in init/do_mounts.c.
3800
3801	rootdelay=	[KNL] Delay (in seconds) to pause before attempting to
3802			mount the root filesystem
3803
3804	rootflags=	[KNL] Set root filesystem mount option string
3805
3806	rootfstype=	[KNL] Set root filesystem type
3807
3808	rootwait	[KNL] Wait (indefinitely) for root device to show up.
3809			Useful for devices that are detected asynchronously
3810			(e.g. USB and MMC devices).
3811
3812	rproc_mem=nn[KMG][@address]
3813			[KNL,ARM,CMA] Remoteproc physical memory block.
3814			Memory area to be used by remote processor image,
3815			managed by CMA.
3816
3817	rw		[KNL] Mount root device read-write on boot
3818
3819	S		[KNL] Run init in single mode
3820
3821	s390_iommu=	[HW,S390]
3822			Set s390 IOTLB flushing mode
3823		strict
3824			With strict flushing every unmap operation will result in
3825			an IOTLB flush. Default is lazy flushing before reuse,
3826			which is faster.
3827
3828	sa1100ir	[NET]
3829			See drivers/net/irda/sa1100_ir.c.
3830
3831	sbni=		[NET] Granch SBNI12 leased line adapter
3832
3833	sched_debug	[KNL] Enables verbose scheduler debug messages.
3834
3835	schedstats=	[KNL,X86] Enable or disable scheduled statistics.
3836			Allowed values are enable and disable. This feature
3837			incurs a small amount of overhead in the scheduler
3838			but is useful for debugging and performance tuning.
3839
3840	skew_tick=	[KNL] Offset the periodic timer tick per cpu to mitigate
3841			xtime_lock contention on larger systems, and/or RCU lock
3842			contention on all systems with CONFIG_MAXSMP set.
3843			Format: { "0" | "1" }
3844			0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3845			1 -- enable.
3846			Note: increases power consumption, thus should only be
3847			enabled if running jitter sensitive (HPC/RT) workloads.
3848
3849	security=	[SECURITY] Choose a security module to enable at boot.
3850			If this boot parameter is not specified, only the first
3851			security module asking for security registration will be
3852			loaded. An invalid security module name will be treated
3853			as if no module has been chosen.
3854
3855	selinux=	[SELINUX] Disable or enable SELinux at boot time.
3856			Format: { "0" | "1" }
3857			See security/selinux/Kconfig help text.
3858			0 -- disable.
3859			1 -- enable.
3860			Default value is set via kernel config option.
3861			If enabled at boot time, /selinux/disable can be used
3862			later to disable prior to initial policy load.
3863
3864	apparmor=	[APPARMOR] Disable or enable AppArmor at boot time
3865			Format: { "0" | "1" }
3866			See security/apparmor/Kconfig help text
3867			0 -- disable.
3868			1 -- enable.
3869			Default value is set via kernel config option.
3870
3871	serialnumber	[BUGS=X86-32]
3872
3873	shapers=	[NET]
3874			Maximal number of shapers.
3875
3876	show_msr=	[x86] show boot-time MSR settings
3877			Format: { <integer> }
3878			Show boot-time (BIOS-initialized) MSR settings.
3879			The parameter means the number of CPUs to show,
3880			for example 1 means boot CPU only.
3881
3882	simeth=		[IA-64]
3883	simscsi=
3884
3885	slram=		[HW,MTD]
3886
3887	slab_nomerge	[MM]
3888			Disable merging of slabs with similar size. May be
3889			necessary if there is some reason to distinguish
3890			allocs to different slabs. Debug options disable
3891			merging on their own.
3892			For more information see Documentation/vm/slub.txt.
3893
3894	slab_max_order=	[MM, SLAB]
3895			Determines the maximum allowed order for slabs.
3896			A high setting may cause OOMs due to memory
3897			fragmentation.  Defaults to 1 for systems with
3898			more than 32MB of RAM, 0 otherwise.
3899
3900	slub_debug[=options[,slabs]]	[MM, SLUB]
3901			Enabling slub_debug allows one to determine the
3902			culprit if slab objects become corrupted. Enabling
3903			slub_debug can create guard zones around objects and
3904			may poison objects when not in use. Also tracks the
3905			last alloc / free. For more information see
3906			Documentation/vm/slub.txt.
3907
3908	slub_max_order= [MM, SLUB]
3909			Determines the maximum allowed order for slabs.
3910			A high setting may cause OOMs due to memory
3911			fragmentation. For more information see
3912			Documentation/vm/slub.txt.
3913
3914	slub_min_objects=	[MM, SLUB]
3915			The minimum number of objects per slab. SLUB will
3916			increase the slab order up to slub_max_order to
3917			generate a sufficiently large slab able to contain
3918			the number of objects indicated. The higher the number
3919			of objects the smaller the overhead of tracking slabs
3920			and the less frequently locks need to be acquired.
3921			For more information see Documentation/vm/slub.txt.
3922
3923	slub_min_order=	[MM, SLUB]
3924			Determines the minimum page order for slabs. Must be
3925			lower than slub_max_order.
3926			For more information see Documentation/vm/slub.txt.
3927
3928	slub_nomerge	[MM, SLUB]
3929			Same with slab_nomerge. This is supported for legacy.
3930			See slab_nomerge for more information.
3931
3932	smart2=		[HW]
3933			Format: <io1>[,<io2>[,...,<io8>]]
3934
3935	smsc-ircc2.nopnp	[HW] Don't use PNP to discover SMC devices
3936	smsc-ircc2.ircc_cfg=	[HW] Device configuration I/O port
3937	smsc-ircc2.ircc_sir=	[HW] SIR base I/O port
3938	smsc-ircc2.ircc_fir=	[HW] FIR base I/O port
3939	smsc-ircc2.ircc_irq=	[HW] IRQ line
3940	smsc-ircc2.ircc_dma=	[HW] DMA channel
3941	smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3942				0: Toshiba Satellite 1800 (GP data pin select)
3943				1: Fast pin select (default)
3944				2: ATC IRMode
3945
3946	smt		[KNL,S390] Set the maximum number of threads (logical
3947			CPUs) to use per physical CPU on systems capable of
3948			symmetric multithreading (SMT). Will be capped to the
3949			actual hardware limit.
3950			Format: <integer>
3951			Default: -1 (no limit)
3952
3953	softlockup_panic=
3954			[KNL] Should the soft-lockup detector generate panics.
3955			Format: <integer>
3956
3957	softlockup_all_cpu_backtrace=
3958			[KNL] Should the soft-lockup detector generate
3959			backtraces on all cpus.
3960			Format: <integer>
3961
3962	sonypi.*=	[HW] Sony Programmable I/O Control Device driver
3963			See Documentation/laptops/sonypi.txt
3964
3965	spectre_v2=	[X86] Control mitigation of Spectre variant 2
3966			(indirect branch speculation) vulnerability.
3967
3968			on   - unconditionally enable
3969			off  - unconditionally disable
3970			auto - kernel detects whether your CPU model is
3971			       vulnerable
3972
3973			Selecting 'on' will, and 'auto' may, choose a
3974			mitigation method at run time according to the
3975			CPU, the available microcode, the setting of the
3976			CONFIG_RETPOLINE configuration option, and the
3977			compiler with which the kernel was built.
3978
3979			Specific mitigations can also be selected manually:
3980
3981			retpoline	  - replace indirect branches
3982			retpoline,generic - google's original retpoline
3983			retpoline,amd     - AMD-specific minimal thunk
3984
3985			Not specifying this option is equivalent to
3986			spectre_v2=auto.
3987
3988	spia_io_base=	[HW,MTD]
3989	spia_fio_base=
3990	spia_pedr=
3991	spia_peddr=
3992
3993	stack_guard_gap=	[MM]
3994			override the default stack gap protection. The value
3995			is in page units and it defines how many pages prior
3996			to (for stacks growing down) resp. after (for stacks
3997			growing up) the main stack are reserved for no other
3998			mapping. Default value is 256 pages.
3999
4000	stacktrace	[FTRACE]
4001			Enabled the stack tracer on boot up.
4002
4003	stacktrace_filter=[function-list]
4004			[FTRACE] Limit the functions that the stack tracer
4005			will trace at boot up. function-list is a comma separated
4006			list of functions. This list can be changed at run
4007			time by the stack_trace_filter file in the debugfs
4008			tracing directory. Note, this enables stack tracing
4009			and the stacktrace above is not needed.
4010
4011	sti=		[PARISC,HW]
4012			Format: <num>
4013			Set the STI (builtin display/keyboard on the HP-PARISC
4014			machines) console (graphic card) which should be used
4015			as the initial boot-console.
4016			See also comment in drivers/video/console/sticore.c.
4017
4018	sti_font=	[HW]
4019			See comment in drivers/video/console/sticore.c.
4020
4021	stifb=		[HW]
4022			Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4023
4024	sunrpc.min_resvport=
4025	sunrpc.max_resvport=
4026			[NFS,SUNRPC]
4027			SunRPC servers often require that client requests
4028			originate from a privileged port (i.e. a port in the
4029			range 0 < portnr < 1024).
4030			An administrator who wishes to reserve some of these
4031			ports for other uses may adjust the range that the
4032			kernel's sunrpc client considers to be privileged
4033			using these two parameters to set the minimum and
4034			maximum port values.
4035
4036	sunrpc.svc_rpc_per_connection_limit=
4037			[NFS,SUNRPC]
4038			Limit the number of requests that the server will
4039			process in parallel from a single connection.
4040			The default value is 0 (no limit).
4041
4042	sunrpc.pool_mode=
4043			[NFS]
4044			Control how the NFS server code allocates CPUs to
4045			service thread pools.  Depending on how many NICs
4046			you have and where their interrupts are bound, this
4047			option will affect which CPUs will do NFS serving.
4048			Note: this parameter cannot be changed while the
4049			NFS server is running.
4050
4051			auto	    the server chooses an appropriate mode
4052				    automatically using heuristics
4053			global	    a single global pool contains all CPUs
4054			percpu	    one pool for each CPU
4055			pernode	    one pool for each NUMA node (equivalent
4056				    to global on non-NUMA machines)
4057
4058	sunrpc.tcp_slot_table_entries=
4059	sunrpc.udp_slot_table_entries=
4060			[NFS,SUNRPC]
4061			Sets the upper limit on the number of simultaneous
4062			RPC calls that can be sent from the client to a
4063			server. Increasing these values may allow you to
4064			improve throughput, but will also increase the
4065			amount of memory reserved for use by the client.
4066
4067	suspend.pm_test_delay=
4068			[SUSPEND]
4069			Sets the number of seconds to remain in a suspend test
4070			mode before resuming the system (see
4071			/sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4072			is set. Default value is 5.
4073
4074	swapaccount=[0|1]
4075			[KNL] Enable accounting of swap in memory resource
4076			controller if no parameter or 1 is given or disable
4077			it if 0 is given (See Documentation/cgroup-v1/memory.txt)
4078
4079	swiotlb=	[ARM,IA-64,PPC,MIPS,X86]
4080			Format: { <int> | force | noforce }
4081			<int> -- Number of I/O TLB slabs
4082			force -- force using of bounce buffers even if they
4083			         wouldn't be automatically used by the kernel
4084			noforce -- Never use bounce buffers (for debugging)
4085
4086	switches=	[HW,M68k]
4087
4088	sysfs.deprecated=0|1 [KNL]
4089			Enable/disable old style sysfs layout for old udev
4090			on older distributions. When this option is enabled
4091			very new udev will not work anymore. When this option
4092			is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4093			in older udev will not work anymore.
4094			Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4095			the kernel configuration.
4096
4097	sysrq_always_enabled
4098			[KNL]
4099			Ignore sysrq setting - this boot parameter will
4100			neutralize any effect of /proc/sys/kernel/sysrq.
4101			Useful for debugging.
4102
4103	tcpmhash_entries= [KNL,NET]
4104			Set the number of tcp_metrics_hash slots.
4105			Default value is 8192 or 16384 depending on total
4106			ram pages. This is used to specify the TCP metrics
4107			cache size. See Documentation/networking/ip-sysctl.txt
4108			"tcp_no_metrics_save" section for more details.
4109
4110	tdfx=		[HW,DRM]
4111
4112	test_suspend=	[SUSPEND][,N]
4113			Specify "mem" (for Suspend-to-RAM) or "standby" (for
4114			standby suspend) or "freeze" (for suspend type freeze)
4115			as the system sleep state during system startup with
4116			the optional capability to repeat N number of times.
4117			The system is woken from this state using a
4118			wakeup-capable RTC alarm.
4119
4120	thash_entries=	[KNL,NET]
4121			Set number of hash buckets for TCP connection
4122
4123	thermal.act=	[HW,ACPI]
4124			-1: disable all active trip points in all thermal zones
4125			<degrees C>: override all lowest active trip points
4126
4127	thermal.crt=	[HW,ACPI]
4128			-1: disable all critical trip points in all thermal zones
4129			<degrees C>: override all critical trip points
4130
4131	thermal.nocrt=	[HW,ACPI]
4132			Set to disable actions on ACPI thermal zone
4133			critical and hot trip points.
4134
4135	thermal.off=	[HW,ACPI]
4136			1: disable ACPI thermal control
4137
4138	thermal.psv=	[HW,ACPI]
4139			-1: disable all passive trip points
4140			<degrees C>: override all passive trip points to this
4141			value
4142
4143	thermal.tzp=	[HW,ACPI]
4144			Specify global default ACPI thermal zone polling rate
4145			<deci-seconds>: poll all this frequency
4146			0: no polling (default)
4147
4148	threadirqs	[KNL]
4149			Force threading of all interrupt handlers except those
4150			marked explicitly IRQF_NO_THREAD.
4151
4152	tmem		[KNL,XEN]
4153			Enable the Transcendent memory driver if built-in.
4154
4155	tmem.cleancache=0|1 [KNL, XEN]
4156			Default is on (1). Disable the usage of the cleancache
4157			API to send anonymous pages to the hypervisor.
4158
4159	tmem.frontswap=0|1 [KNL, XEN]
4160			Default is on (1). Disable the usage of the frontswap
4161			API to send swap pages to the hypervisor. If disabled
4162			the selfballooning and selfshrinking are force disabled.
4163
4164	tmem.selfballooning=0|1 [KNL, XEN]
4165			Default is on (1). Disable the driving of swap pages
4166			to the hypervisor.
4167
4168	tmem.selfshrinking=0|1 [KNL, XEN]
4169			Default is on (1). Partial swapoff that immediately
4170			transfers pages from Xen hypervisor back to the
4171			kernel based on different criteria.
4172
4173	topology=	[S390]
4174			Format: {off | on}
4175			Specify if the kernel should make use of the cpu
4176			topology information if the hardware supports this.
4177			The scheduler will make use of this information and
4178			e.g. base its process migration decisions on it.
4179			Default is on.
4180
4181	topology_updates= [KNL, PPC, NUMA]
4182			Format: {off}
4183			Specify if the kernel should ignore (off)
4184			topology updates sent by the hypervisor to this
4185			LPAR.
4186
4187	tp720=		[HW,PS2]
4188
4189	tpm_suspend_pcr=[HW,TPM]
4190			Format: integer pcr id
4191			Specify that at suspend time, the tpm driver
4192			should extend the specified pcr with zeros,
4193			as a workaround for some chips which fail to
4194			flush the last written pcr on TPM_SaveState.
4195			This will guarantee that all the other pcrs
4196			are saved.
4197
4198	trace_buf_size=nn[KMG]
4199			[FTRACE] will set tracing buffer size on each cpu.
4200
4201	trace_event=[event-list]
4202			[FTRACE] Set and start specified trace events in order
4203			to facilitate early boot debugging. The event-list is a
4204			comma separated list of trace events to enable. See
4205			also Documentation/trace/events.txt
4206
4207	trace_options=[option-list]
4208			[FTRACE] Enable or disable tracer options at boot.
4209			The option-list is a comma delimited list of options
4210			that can be enabled or disabled just as if you were
4211			to echo the option name into
4212
4213			    /sys/kernel/debug/tracing/trace_options
4214
4215			For example, to enable stacktrace option (to dump the
4216			stack trace of each event), add to the command line:
4217
4218			      trace_options=stacktrace
4219
4220			See also Documentation/trace/ftrace.txt "trace options"
4221			section.
4222
4223	tp_printk[FTRACE]
4224			Have the tracepoints sent to printk as well as the
4225			tracing ring buffer. This is useful for early boot up
4226			where the system hangs or reboots and does not give the
4227			option for reading the tracing buffer or performing a
4228			ftrace_dump_on_oops.
4229
4230			To turn off having tracepoints sent to printk,
4231			 echo 0 > /proc/sys/kernel/tracepoint_printk
4232			Note, echoing 1 into this file without the
4233			tracepoint_printk kernel cmdline option has no effect.
4234
4235			** CAUTION **
4236
4237			Having tracepoints sent to printk() and activating high
4238			frequency tracepoints such as irq or sched, can cause
4239			the system to live lock.
4240
4241	traceoff_on_warning
4242			[FTRACE] enable this option to disable tracing when a
4243			warning is hit. This turns off "tracing_on". Tracing can
4244			be enabled again by echoing '1' into the "tracing_on"
4245			file located in /sys/kernel/debug/tracing/
4246
4247			This option is useful, as it disables the trace before
4248			the WARNING dump is called, which prevents the trace to
4249			be filled with content caused by the warning output.
4250
4251			This option can also be set at run time via the sysctl
4252			option:  kernel/traceoff_on_warning
4253
4254	transparent_hugepage=
4255			[KNL]
4256			Format: [always|madvise|never]
4257			Can be used to control the default behavior of the system
4258			with respect to transparent hugepages.
4259			See Documentation/vm/transhuge.txt for more details.
4260
4261	tsc=		Disable clocksource stability checks for TSC.
4262			Format: <string>
4263			[x86] reliable: mark tsc clocksource as reliable, this
4264			disables clocksource verification at runtime, as well
4265			as the stability checks done at bootup.	Used to enable
4266			high-resolution timer mode on older hardware, and in
4267			virtualized environment.
4268			[x86] noirqtime: Do not use TSC to do irq accounting.
4269			Used to run time disable IRQ_TIME_ACCOUNTING on any
4270			platforms where RDTSC is slow and this accounting
4271			can add overhead.
4272
4273	turbografx.map[2|3]=	[HW,JOY]
4274			TurboGraFX parallel port interface
4275			Format:
4276			<port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4277			See also Documentation/input/joystick-parport.txt
4278
4279	udbg-immortal	[PPC] When debugging early kernel crashes that
4280			happen after console_init() and before a proper
4281			console driver takes over, this boot options might
4282			help "seeing" what's going on.
4283
4284	uhash_entries=	[KNL,NET]
4285			Set number of hash buckets for UDP/UDP-Lite connections
4286
4287	uhci-hcd.ignore_oc=
4288			[USB] Ignore overcurrent events (default N).
4289			Some badly-designed motherboards generate lots of
4290			bogus events, for ports that aren't wired to
4291			anything.  Set this parameter to avoid log spamming.
4292			Note that genuine overcurrent events won't be
4293			reported either.
4294
4295	unknown_nmi_panic
4296			[X86] Cause panic on unknown NMI.
4297
4298	usbcore.authorized_default=
4299			[USB] Default USB device authorization:
4300			(default -1 = authorized except for wireless USB,
4301			0 = not authorized, 1 = authorized)
4302
4303	usbcore.autosuspend=
4304			[USB] The autosuspend time delay (in seconds) used
4305			for newly-detected USB devices (default 2).  This
4306			is the time required before an idle device will be
4307			autosuspended.  Devices for which the delay is set
4308			to a negative value won't be autosuspended at all.
4309
4310	usbcore.usbfs_snoop=
4311			[USB] Set to log all usbfs traffic (default 0 = off).
4312
4313	usbcore.usbfs_snoop_max=
4314			[USB] Maximum number of bytes to snoop in each URB
4315			(default = 65536).
4316
4317	usbcore.blinkenlights=
4318			[USB] Set to cycle leds on hubs (default 0 = off).
4319
4320	usbcore.old_scheme_first=
4321			[USB] Start with the old device initialization
4322			scheme (default 0 = off).
4323
4324	usbcore.usbfs_memory_mb=
4325			[USB] Memory limit (in MB) for buffers allocated by
4326			usbfs (default = 16, 0 = max = 2047).
4327
4328	usbcore.use_both_schemes=
4329			[USB] Try the other device initialization scheme
4330			if the first one fails (default 1 = enabled).
4331
4332	usbcore.initial_descriptor_timeout=
4333			[USB] Specifies timeout for the initial 64-byte
4334                        USB_REQ_GET_DESCRIPTOR request in milliseconds
4335			(default 5000 = 5.0 seconds).
4336
4337	usbcore.nousb	[USB] Disable the USB subsystem
4338
4339	usbhid.mousepoll=
4340			[USBHID] The interval which mice are to be polled at.
4341
4342	usb-storage.delay_use=
4343			[UMS] The delay in seconds before a new device is
4344			scanned for Logical Units (default 1).
4345
4346	usb-storage.quirks=
4347			[UMS] A list of quirks entries to supplement or
4348			override the built-in unusual_devs list.  List
4349			entries are separated by commas.  Each entry has
4350			the form VID:PID:Flags where VID and PID are Vendor
4351			and Product ID values (4-digit hex numbers) and
4352			Flags is a set of characters, each corresponding
4353			to a common usb-storage quirk flag as follows:
4354				a = SANE_SENSE (collect more than 18 bytes
4355					of sense data);
4356				b = BAD_SENSE (don't collect more than 18
4357					bytes of sense data);
4358				c = FIX_CAPACITY (decrease the reported
4359					device capacity by one sector);
4360				d = NO_READ_DISC_INFO (don't use
4361					READ_DISC_INFO command);
4362				e = NO_READ_CAPACITY_16 (don't use
4363					READ_CAPACITY_16 command);
4364				f = NO_REPORT_OPCODES (don't use report opcodes
4365					command, uas only);
4366				g = MAX_SECTORS_240 (don't transfer more than
4367					240 sectors at a time, uas only);
4368				h = CAPACITY_HEURISTICS (decrease the
4369					reported device capacity by one
4370					sector if the number is odd);
4371				i = IGNORE_DEVICE (don't bind to this
4372					device);
4373				j = NO_REPORT_LUNS (don't use report luns
4374					command, uas only);
4375				l = NOT_LOCKABLE (don't try to lock and
4376					unlock ejectable media);
4377				m = MAX_SECTORS_64 (don't transfer more
4378					than 64 sectors = 32 KB at a time);
4379				n = INITIAL_READ10 (force a retry of the
4380					initial READ(10) command);
4381				o = CAPACITY_OK (accept the capacity
4382					reported by the device);
4383				p = WRITE_CACHE (the device cache is ON
4384					by default);
4385				r = IGNORE_RESIDUE (the device reports
4386					bogus residue values);
4387				s = SINGLE_LUN (the device has only one
4388					Logical Unit);
4389				t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4390					commands, uas only);
4391				u = IGNORE_UAS (don't bind to the uas driver);
4392				w = NO_WP_DETECT (don't test whether the
4393					medium is write-protected).
4394				y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4395					even if the device claims no cache)
4396			Example: quirks=0419:aaf5:rl,0421:0433:rc
4397
4398	user_debug=	[KNL,ARM]
4399			Format: <int>
4400			See arch/arm/Kconfig.debug help text.
4401				 1 - undefined instruction events
4402				 2 - system calls
4403				 4 - invalid data aborts
4404				 8 - SIGSEGV faults
4405				16 - SIGBUS faults
4406			Example: user_debug=31
4407
4408	userpte=
4409			[X86] Flags controlling user PTE allocations.
4410
4411				nohigh = do not allocate PTE pages in
4412					HIGHMEM regardless of setting
4413					of CONFIG_HIGHPTE.
4414
4415	vdso=		[X86,SH]
4416			On X86_32, this is an alias for vdso32=.  Otherwise:
4417
4418			vdso=1: enable VDSO (the default)
4419			vdso=0: disable VDSO mapping
4420
4421	vdso32=		[X86] Control the 32-bit vDSO
4422			vdso32=1: enable 32-bit VDSO
4423			vdso32=0 or vdso32=2: disable 32-bit VDSO
4424
4425			See the help text for CONFIG_COMPAT_VDSO for more
4426			details.  If CONFIG_COMPAT_VDSO is set, the default is
4427			vdso32=0; otherwise, the default is vdso32=1.
4428
4429			For compatibility with older kernels, vdso32=2 is an
4430			alias for vdso32=0.
4431
4432			Try vdso32=0 if you encounter an error that says:
4433			dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4434
4435	vector=		[IA-64,SMP]
4436			vector=percpu: enable percpu vector domain
4437
4438	video=		[FB] Frame buffer configuration
4439			See Documentation/fb/modedb.txt.
4440
4441	video.brightness_switch_enabled= [0,1]
4442			If set to 1, on receiving an ACPI notify event
4443			generated by hotkey, video driver will adjust brightness
4444			level and then send out the event to user space through
4445			the allocated input device; If set to 0, video driver
4446			will only send out the event without touching backlight
4447			brightness level.
4448			default: 1
4449
4450	virtio_mmio.device=
4451			[VMMIO] Memory mapped virtio (platform) device.
4452
4453				<size>@<baseaddr>:<irq>[:<id>]
4454			where:
4455				<size>     := size (can use standard suffixes
4456						like K, M and G)
4457				<baseaddr> := physical base address
4458				<irq>      := interrupt number (as passed to
4459						request_irq())
4460				<id>       := (optional) platform device id
4461			example:
4462				virtio_mmio.device=1K@0x100b0000:48:7
4463
4464			Can be used multiple times for multiple devices.
4465
4466	vga=		[BOOT,X86-32] Select a particular video mode
4467			See Documentation/x86/boot.txt and
4468			Documentation/svga.txt.
4469			Use vga=ask for menu.
4470			This is actually a boot loader parameter; the value is
4471			passed to the kernel using a special protocol.
4472
4473	vmalloc=nn[KMG]	[KNL,BOOT] Forces the vmalloc area to have an exact
4474			size of <nn>. This can be used to increase the
4475			minimum size (128MB on x86). It can also be used to
4476			decrease the size and leave more room for directly
4477			mapped kernel RAM.
4478
4479	vmhalt=		[KNL,S390] Perform z/VM CP command after system halt.
4480			Format: <command>
4481
4482	vmpanic=	[KNL,S390] Perform z/VM CP command after kernel panic.
4483			Format: <command>
4484
4485	vmpoff=		[KNL,S390] Perform z/VM CP command after power off.
4486			Format: <command>
4487
4488	vsyscall=	[X86-64]
4489			Controls the behavior of vsyscalls (i.e. calls to
4490			fixed addresses of 0xffffffffff600x00 from legacy
4491			code).  Most statically-linked binaries and older
4492			versions of glibc use these calls.  Because these
4493			functions are at fixed addresses, they make nice
4494			targets for exploits that can control RIP.
4495
4496			emulate     [default] Vsyscalls turn into traps and are
4497			            emulated reasonably safely.
4498
4499			native      Vsyscalls are native syscall instructions.
4500			            This is a little bit faster than trapping
4501			            and makes a few dynamic recompilers work
4502			            better than they would in emulation mode.
4503			            It also makes exploits much easier to write.
4504
4505			none        Vsyscalls don't work at all.  This makes
4506			            them quite hard to use for exploits but
4507			            might break your system.
4508
4509	vt.color=	[VT] Default text color.
4510			Format: 0xYX, X = foreground, Y = background.
4511			Default: 0x07 = light gray on black.
4512
4513	vt.cur_default=	[VT] Default cursor shape.
4514			Format: 0xCCBBAA, where AA, BB, and CC are the same as
4515			the parameters of the <Esc>[?A;B;Cc escape sequence;
4516			see VGA-softcursor.txt. Default: 2 = underline.
4517
4518	vt.default_blu=	[VT]
4519			Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4520			Change the default blue palette of the console.
4521			This is a 16-member array composed of values
4522			ranging from 0-255.
4523
4524	vt.default_grn=	[VT]
4525			Format: <green0>,<green1>,<green2>,...,<green15>
4526			Change the default green palette of the console.
4527			This is a 16-member array composed of values
4528			ranging from 0-255.
4529
4530	vt.default_red=	[VT]
4531			Format: <red0>,<red1>,<red2>,...,<red15>
4532			Change the default red palette of the console.
4533			This is a 16-member array composed of values
4534			ranging from 0-255.
4535
4536	vt.default_utf8=
4537			[VT]
4538			Format=<0|1>
4539			Set system-wide default UTF-8 mode for all tty's.
4540			Default is 1, i.e. UTF-8 mode is enabled for all
4541			newly opened terminals.
4542
4543	vt.global_cursor_default=
4544			[VT]
4545			Format=<-1|0|1>
4546			Set system-wide default for whether a cursor
4547			is shown on new VTs. Default is -1,
4548			i.e. cursors will be created by default unless
4549			overridden by individual drivers. 0 will hide
4550			cursors, 1 will display them.
4551
4552	vt.italic=	[VT] Default color for italic text; 0-15.
4553			Default: 2 = green.
4554
4555	vt.underline=	[VT] Default color for underlined text; 0-15.
4556			Default: 3 = cyan.
4557
4558	watchdog timers	[HW,WDT] For information on watchdog timers,
4559			see Documentation/watchdog/watchdog-parameters.txt
4560			or other driver-specific files in the
4561			Documentation/watchdog/ directory.
4562
4563	workqueue.watchdog_thresh=
4564			If CONFIG_WQ_WATCHDOG is configured, workqueue can
4565			warn stall conditions and dump internal state to
4566			help debugging.  0 disables workqueue stall
4567			detection; otherwise, it's the stall threshold
4568			duration in seconds.  The default value is 30 and
4569			it can be updated at runtime by writing to the
4570			corresponding sysfs file.
4571
4572	workqueue.disable_numa
4573			By default, all work items queued to unbound
4574			workqueues are affine to the NUMA nodes they're
4575			issued on, which results in better behavior in
4576			general.  If NUMA affinity needs to be disabled for
4577			whatever reason, this option can be used.  Note
4578			that this also can be controlled per-workqueue for
4579			workqueues visible under /sys/bus/workqueue/.
4580
4581	workqueue.power_efficient
4582			Per-cpu workqueues are generally preferred because
4583			they show better performance thanks to cache
4584			locality; unfortunately, per-cpu workqueues tend to
4585			be more power hungry than unbound workqueues.
4586
4587			Enabling this makes the per-cpu workqueues which
4588			were observed to contribute significantly to power
4589			consumption unbound, leading to measurably lower
4590			power usage at the cost of small performance
4591			overhead.
4592
4593			The default value of this parameter is determined by
4594			the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4595
4596	workqueue.debug_force_rr_cpu
4597			Workqueue used to implicitly guarantee that work
4598			items queued without explicit CPU specified are put
4599			on the local CPU.  This guarantee is no longer true
4600			and while local CPU is still preferred work items
4601			may be put on foreign CPUs.  This debug option
4602			forces round-robin CPU selection to flush out
4603			usages which depend on the now broken guarantee.
4604			When enabled, memory and cache locality will be
4605			impacted.
4606
4607	x2apic_phys	[X86-64,APIC] Use x2apic physical mode instead of
4608			default x2apic cluster mode on platforms
4609			supporting x2apic.
4610
4611	x86_intel_mid_timer= [X86-32,APBT]
4612			Choose timer option for x86 Intel MID platform.
4613			Two valid options are apbt timer only and lapic timer
4614			plus one apbt timer for broadcast timer.
4615			x86_intel_mid_timer=apbt_only | lapic_and_apbt
4616
4617	xen_512gb_limit		[KNL,X86-64,XEN]
4618			Restricts the kernel running paravirtualized under Xen
4619			to use only up to 512 GB of RAM. The reason to do so is
4620			crash analysis tools and Xen tools for doing domain
4621			save/restore/migration must be enabled to handle larger
4622			domains.
4623
4624	xen_emul_unplug=		[HW,X86,XEN]
4625			Unplug Xen emulated devices
4626			Format: [unplug0,][unplug1]
4627			ide-disks -- unplug primary master IDE devices
4628			aux-ide-disks -- unplug non-primary-master IDE devices
4629			nics -- unplug network devices
4630			all -- unplug all emulated devices (NICs and IDE disks)
4631			unnecessary -- unplugging emulated devices is
4632				unnecessary even if the host did not respond to
4633				the unplug protocol
4634			never -- do not unplug even if version check succeeds
4635
4636	xen_nopvspin	[X86,XEN]
4637			Disables the ticketlock slowpath using Xen PV
4638			optimizations.
4639
4640	xen_nopv	[X86]
4641			Disables the PV optimizations forcing the HVM guest to
4642			run as generic HVM guest with no PV drivers.
4643
4644	xirc2ps_cs=	[NET,PCMCIA]
4645			Format:
4646			<irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
4647
4648______________________________________________________________________
4649
4650TODO:
4651
4652	Add more DRM drivers.
4653