1 acpi= [HW,ACPI,X86,ARM64]
2 Advanced Configuration and Power Interface
3 Format: { force | on | off | strict | noirq | rsdt |
5 force -- enable ACPI if default was off
6 on -- enable ACPI but allow fallback to DT [arm64]
7 off -- disable ACPI if default was on
8 noirq -- do not use ACPI for IRQ routing
9 strict -- Be less tolerant of platforms that are not
10 strictly ACPI specification compliant.
11 rsdt -- prefer RSDT over (default) XSDT
12 copy_dsdt -- copy DSDT to memory
13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force"
16 See also Documentation/power/runtime_pm.txt, pci=noacpi
18 acpi_apic_instance= [ACPI, IOAPIC]
20 2: use 2nd APIC table, if available
21 1,0: use 1st APIC table
24 acpi_backlight= [HW,ACPI]
27 If set to vendor, prefer vendor specific driver
28 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
29 of the ACPI video.ko driver.
31 acpi_force_32bit_fadt_addr
32 force FADT to use 32 bit addresses rather than the
33 64 bit X_* addresses. Some firmware have broken 64
34 bit addresses for force ACPI ignore these and use
35 the older legacy 32 bit addresses.
37 acpica_no_return_repair [HW, ACPI]
38 Disable AML predefined validation mechanism
39 This mechanism can repair the evaluation result to make
40 the return objects more ACPI specification compliant.
41 This option is useful for developers to identify the
42 root cause of an AML interpreter issue when the issue
43 has something to do with the repair mechanism.
45 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
46 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
48 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
49 debug output. Bits in debug_layer correspond to a
50 _COMPONENT in an ACPI source file, e.g.,
51 #define _COMPONENT ACPI_PCI_COMPONENT
52 Bits in debug_level correspond to a level in
53 ACPI_DEBUG_PRINT statements, e.g.,
54 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
55 The debug_level mask defaults to "info". See
56 Documentation/acpi/debug.txt for more information about
57 debug layers and levels.
59 Enable processor driver info messages:
60 acpi.debug_layer=0x20000000
61 Enable PCI/PCI interrupt routing info messages:
62 acpi.debug_layer=0x400000
63 Enable AML "Debug" output, i.e., stores to the Debug
64 object while interpreting AML:
65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
66 Enable all messages related to ACPI hardware:
67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
69 Some values produce so much output that the system is
70 unusable. The "log_buf_len" parameter may be useful
71 if you need to capture more output.
73 acpi_enforce_resources= [ACPI]
75 Check for resource conflicts between native drivers
76 and ACPI OperationRegions (SystemIO and SystemMemory
77 only). IO ports and memory declared in ACPI might be
78 used by the ACPI subsystem in arbitrary AML code and
79 can interfere with legacy drivers.
80 strict (default): access to resources claimed by ACPI
81 is denied; legacy drivers trying to access reserved
82 resources will fail to bind to device using them.
83 lax: access to resources claimed by ACPI is allowed;
84 legacy drivers trying to access reserved resources
85 will bind successfully but a warning message is logged.
86 no: ACPI OperationRegions are not marked as reserved,
87 no further checks are performed.
89 acpi_force_table_verification [HW,ACPI]
90 Enable table checksum verification during early stage.
91 By default, this is disabled due to x86 early mapping
94 acpi_irq_balance [HW,ACPI]
95 ACPI will balance active IRQs
98 acpi_irq_nobalance [HW,ACPI]
99 ACPI will not move active IRQs (default)
102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
103 Format: <irq>,<irq>...
105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
107 Format: <irq>,<irq>...
109 acpi_mask_gpe= [HW,ACPI]
110 Due to the existence of _Lxx/_Exx, some GPEs triggered
111 by unsupported hardware/firmware features can result in
112 GPE floodings that cannot be automatically disabled by
114 This facility can be used to prevent such uncontrolled
117 Support masking of GPEs numbered from 0x00 to 0x7f.
119 acpi_no_auto_serialize [HW,ACPI]
120 Disable auto-serialization of AML methods
121 AML control methods that contain the opcodes to create
122 named objects will be marked as "Serialized" by the
123 auto-serialization feature.
124 This feature is enabled by default.
125 This option allows to turn off the feature.
127 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
130 acpi_no_static_ssdt [HW,ACPI]
131 Disable installation of static SSDTs at early boot time
132 By default, SSDTs contained in the RSDT/XSDT will be
133 installed automatically and they will appear under
134 /sys/firmware/acpi/tables.
135 This option turns off this feature.
136 Note that specifying this option does not affect
137 dynamic table installation which will install SSDT
138 tables to /sys/firmware/acpi/tables/dynamic.
140 acpi_rsdp= [ACPI,EFI,KEXEC]
141 Pass the RSDP address to the kernel, mostly used
142 on machines running EFI runtime service to boot the
143 second kernel for kdump.
145 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
146 Format: To spoof as Windows 98: ="Microsoft Windows"
148 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
149 of 2 which is mandated by ACPI 6) as the supported ACPI
150 specification revision (when using this switch, it may
151 be necessary to carry out a cold reboot _twice_ in a
152 row to make it take effect on the platform firmware).
154 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
155 acpi_osi="string1" # add string1
156 acpi_osi="!string2" # remove string2
157 acpi_osi=!* # remove all strings
158 acpi_osi=! # disable all built-in OS vendor
160 acpi_osi=!! # enable all built-in OS vendor
162 acpi_osi= # disable all strings
164 'acpi_osi=!' can be used in combination with single or
165 multiple 'acpi_osi="string1"' to support specific OS
166 vendor string(s). Note that such command can only
167 affect the default state of the OS vendor strings, thus
168 it cannot affect the default state of the feature group
169 strings and the current state of the OS vendor strings,
170 specifying it multiple times through kernel command line
171 is meaningless. This command is useful when one do not
172 care about the state of the feature group strings which
173 should be controlled by the OSPM.
175 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
176 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
177 can make '_OSI("Windows 2000")' TRUE.
179 'acpi_osi=' cannot be used in combination with other
180 'acpi_osi=' command lines, the _OSI method will not
181 exist in the ACPI namespace. NOTE that such command can
182 only affect the _OSI support state, thus specifying it
183 multiple times through kernel command line is also
186 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
189 'acpi_osi=!*' can be used in combination with single or
190 multiple 'acpi_osi="string1"' to support specific
191 string(s). Note that such command can affect the
192 current state of both the OS vendor strings and the
193 feature group strings, thus specifying it multiple times
194 through kernel command line is meaningful. But it may
195 still not able to affect the final state of a string if
196 there are quirks related to this string. This command
197 is useful when one want to control the state of the
198 feature group strings to debug BIOS issues related to
201 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
202 '_OSI("Module Device")' FALSE.
203 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
204 '_OSI("Module Device")' TRUE.
205 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
207 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
209 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
210 they all will make '_OSI("Windows 2000")' TRUE.
213 Override the pmtimer bug detection: force the kernel
214 to assume that this machine's pmtimer latches its value
215 and always returns good values.
217 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
218 Format: { level | edge | high | low }
220 acpi_skip_timer_override [HW,ACPI]
221 Recognize and ignore IRQ0/pin2 Interrupt Override.
222 For broken nForce2 BIOS resulting in XT-PIC timer.
224 acpi_sleep= [HW,ACPI] Sleep options
225 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
226 old_ordering, nonvs, sci_force_enable }
227 See Documentation/power/video.txt for information on
229 s3_beep is for debugging; it makes the PC's speaker beep
230 as soon as the kernel's real-mode entry point is called.
231 s4_nohwsig prevents ACPI hardware signature from being
232 used during resume from hibernation.
233 old_ordering causes the ACPI 1.0 ordering of the _PTS
234 control method, with respect to putting devices into
235 low power states, to be enforced (the ACPI 2.0 ordering
236 of _PTS is used by default).
237 nonvs prevents the kernel from saving/restoring the
238 ACPI NVS memory during suspend/hibernation and resume.
239 sci_force_enable causes the kernel to set SCI_EN directly
240 on resume from S1/S3 (which is against the ACPI spec,
241 but some broken systems don't work without it).
243 acpi_use_timer_override [HW,ACPI]
244 Use timer override. For some broken Nvidia NF5 boards
245 that require a timer override, but don't have HPET
247 add_efi_memmap [EFI; X86] Include EFI memory map in
248 kernel's map of available physical RAM.
251 { off | try_unsupported }
252 off: disable AGP support
253 try_unsupported: try to drive unsupported chipsets
254 (may crash computer or cause data corruption)
257 See Documentation/sound/alsa/alsa-parameters.txt
260 Allow the default userspace alignment fault handler
261 behaviour to be specified. Bit 0 enables warnings,
262 bit 1 enables fixups, and bit 2 sends a segfault.
264 align_va_addr= [X86-64]
265 Align virtual addresses by clearing slice [14:12] when
266 allocating a VMA at process creation time. This option
267 gives you up to 3% performance improvement on AMD F15h
268 machines (where it is enabled by default) for a
269 CPU-intensive style benchmark, and it can vary highly in
270 a microbenchmark depending on workload and compiler.
272 32: only for 32-bit processes
273 64: only for 64-bit processes
274 on: enable for both 32- and 64-bit processes
275 off: disable for both 32- and 64-bit processes
277 alloc_snapshot [FTRACE]
278 Allocate the ftrace snapshot buffer on boot up when the
279 main buffer is allocated. This is handy if debugging
280 and you need to use tracing_snapshot() on boot up, and
281 do not want to use tracing_snapshot_alloc() as it needs
282 to be done where GFP_KERNEL allocations are allowed.
284 amd_iommu= [HW,X86-64]
285 Pass parameters to the AMD IOMMU driver in the system.
287 fullflush - enable flushing of IO/TLB entries when
288 they are unmapped. Otherwise they are
289 flushed before they will be reused, which
291 off - do not initialize any AMD IOMMU found in
293 force_isolation - Force device isolation for all
294 devices. The IOMMU driver is not
295 allowed anymore to lift isolation
296 requirements as needed. This option
297 does not override iommu=pt
299 amd_iommu_dump= [HW,X86-64]
300 Enable AMD IOMMU driver option to dump the ACPI table
301 for AMD IOMMU. With this option enabled, AMD IOMMU
302 driver will print ACPI tables for AMD IOMMU during
303 IOMMU initialization.
305 amd_iommu_intr= [HW,X86-64]
306 Specifies one of the following AMD IOMMU interrupt
308 legacy - Use legacy interrupt remapping mode.
309 vapic - Use virtual APIC mode, which allows IOMMU
310 to inject interrupts directly into guest.
311 This mode requires kvm-amd.avic=1.
312 (Default when IOMMU HW support is present.)
314 amijoy.map= [HW,JOY] Amiga joystick support
315 Map of devices attached to JOY0DAT and JOY1DAT
317 See also Documentation/input/joystick.txt
319 analog.map= [HW,JOY] Analog joystick and gamepad support
320 Specifies type or capabilities of an analog joystick
321 connected to one of 16 gameports
322 Format: <type1>,<type2>,..<type16>
325 Power management functions (SPARCstation-4/5 + deriv.)
327 Disable APC CPU standby support. SPARCstation-Fox does
328 not play well with APC CPU idle - disable it if you have
329 APC and your system crashes randomly.
331 apic= [APIC,X86-32] Advanced Programmable Interrupt Controller
332 Change the output verbosity whilst booting
333 Format: { quiet (default) | verbose | debug }
334 Change the amount of debugging information output
335 when initialising the APIC and IO-APIC components.
337 apic_extnmi= [APIC,X86] External NMI delivery setting
338 Format: { bsp (default) | all | none }
339 bsp: External NMI is delivered only to CPU 0
340 all: External NMIs are broadcast to all CPUs as a
342 none: External NMI is masked for all CPUs. This is
343 useful so that a dump capture kernel won't be
347 See Documentation/networking/ipv6.txt.
349 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
350 Limit apic dumping. The parameter defines the maximal
351 number of local apics being dumped. Also it is possible
352 to set it to "all" by meaning -- no limit here.
353 Format: { 1 (default) | 2 | ... | all }.
354 The parameter valid if only apic=debug or
355 apic=verbose is specified.
356 Example: apic=debug show_lapic=all
358 apm= [APM] Advanced Power Management
359 See header of arch/x86/kernel/apm_32.c.
361 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
362 Format: <io>,<irq>,<nodeID>
366 atarimouse= [HW,MOUSE] Atari Mouse
368 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
369 EzKey and similar keyboards
371 atkbd.reset= [HW] Reset keyboard during initialization
373 atkbd.set= [HW] Select keyboard code set
374 Format: <int> (2 = AT (default), 3 = PS/2)
376 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
379 atkbd.softraw= [HW] Choose between synthetic and real raw mode
380 Format: <bool> (0 = real, 1 = synthetic (default))
382 atkbd.softrepeat= [HW]
383 Use software keyboard repeat
385 audit= [KNL] Enable the audit sub-system
386 Format: { "0" | "1" } (0 = disabled, 1 = enabled)
387 0 - kernel audit is disabled and can not be enabled
388 until the next reboot
389 unset - kernel audit is initialized but disabled and
390 will be fully enabled by the userspace auditd.
391 1 - kernel audit is initialized and partially enabled,
392 storing at most audit_backlog_limit messages in
393 RAM until it is fully enabled by the userspace
397 audit_backlog_limit= [KNL] Set the audit queue size limit.
398 Format: <int> (must be >=0)
401 bau= [X86_UV] Enable the BAU on SGI UV. The default
402 behavior is to disable the BAU (i.e. bau=0).
403 Format: { "0" | "1" }
406 unset - Disable the BAU.
408 baycom_epp= [HW,AX25]
411 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
413 See header of drivers/net/hamradio/baycom_par.c.
415 baycom_ser_fdx= [HW,AX25]
416 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
417 Format: <io>,<irq>,<mode>[,<baud>]
418 See header of drivers/net/hamradio/baycom_ser_fdx.c.
420 baycom_ser_hdx= [HW,AX25]
421 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
422 Format: <io>,<irq>,<mode>
423 See header of drivers/net/hamradio/baycom_ser_hdx.c.
425 blkdevparts= Manual partition parsing of block device(s) for
426 embedded devices based on command line input.
427 See Documentation/block/cmdline-partition.txt
429 boot_delay= Milliseconds to delay each printk during boot.
430 Values larger than 10 seconds (10000) are changed to
434 bootmem_debug [KNL] Enable bootmem allocator debug messages.
437 Disable BERT OS support on buggy BIOSes.
439 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
440 bttv.radio= Most important insmod options are available as
442 bttv.pll= See Documentation/video4linux/bttv/Insmod-options
445 bulk_remove=off [PPC] This parameter disables the use of the pSeries
446 firmware feature for flushing multiple hpte entries
449 c101= [NET] Moxa C101 synchronous serial card
451 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
452 Sometimes CPU hardware bugs make them report the cache
453 size incorrectly. The kernel will attempt work arounds
454 to fix known problems, but for some CPUs it is not
455 possible to determine what the correct size should be.
456 This option provides an override for these situations.
458 ca_keys= [KEYS] This parameter identifies a specific key(s) on
459 the system trusted keyring to be used for certificate
461 format: { id:<keyid> | builtin }
463 cca= [MIPS] Override the kernel pages' cache coherency
464 algorithm. Accepted values range from 0 to 7
465 inclusive. See arch/mips/include/asm/pgtable-bits.h
466 for platform specific values (SB1, Loongson3 and
469 ccw_timeout_log [S390]
470 See Documentation/s390/CommonIO for details.
472 cgroup_disable= [KNL] Disable a particular controller
473 Format: {name of the controller(s) to disable}
474 The effects of cgroup_disable=foo are:
475 - foo isn't auto-mounted if you mount all cgroups in
477 - foo isn't visible as an individually mountable
479 {Currently only "memory" controller deal with this and
480 cut the overhead, others just disable the usage. So
481 only cgroup_disable=memory is actually worthy}
483 cgroup_no_v1= [KNL] Disable one, multiple, all cgroup controllers in v1
484 Format: { controller[,controller...] | "all" }
485 Like cgroup_disable, but only applies to cgroup v1;
486 the blacklisted controllers remain available in cgroup2.
488 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
490 nosocket -- Disable socket memory accounting.
491 nokmem -- Disable kernel memory accounting.
493 checkreqprot [SELINUX] Set initial checkreqprot flag value.
494 Format: { "0" | "1" }
495 See security/selinux/Kconfig help text.
496 0 -- check protection applied by kernel (includes
497 any implied execute protection).
498 1 -- check protection requested by application.
499 Default value is set via a kernel config option.
500 Value can be changed at runtime via
501 /selinux/checkreqprot.
504 See Documentation/s390/CommonIO for details.
507 Prevents the clock framework from automatically gating
508 clocks that have not been explicitly enabled by a Linux
509 device driver but are enabled in hardware at reset or
510 by the bootloader/firmware. Note that this does not
511 force such clocks to be always-on nor does it reserve
512 those clocks in any way. This parameter is useful for
513 debug and development, but should not be needed on a
514 platform with proper driver support. For more
515 information, see Documentation/clk.txt.
517 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
519 Forces specified clocksource (if available) to be used
520 when calculating gettimeofday(). If specified
521 clocksource is not available, it defaults to PIT.
522 Format: { pit | tsc | cyclone | pmtmr }
524 clocksource= Override the default clocksource
526 Override the default clocksource and use the clocksource
527 with the name specified.
528 Some clocksource names to choose from, depending on
530 [all] jiffies (this is the base, fallback clocksource)
532 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
533 pxa_timer,timer3,32k_counter,timer0_1
534 [X86-32] pit,hpet,tsc;
535 scx200_hrt on Geode; cyclone on IBM x440
543 clocksource.arm_arch_timer.evtstrm=
546 Enable/disable the eventstream feature of the ARM
547 architected timer so that code using WFE-based polling
548 loops can be debugged more effectively on production
551 clearcpuid=BITNUM [X86]
552 Disable CPUID feature X for the kernel. See
553 arch/x86/include/asm/cpufeatures.h for the valid bit
554 numbers. Note the Linux specific bits are not necessarily
555 stable over kernel options, but the vendor specific
557 Also note that user programs calling CPUID directly
558 or using the feature without checking anything
559 will still see it. This just prevents it from
560 being used by the kernel or shown in /proc/cpuinfo.
561 Also note the kernel might malfunction if you disable
564 cma=nn[MG]@[start[MG][-end[MG]]]
566 Sets the size of kernel global memory area for
567 contiguous memory allocations and optionally the
568 placement constraint by the physical address range of
569 memory allocations. A value of 0 disables CMA
570 altogether. For more information, see
571 include/linux/dma-contiguous.h
573 cmo_free_hint= [PPC] Format: { yes | no }
574 Specify whether pages are marked as being inactive
575 when they are freed. This is used in CMO environments
576 to determine OS memory pressure for page stealing by
580 coherent_pool=nn[KMG] [ARM,KNL]
581 Sets the size of memory pool for coherent, atomic dma
582 allocations, by default set to 256K.
584 code_bytes [X86] How many bytes of object code to print
589 com20020= [HW,NET] ARCnet - COM20020 chipset
591 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
593 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
597 ARCnet - COM90xx chipset (memory-mapped buffers)
598 Format: <io>[,<irq>[,<memstart>]]
600 condev= [HW,S390] console device
603 console= [KNL] Output console device and options.
605 tty<n> Use the virtual console device <n>.
609 Use the specified serial port. The options are of
610 the form "bbbbpnf", where "bbbb" is the baud rate,
611 "p" is parity ("n", "o", or "e"), "n" is number of
612 bits, and "f" is flow control ("r" for RTS or
613 omit it). Default is "9600n8".
615 See Documentation/admin-guide/serial-console.rst for more
617 Documentation/networking/netconsole.txt for an
620 uart[8250],io,<addr>[,options]
621 uart[8250],mmio,<addr>[,options]
622 uart[8250],mmio16,<addr>[,options]
623 uart[8250],mmio32,<addr>[,options]
624 uart[8250],0x<addr>[,options]
625 Start an early, polled-mode console on the 8250/16550
626 UART at the specified I/O port or MMIO address,
627 switching to the matching ttyS device later.
628 MMIO inter-register address stride is either 8-bit
629 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
630 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
631 to be equivalent to 'mmio'. 'options' are specified in
632 the same format described for ttyS above; if unspecified,
633 the h/w is not re-initialized.
635 hvc<n> Use the hypervisor console device <n>. This is for
636 both Xen and PowerPC hypervisors.
638 If the device connected to the port is not a TTY but a braille
639 device, prepend "brl," before the device type, for instance
641 For now, only VisioBraille is supported.
643 consoleblank= [KNL] The console blank (screen saver) timeout in
644 seconds. Defaults to 10*60 = 10mins. A value of 0
645 disables the blank timer.
648 [KNL] Change the default value for
649 /proc/<pid>/coredump_filter.
650 See also Documentation/filesystems/proc.txt.
652 coresight_cpu_debug.enable
655 Enable/disable the CPU sampling based debugging.
656 0: default value, disable debugging
657 1: enable debugging at boot time
659 cpuidle.off=1 [CPU_IDLE]
660 disable the cpuidle sub-system
662 cpufreq.off=1 [CPU_FREQ]
663 disable the cpufreq sub-system
666 [X86] Delay for N microsec between assert and de-assert
667 of APIC INIT to start processors. This delay occurs
668 on every CPU online, such as boot, and resume from suspend.
671 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
673 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
675 crashkernel=size[KMG][@offset[KMG]]
676 [KNL] Using kexec, Linux can switch to a 'crash kernel'
677 upon panic. This parameter reserves the physical
678 memory region [offset, offset + size] for that kernel
679 image. If '@offset' is omitted, then a suitable offset
680 is selected automatically. Check
681 Documentation/kdump/kdump.txt for further details.
683 crashkernel=range1:size1[,range2:size2,...][@offset]
684 [KNL] Same as above, but depends on the memory
685 in the running system. The syntax of range is
686 start-[end] where start and end are both
687 a memory unit (amount[KMG]). See also
688 Documentation/kdump/kdump.txt for an example.
690 crashkernel=size[KMG],high
691 [KNL, x86_64] range could be above 4G. Allow kernel
692 to allocate physical memory region from top, so could
693 be above 4G if system have more than 4G ram installed.
694 Otherwise memory region will be allocated below 4G, if
696 It will be ignored if crashkernel=X is specified.
697 crashkernel=size[KMG],low
698 [KNL, x86_64] range under 4G. When crashkernel=X,high
699 is passed, kernel could allocate physical memory region
700 above 4G, that cause second kernel crash on system
701 that require some amount of low memory, e.g. swiotlb
702 requires at least 64M+32K low memory, also enough extra
703 low memory is needed to make sure DMA buffers for 32-bit
704 devices won't run out. Kernel would try to allocate at
705 at least 256M below 4G automatically.
706 This one let user to specify own low range under 4G
707 for second kernel instead.
708 0: to disable low allocation.
709 It will be ignored when crashkernel=X,high is not used
710 or memory reserved is below 4G.
713 [KNL] Disable crypto self-tests
718 cs89x0_media= [HW,NET]
719 Format: { rj45 | aui | bnc }
722 See header of drivers/s390/block/dasd_devmap.c.
724 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
725 (one device per port)
726 Format: <port#>,<type>
727 See also Documentation/input/joystick-parport.txt
729 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
731 Documentation/admin-guide/dynamic-debug-howto.rst for
732 details. Deprecated, see dyndbg.
734 debug [KNL] Enable kernel debugging (events log level).
737 [KNL] verbose self-tests
739 Print debugging info while doing the locking API
741 We default to 0 (no extra messages), setting it to
742 1 will print _a lot_ more information - normally
743 only useful to kernel developers.
745 debug_objects [KNL] Enable object debugging
748 [KNL] Disable object debugging
750 debug_guardpage_minorder=
751 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
752 parameter allows control of the order of pages that will
753 be intentionally kept free (and hence protected) by the
754 buddy allocator. Bigger value increase the probability
755 of catching random memory corruption, but reduce the
756 amount of memory for normal system use. The maximum
757 possible value is MAX_ORDER/2. Setting this parameter
758 to 1 or 2 should be enough to identify most random
759 memory corruption problems caused by bugs in kernel or
760 driver code when a CPU writes to (or reads from) a
761 random memory location. Note that there exists a class
762 of memory corruptions problems caused by buggy H/W or
763 F/W or by drivers badly programing DMA (basically when
764 memory is written at bus level and the CPU MMU is
765 bypassed) which are not detectable by
766 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
767 tracking down these problems.
770 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
771 parameter enables the feature at boot time. In
772 default, it is disabled. We can avoid allocating huge
773 chunk of memory for debug pagealloc if we don't enable
774 it at boot time and the system will work mostly same
775 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
776 on: enable the feature
778 debugpat [X86] Enable PAT debugging
780 decnet.addr= [HW,NET]
781 Format: <area>[,<node>]
782 See also Documentation/networking/decnet.txt.
785 [same as hugepagesz=] The size of the default
786 HugeTLB page size. This is the size represented by
787 the legacy /proc/ hugepages APIs, used for SHM, and
788 default size when mounting hugetlbfs filesystems.
789 Defaults to the default architecture's huge page size
793 Set number of hash buckets for dentry cache.
795 disable_1tb_segments [PPC]
796 Disables the use of 1TB hash page table segments. This
797 causes the kernel to fall back to 256MB segments which
798 can be useful when debugging issues that require an SLB
802 See Documentation/networking/ipv6.txt.
805 Disable RADIX MMU mode on POWER9
807 disable_cpu_apicid= [X86,APIC,SMP]
809 The number of initial APIC ID for the
810 corresponding CPU to be disabled at boot,
811 mostly used for the kdump 2nd kernel to
812 disable BSP to wake up multiple CPUs without
813 causing system reset or hang due to sending
816 disable_ddw [PPC/PSERIES]
817 Disable Dynamic DMA Window support. Use this if
818 to workaround buggy firmware.
821 See Documentation/networking/ipv6.txt.
823 disable_mtrr_cleanup [X86]
824 The kernel tries to adjust MTRR layout from continuous
825 to discrete, to make X server driver able to add WB
826 entry later. This parameter disables that.
828 disable_mtrr_trim [X86, Intel and AMD only]
829 By default the kernel will trim any uncacheable
830 memory out of your available memory pool based on
831 MTRR settings. This parameter disables that behavior,
832 possibly causing your machine to run very slowly.
834 disable_timer_pin_1 [X86]
835 Disable PIN 1 of APIC timer
836 Can be useful to work around chipset bugs.
838 dis_ucode_ldr [X86] Disable the microcode loader.
840 dm= [DM] Allows early creation of a device-mapper device.
841 See Documentation/device-mapper/boot.txt.
843 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
844 this option disables the debugging code at boot.
846 dma_debug_entries=<number>
847 This option allows to tune the number of preallocated
848 entries for DMA-API debugging code. One entry is
849 required per DMA-API allocation. Use this if the
850 DMA-API debugging code disables itself because the
851 architectural default is too low.
853 dma_debug_driver=<driver_name>
854 With this option the DMA-API debugging driver
855 filter feature can be enabled at boot time. Just
856 pass the driver to filter for as the parameter.
857 The filter can be disabled or changed to another
858 driver later using sysfs.
860 drm_kms_helper.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
861 Broken monitors, graphic adapters, KVMs and EDIDless
862 panels may send no or incorrect EDID data sets.
863 This parameter allows to specify an EDID data sets
864 in the /lib/firmware directory that are used instead.
865 Generic built-in EDID data sets are used, if one of
866 edid/1024x768.bin, edid/1280x1024.bin,
867 edid/1680x1050.bin, or edid/1920x1080.bin is given
868 and no file with the same name exists. Details and
869 instructions how to build your own EDID data are
870 available in Documentation/EDID/HOWTO.txt. An EDID
871 data set will only be used for a particular connector,
872 if its name and a colon are prepended to the EDID
873 name. Each connector may use a unique EDID data
874 set by separating the files with a comma. An EDID
875 data set with no connector name will be used for
876 any connectors not explicitly specified.
881 Format: {"off" | "known"}
882 Control how the dt_cpu_ftrs device-tree binding is
883 used for CPU feature discovery and setup (if it
885 off: Do not use it, fall back to legacy cpu table.
886 known: Do not pass through unknown features to guests
887 or userspace, only those that the kernel is aware of.
889 dump_apple_properties [X86]
890 Dump name and content of EFI device properties on
891 x86 Macs. Useful for driver authors to determine
892 what data is available or for reverse-engineering.
894 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
895 module.dyndbg[="val"]
896 Enable debug messages at boot time. See
897 Documentation/admin-guide/dynamic-debug-howto.rst
900 nompx [X86] Disables Intel Memory Protection Extensions.
901 See Documentation/x86/intel_mpx.txt for more
902 information about the feature.
904 nopku [X86] Disable Memory Protection Keys CPU feature found
907 module.async_probe [KNL]
908 Enable asynchronous probe on this module.
910 early_ioremap_debug [KNL]
911 Enable debug messages in early_ioremap support. This
912 is useful for tracking down temporary early mappings
913 which are not unmapped.
915 earlycon= [KNL] Output early console device and options.
917 When used with no options, the early console is
918 determined by the stdout-path property in device
921 cdns,<addr>[,options]
922 Start an early, polled-mode console on a Cadence
923 (xuartps) serial port at the specified address. Only
924 supported option is baud rate. If baud rate is not
925 specified, the serial port must already be setup and
928 uart[8250],io,<addr>[,options]
929 uart[8250],mmio,<addr>[,options]
930 uart[8250],mmio32,<addr>[,options]
931 uart[8250],mmio32be,<addr>[,options]
932 uart[8250],0x<addr>[,options]
933 Start an early, polled-mode console on the 8250/16550
934 UART at the specified I/O port or MMIO address.
935 MMIO inter-register address stride is either 8-bit
936 (mmio) or 32-bit (mmio32 or mmio32be).
937 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
938 to be equivalent to 'mmio'. 'options' are specified
939 in the same format described for "console=ttyS<n>"; if
940 unspecified, the h/w is not initialized.
944 Start an early, polled-mode console on a pl011 serial
945 port at the specified address. The pl011 serial port
946 must already be setup and configured. Options are not
947 yet supported. If 'mmio32' is specified, then only
948 the driver will use only 32-bit accessors to read/write
949 the device registers.
952 Start an early, polled-mode console on a meson serial
953 port at the specified address. The serial port must
954 already be setup and configured. Options are not yet
958 Start an early, polled-mode console on an msm serial
959 port at the specified address. The serial port
960 must already be setup and configured. Options are not
964 Start an early, polled-mode console on an msm serial
965 dm port at the specified address. The serial port
966 must already be setup and configured. Options are not
970 Start an early, polled-mode console on a serial port
971 of an Actions Semi SoC, such as S500 or S900, at the
972 specified address. The serial port must already be
973 setup and configured. Options are not yet supported.
975 smh Use ARM semihosting calls for early console.
983 Use early console provided by serial driver available
984 on Samsung SoCs, requires selecting proper type and
985 a correct base address of the selected UART port. The
986 serial port must already be setup and configured.
987 Options are not yet supported.
990 Start an early, polled-mode console on a lantiq serial
991 (lqasc) port at the specified address. The serial port
992 must already be setup and configured. Options are not
997 Use early console provided by Freescale LP UART driver
998 found on Freescale Vybrid and QorIQ LS1021A processors.
999 A valid base address must be provided, and the serial
1000 port must already be setup and configured.
1003 Start an early, polled-mode console on the
1004 Armada 3700 serial port at the specified
1005 address. The serial port must already be setup
1006 and configured. Options are not yet supported.
1008 earlyprintk= [X86,SH,BLACKFIN,ARM,M68k,S390]
1013 earlyprintk=serial[,ttySn[,baudrate]]
1014 earlyprintk=serial[,0x...[,baudrate]]
1015 earlyprintk=ttySn[,baudrate]
1016 earlyprintk=dbgp[debugController#]
1017 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1018 earlyprintk=xdbc[xhciController#]
1020 earlyprintk is useful when the kernel crashes before
1021 the normal console is initialized. It is not enabled by
1022 default because it has some cosmetic problems.
1024 Append ",keep" to not disable it when the real console
1027 Only one of vga, efi, serial, or usb debug port can
1030 Currently only ttyS0 and ttyS1 may be specified by
1031 name. Other I/O ports may be explicitly specified
1032 on some architectures (x86 and arm at least) by
1033 replacing ttySn with an I/O port address, like this:
1034 earlyprintk=serial,0x1008,115200
1035 You can find the port for a given device in
1036 /proc/tty/driver/serial:
1037 2: uart:ST16650V2 port:00001008 irq:18 ...
1039 Interaction with the standard serial driver is not
1042 The VGA and EFI output is eventually overwritten by
1045 The xen output can only be used by Xen PV guests.
1047 The sclp output can only be used on s390.
1049 The optional "force" to "pciserial" enables use of a
1050 PCI device even when its classcode is not of the
1053 edac_report= [HW,EDAC] Control how to report EDAC event
1054 Format: {"on" | "off" | "force"}
1055 on: enable EDAC to report H/W event. May be overridden
1056 by other higher priority error reporting module.
1057 off: disable H/W event reporting through EDAC.
1058 force: enforce the use of EDAC to report H/W event.
1061 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1064 This is designed to be used in conjunction with
1065 the boot argument: earlyprintk=vga
1068 Format: {"off" | "on" | "skip[mbr]"}
1071 Format: { "old_map", "nochunk", "noruntime", "debug" }
1072 old_map [X86-64]: switch to the old ioremap-based EFI
1073 runtime services mapping. 32-bit still uses this one by
1075 nochunk: disable reading files in "chunks" in the EFI
1076 boot stub, as chunking can cause problems with some
1077 firmware implementations.
1078 noruntime : disable EFI runtime services support
1079 debug: enable misc debug output
1081 efi_no_storage_paranoia [EFI; X86]
1082 Using this parameter you can use more than 50% of
1083 your efi variable storage. Use this parameter only if
1084 you are really sure that your UEFI does sane gc and
1085 fulfills the spec otherwise your board may brick.
1087 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1088 Add arbitrary attribute to specific memory range by
1089 updating original EFI memory map.
1090 Region of memory which aa attribute is added to is
1092 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1093 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1094 attribute is added to range 0x100000000-0x180000000 and
1095 0x10a0000000-0x1120000000.
1097 Using this parameter you can do debugging of EFI memmap
1098 related feature. For example, you can do debugging of
1099 Address Range Mirroring feature even if your box
1102 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1103 that is to be dynamically loaded by Linux. If there are
1104 multiple variables with the same name but with different
1105 vendor GUIDs, all of them will be loaded. See
1106 Documentation/acpi/ssdt-overlays.txt for details.
1109 eisa_irq_edge= [PARISC,HW]
1110 See header of drivers/parisc/eisa.c.
1113 See comment before function elanfreq_setup() in
1114 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1117 Format: {"cfq" | "deadline" | "noop"}
1118 See Documentation/block/cfq-iosched.txt and
1119 Documentation/block/deadline-iosched.txt for details.
1121 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1122 Specifies physical address of start of kernel core
1123 image elf header and optionally the size. Generally
1124 kexec loader will pass this option to capture kernel.
1125 See Documentation/kdump/kdump.txt for details.
1127 enable_mtrr_cleanup [X86]
1128 The kernel tries to adjust MTRR layout from continuous
1129 to discrete, to make X server driver able to add WB
1130 entry later. This parameter enables that.
1132 enable_timer_pin_1 [X86]
1133 Enable PIN 1 of APIC timer
1134 Can be useful to work around chipset bugs
1135 (in particular on some ATI chipsets).
1136 The kernel tries to set a reasonable default.
1138 enforcing [SELINUX] Set initial enforcing status.
1140 See security/selinux/Kconfig help text.
1141 0 -- permissive (log only, no denials).
1142 1 -- enforcing (deny and log).
1144 Value can be changed at runtime via /selinux/enforce.
1147 Disable Error Record Serialization Table (ERST)
1150 ether= [HW,NET] Ethernet cards parameters
1151 This option is obsoleted by the "netdev=" option, which
1152 has equivalent usage. See its documentation for details.
1156 Permit 'security.evm' to be updated regardless of
1157 current integrity status.
1161 fail_make_request=[KNL]
1162 General fault injection mechanism.
1163 Format: <interval>,<probability>,<space>,<times>
1164 See also Documentation/fault-injection/.
1167 See Documentation/blockdev/floppy.txt.
1169 force_pal_cache_flush
1170 [IA-64] Avoid check_sal_cache_flush which may hang on
1171 buggy SAL_CACHE_FLUSH implementations. Using this
1172 parameter will force ia64_sal_cache_flush to call
1173 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1176 Forcefully enable Physical Address Extension (PAE).
1177 Many Pentium M systems disable PAE but may have a
1178 functionally usable PAE implementation.
1179 Warning: use of this parameter will taint the kernel
1180 and may cause unknown problems.
1183 [FTRACE] will set and start the specified tracer
1184 as early as possible in order to facilitate early
1187 ftrace_dump_on_oops[=orig_cpu]
1188 [FTRACE] will dump the trace buffers on oops.
1189 If no parameter is passed, ftrace will dump
1190 buffers of all CPUs, but if you pass orig_cpu, it will
1191 dump only the buffer of the CPU that triggered the
1194 ftrace_filter=[function-list]
1195 [FTRACE] Limit the functions traced by the function
1196 tracer at boot up. function-list is a comma separated
1197 list of functions. This list can be changed at run
1198 time by the set_ftrace_filter file in the debugfs
1201 ftrace_notrace=[function-list]
1202 [FTRACE] Do not trace the functions specified in
1203 function-list. This list can be changed at run time
1204 by the set_ftrace_notrace file in the debugfs
1207 ftrace_graph_filter=[function-list]
1208 [FTRACE] Limit the top level callers functions traced
1209 by the function graph tracer at boot up.
1210 function-list is a comma separated list of functions
1211 that can be changed at run time by the
1212 set_graph_function file in the debugfs tracing directory.
1214 ftrace_graph_notrace=[function-list]
1215 [FTRACE] Do not trace from the functions specified in
1216 function-list. This list is a comma separated list of
1217 functions that can be changed at run time by the
1218 set_graph_notrace file in the debugfs tracing directory.
1220 ftrace_graph_max_depth=<uint>
1221 [FTRACE] Used with the function graph tracer. This is
1222 the max depth it will trace into a function. This value
1223 can be changed at run time by the max_graph_depth file
1224 in the tracefs tracing directory. default: 0 (no limit)
1227 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1228 support via parallel port (up to 5 devices per port)
1229 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1230 See also Documentation/input/joystick-parport.txt
1234 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1238 gcov_persist= [GCOV] When non-zero (default), profiling data for
1239 kernel modules is saved and remains accessible via
1240 debugfs, even when the module is unloaded/reloaded.
1241 When zero, profiling data is discarded and associated
1242 debugfs files are removed at module unload time.
1244 goldfish [X86] Enable the goldfish android emulator platform.
1245 Don't use this when you are not running on the
1248 gpt [EFI] Forces disk with valid GPT signature but
1249 invalid Protective MBR to be treated as GPT. If the
1250 primary GPT is corrupted, it enables the backup/alternate
1251 GPT to be used instead.
1253 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1254 the "Enable 0" bit of the configuration register.
1257 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1258 the "Enable 0" bit of the configuration register.
1261 grcan.select= [HW] Select which physical interface to use.
1264 grcan.txsize= [HW] Sets the size of the tx buffer.
1265 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1267 grcan.rxsize= [HW] Sets the size of the rx buffer.
1268 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1271 gpio-mockup.gpio_mockup_ranges
1272 [HW] Sets the ranges of gpiochip of for this device.
1273 Format: <start1>,<end1>,<start2>,<end2>...
1275 hardlockup_all_cpu_backtrace=
1276 [KNL] Should the hard-lockup detector generate
1277 backtraces on all cpus.
1280 hashdist= [KNL,NUMA] Large hashes allocated during boot
1281 are distributed across NUMA nodes. Defaults on
1282 for 64-bit NUMA, off otherwise.
1283 Format: 0 | 1 (for off | on)
1285 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1287 hd= [EIDE] (E)IDE hard drive subsystem geometry
1288 Format: <cyl>,<head>,<sect>
1291 Disable Hardware Error Source Table (HEST) support;
1292 corresponding firmware-first mode error processing
1293 logic will be disabled.
1295 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1296 size of <nn>. This works even on boxes that have no
1297 highmem otherwise. This also works to reduce highmem
1298 size on bigger boxes.
1300 highres= [KNL] Enable/disable high resolution timer mode.
1301 Valid parameters: "on", "off"
1305 See Documentation/isdn/README.HiSax.
1309 hpet= [X86-32,HPET] option to control HPET usage
1310 Format: { enable (default) | disable | force |
1312 disable: disable HPET and use PIT instead
1313 force: allow force enabled of undocumented chips (ICH4,
1315 verbose: show contents of HPET registers during setup
1317 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1318 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1320 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1321 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1322 On x86-64 and powerpc, this option can be specified
1323 multiple times interleaved with hugepages= to reserve
1324 huge pages of different sizes. Valid pages sizes on
1325 x86-64 are 2M (when the CPU supports "pse") and 1G
1326 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1328 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1329 terminal devices. Valid values: 0..8
1330 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1331 If specified, z/VM IUCV HVC accepts connections
1332 from listed z/VM user IDs only.
1334 hwthread_map= [METAG] Comma-separated list of Linux cpu id to
1335 hardware thread id mappings.
1336 Format: <cpu>:<hwthread>
1339 Do not unregister boot console at start. This is only
1340 useful for debugging when something happens in the window
1341 between unregistering the boot console and initializing
1344 i2c_bus= [HW] Override the default board specific I2C bus speed
1345 or register an additional I2C bus that is not
1346 registered from board initialization code.
1350 i8042.debug [HW] Toggle i8042 debug mode
1351 i8042.unmask_kbd_data
1352 [HW] Enable printing of interrupt data from the KBD port
1353 (disabled by default, and as a pre-condition
1354 requires that i8042.debug=1 be enabled)
1355 i8042.direct [HW] Put keyboard port into non-translated mode
1356 i8042.dumbkbd [HW] Pretend that controller can only read data from
1357 keyboard and cannot control its state
1358 (Don't attempt to blink the leds)
1359 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1360 i8042.nokbd [HW] Don't check/create keyboard port
1361 i8042.noloop [HW] Disable the AUX Loopback command while probing
1363 i8042.nomux [HW] Don't check presence of an active multiplexing
1365 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1367 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1368 i8042.reset [HW] Reset the controller during init, cleanup and
1369 suspend-to-ram transitions, only during s2r
1370 transitions, or never reset
1371 Format: { 1 | Y | y | 0 | N | n }
1372 1, Y, y: always reset controller
1373 0, N, n: don't ever reset controller
1374 Default: only on s2r transitions on x86; most other
1375 architectures force reset to be always executed
1376 i8042.unlock [HW] Unlock (ignore) the keylock
1377 i8042.kbdreset [HW] Reset device connected to KBD port
1381 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1382 indicates that the driver is running on unsupported
1384 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1385 does not match list of supported models.
1387 [HW] Report power status in /proc/i8k
1388 (disabled by default)
1389 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1392 i915.invert_brightness=
1393 [DRM] Invert the sense of the variable that is used to
1394 set the brightness of the panel backlight. Normally a
1395 brightness value of 0 indicates backlight switched off,
1396 and the maximum of the brightness value sets the backlight
1397 to maximum brightness. If this parameter is set to 0
1398 (default) and the machine requires it, or this parameter
1399 is set to 1, a brightness value of 0 sets the backlight
1400 to maximum brightness, and the maximum of the brightness
1401 value switches the backlight off.
1402 -1 -- never invert brightness
1403 0 -- machine default
1404 1 -- force brightness inversion
1407 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1409 ide-core.nodma= [HW] (E)IDE subsystem
1410 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1411 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1412 .cdrom .chs .ignore_cable are additional options
1413 See Documentation/ide/ide.txt.
1415 ide-generic.probe-mask= [HW] (E)IDE subsystem
1417 Probe mask for legacy ISA IDE ports. Depending on
1418 platform up to 6 ports are supported, enabled by
1419 setting corresponding bits in the mask to 1. The
1420 default value is 0x0, which has a special meaning.
1421 On systems that have PCI, it triggers scanning the
1422 PCI bus for the first and the second port, which
1423 are then probed. On systems without PCI the value
1424 of 0x0 enables probing the two first ports as if it
1427 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1428 Claim all unknown PCI IDE storage controllers.
1431 Format: idle=poll, idle=halt, idle=nomwait
1432 Poll forces a polling idle loop that can slightly
1433 improve the performance of waking up a idle CPU, but
1434 will use a lot of power and make the system run hot.
1436 idle=halt: Halt is forced to be used for CPU idle.
1437 In such case C2/C3 won't be used again.
1438 idle=nomwait: Disable mwait for CPU C-states
1440 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1441 Format: { strict | legacy | 2008 | relaxed }
1444 Choose which programs will be accepted for execution
1445 based on the IEEE 754 NaN encoding(s) supported by
1446 the FPU and the NaN encoding requested with the value
1447 of an ELF file header flag individually set by each
1448 binary. Hardware implementations are permitted to
1449 support either or both of the legacy and the 2008 NaN
1452 Available settings are as follows:
1453 strict accept binaries that request a NaN encoding
1454 supported by the FPU
1455 legacy only accept legacy-NaN binaries, if supported
1457 2008 only accept 2008-NaN binaries, if supported
1459 relaxed accept any binaries regardless of whether
1460 supported by the FPU
1462 The FPU emulator is always able to support both NaN
1463 encodings, so if no FPU hardware is present or it has
1464 been disabled with 'nofpu', then the settings of
1465 'legacy' and '2008' strap the emulator accordingly,
1466 'relaxed' straps the emulator for both legacy-NaN and
1467 2008-NaN, whereas 'strict' enables legacy-NaN only on
1468 legacy processors and both NaN encodings on MIPS32 or
1471 The setting for ABS.fmt/NEG.fmt instruction execution
1472 mode generally follows that for the NaN encoding,
1473 except where unsupported by hardware.
1475 ignore_loglevel [KNL]
1476 Ignore loglevel setting - this will print /all/
1477 kernel messages to the console. Useful for debugging.
1478 We also add it as printk module parameter, so users
1479 could change it dynamically, usually by
1480 /sys/module/printk/parameters/ignore_loglevel.
1483 Ignore RLIMIT_DATA setting for data mappings,
1484 print warning at first misuse. Can be changed via
1485 /sys/module/kernel/parameters/ignore_rlimit_data.
1487 ihash_entries= [KNL]
1488 Set number of hash buckets for inode cache.
1490 ima_appraise= [IMA] appraise integrity measurements
1491 Format: { "off" | "enforce" | "fix" | "log" }
1494 ima_appraise_tcb [IMA]
1495 The builtin appraise policy appraises all files
1498 ima_canonical_fmt [IMA]
1499 Use the canonical format for the binary runtime
1500 measurements, instead of host native format.
1503 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1507 The list of supported hash algorithms is defined
1508 in crypto/hash_info.h.
1511 The builtin policies to load during IMA setup.
1512 Format: "tcb | appraise_tcb | secure_boot"
1514 The "tcb" policy measures all programs exec'd, files
1515 mmap'd for exec, and all files opened with the read
1516 mode bit set by either the effective uid (euid=0) or
1519 The "appraise_tcb" policy appraises the integrity of
1520 all files owned by root. (This is the equivalent
1521 of ima_appraise_tcb.)
1523 The "secure_boot" policy appraises the integrity
1524 of files (eg. kexec kernel image, kernel modules,
1525 firmware, policy, etc) based on file signatures.
1527 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1528 Load a policy which meets the needs of the Trusted
1529 Computing Base. This means IMA will measure all
1530 programs exec'd, files mmap'd for exec, and all files
1531 opened for read by uid=0.
1534 Select one of defined IMA measurements template formats.
1535 Formats: { "ima" | "ima-ng" | "ima-sig" }
1539 [IMA] Define a custom template format.
1540 Format: { "field1|...|fieldN" }
1542 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1543 Format: <min_file_size>
1544 Set the minimal file size for using asynchronous hash.
1545 If left unspecified, ahash usage is disabled.
1547 ahash performance varies for different data sizes on
1548 different crypto accelerators. This option can be used
1549 to achieve the best performance for a particular HW.
1551 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1553 Set hashing buffer size. Default: 4k.
1555 ahash performance varies for different chunk sizes on
1556 different crypto accelerators. This option can be used
1557 to achieve best performance for particular HW.
1561 Run specified binary instead of /sbin/init as init
1564 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1565 for working out where the kernel is dying during
1568 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1569 initcall functions. Useful for debugging built-in
1570 modules and initcalls.
1572 initrd= [BOOT] Specify the location of the initial ramdisk
1574 init_pkru= [x86] Specify the default memory protection keys rights
1575 register contents for all processes. 0x55555554 by
1576 default (disallow access to all but pkey 0). Can
1577 override in debugfs after boot.
1579 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1582 int_pln_enable [x86] Enable power limit notification interrupt
1584 integrity_audit=[IMA]
1585 Format: { "0" | "1" }
1586 0 -- basic integrity auditing messages. (Default)
1587 1 -- additional integrity auditing messages.
1589 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1591 Enable intel iommu driver.
1593 Disable intel iommu driver.
1594 igfx_off [Default Off]
1595 By default, gfx is mapped as normal device. If a gfx
1596 device has a dedicated DMAR unit, the DMAR unit is
1597 bypassed by not enabling DMAR with this option. In
1598 this case, gfx device will use physical address for
1601 With this option iommu will not optimize to look
1602 for io virtual address below 32-bit forcing dual
1603 address cycle on pci bus for cards supporting greater
1604 than 32-bit addressing. The default is to look
1605 for translation below 32-bit and if not available
1606 then look in the higher range.
1607 strict [Default Off]
1608 With this option on every unmap_single operation will
1609 result in a hardware IOTLB flush operation as opposed
1610 to batching them for performance.
1611 sp_off [Default Off]
1612 By default, super page will be supported if Intel IOMMU
1613 has the capability. With this option, super page will
1615 ecs_off [Default Off]
1616 By default, extended context tables will be supported if
1617 the hardware advertises that it has support both for the
1618 extended tables themselves, and also PASID support. With
1619 this option set, extended tables will not be used even
1620 on hardware which claims to support them.
1621 tboot_noforce [Default Off]
1622 Do not force the Intel IOMMU enabled under tboot.
1623 By default, tboot will force Intel IOMMU on, which
1624 could harm performance of some high-throughput
1625 devices like 40GBit network cards, even if identity
1627 Note that using this option lowers the security
1628 provided by tboot because it makes the system
1629 vulnerable to DMA attacks.
1631 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1632 0 disables intel_idle and fall back on acpi_idle.
1633 1 to 9 specify maximum depth of C-state.
1637 Do not enable intel_pstate as the default
1638 scaling driver for the supported processors
1640 Use intel_pstate as a scaling driver, but configure it
1641 to work with generic cpufreq governors (instead of
1642 enabling its internal governor). This mode cannot be
1643 used along with the hardware-managed P-states (HWP)
1646 Enable intel_pstate on systems that prohibit it by default
1647 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1648 instead of acpi-cpufreq may disable platform features, such
1649 as thermal controls and power capping, that rely on ACPI
1650 P-States information being indicated to OSPM and therefore
1651 should be used with caution. This option does not work with
1652 processors that aren't supported by the intel_pstate driver
1653 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1655 Do not enable hardware P state control (HWP)
1658 Only load intel_pstate on systems which support
1659 hardware P state control (HWP) if available.
1661 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1662 Description Table, specifies preferred power management
1663 profile as "Enterprise Server" or "Performance Server",
1664 then this feature is turned on by default.
1666 Allow per-logical-CPU P-State performance control limits using
1667 cpufreq sysfs interface
1669 intremap= [X86-64, Intel-IOMMU]
1670 on enable Interrupt Remapping (default)
1671 off disable Interrupt Remapping
1672 nosid disable Source ID checking
1674 BIOS x2APIC opt-out request will be ignored
1675 nopost disable Interrupt Posting
1677 iomem= Disable strict checking of access to MMIO memory
1678 strict regions from userspace.
1693 nobypass [PPC/POWERNV]
1694 Disable IOMMU bypass, using IOMMU for PCI devices.
1697 [ARM64] Configure DMA to bypass the IOMMU by default.
1698 Format: { "0" | "1" }
1699 0 - Use IOMMU translation for DMA.
1700 1 - Bypass the IOMMU for DMA.
1701 unset - Use IOMMU translation for DMA.
1703 io7= [HW] IO7 for Marvel based alpha systems
1704 See comment before marvel_specify_io7 in
1705 arch/alpha/kernel/core_marvel.c.
1707 io_delay= [X86] I/O delay method
1709 Standard port 0x80 based delay
1711 Alternate port 0xed based delay (needed on some systems)
1713 Simple two microseconds delay
1718 See Documentation/filesystems/nfs/nfsroot.txt.
1720 irqaffinity= [SMP] Set the default irq affinity mask
1721 The argument is a cpu list, as described above.
1724 When an interrupt is not handled search all handlers
1725 for it. Intended to get systems with badly broken
1729 When an interrupt is not handled search all handlers
1730 for it. Also check all handlers each timer
1731 interrupt. Intended to get systems with badly broken
1735 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1737 isolcpus= [KNL,SMP] Isolate CPUs from the general scheduler.
1738 The argument is a cpu list, as described above.
1740 This option can be used to specify one or more CPUs
1741 to isolate from the general SMP balancing and scheduling
1742 algorithms. You can move a process onto or off an
1743 "isolated" CPU via the CPU affinity syscalls or cpuset.
1744 <cpu number> begins at 0 and the maximum value is
1745 "number of CPUs in system - 1".
1747 This option is the preferred way to isolate CPUs. The
1748 alternative -- manually setting the CPU mask of all
1749 tasks in the system -- can cause problems and
1750 suboptimal load balancer performance.
1754 ivrs_ioapic [HW,X86_64]
1755 Provide an override to the IOAPIC-ID<->DEVICE-ID
1756 mapping provided in the IVRS ACPI table. For
1757 example, to map IOAPIC-ID decimal 10 to
1758 PCI device 00:14.0 write the parameter as:
1759 ivrs_ioapic[10]=00:14.0
1761 ivrs_hpet [HW,X86_64]
1762 Provide an override to the HPET-ID<->DEVICE-ID
1763 mapping provided in the IVRS ACPI table. For
1764 example, to map HPET-ID decimal 0 to
1765 PCI device 00:14.0 write the parameter as:
1766 ivrs_hpet[0]=00:14.0
1768 ivrs_acpihid [HW,X86_64]
1769 Provide an override to the ACPI-HID:UID<->DEVICE-ID
1770 mapping provided in the IVRS ACPI table. For
1771 example, to map UART-HID:UID AMD0020:0 to
1772 PCI device 00:14.5 write the parameter as:
1773 ivrs_acpihid[00:14.5]=AMD0020:0
1775 js= [HW,JOY] Analog joystick
1776 See Documentation/input/joystick.txt.
1779 When CONFIG_RANDOMIZE_BASE is set, this disables
1780 kernel and module base offset ASLR (Address Space
1781 Layout Randomization).
1784 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
1785 report on every invalid memory access. Without this
1786 parameter KASAN will print report only for the first
1791 kernelcore= [KNL,X86,IA-64,PPC]
1792 Format: nn[KMGTPE] | "mirror"
1794 specifies the amount of memory usable by the kernel
1795 for non-movable allocations. The requested amount is
1796 spread evenly throughout all nodes in the system. The
1797 remaining memory in each node is used for Movable
1798 pages. In the event, a node is too small to have both
1799 kernelcore and Movable pages, kernelcore pages will
1800 take priority and other nodes will have a larger number
1801 of Movable pages. The Movable zone is used for the
1802 allocation of pages that may be reclaimed or moved
1803 by the page migration subsystem. This means that
1804 HugeTLB pages may not be allocated from this zone.
1805 Note that allocations like PTEs-from-HighMem still
1806 use the HighMem zone if it exists, and the Normal
1807 zone if it does not.
1809 Instead of specifying the amount of memory (nn[KMGTPE]),
1810 you can specify "mirror" option. In case "mirror"
1811 option is specified, mirrored (reliable) memory is used
1812 for non-movable allocations and remaining memory is used
1813 for Movable pages. nn[KMGTPE] and "mirror" are exclusive,
1814 so you can NOT specify nn[KMGTPE] and "mirror" at the same
1817 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1818 Format: <Controller#>[,poll interval]
1819 The controller # is the number of the ehci usb debug
1820 port as it is probed via PCI. The poll interval is
1821 optional and is the number seconds in between
1822 each poll cycle to the debug port in case you need
1823 the functionality for interrupting the kernel with
1824 gdb or control-c on the dbgp connection. When
1825 not using this parameter you use sysrq-g to break into
1826 the kernel debugger.
1828 kgdboc= [KGDB,HW] kgdb over consoles.
1829 Requires a tty driver that supports console polling,
1830 or a supported polling keyboard driver (non-usb).
1831 Serial only format: <serial_device>[,baud]
1832 keyboard only format: kbd
1833 keyboard and serial format: kbd,<serial_device>[,baud]
1834 Optional Kernel mode setting:
1835 kms, kbd format: kms,kbd
1836 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1838 kgdbwait [KGDB] Stop kernel execution and enter the
1839 kernel debugger at the earliest opportunity.
1841 kmac= [MIPS] korina ethernet MAC address.
1842 Configure the RouterBoard 532 series on-chip
1843 Ethernet adapter MAC address.
1845 kmemleak= [KNL] Boot-time kmemleak enable/disable
1846 Valid arguments: on, off
1848 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
1851 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
1852 Default is 0 (don't ignore, but inject #GP)
1854 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
1858 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
1859 Default is 1 (enabled)
1861 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
1863 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
1865 kvm-arm.vgic_v3_group0_trap=
1866 [KVM,ARM] Trap guest accesses to GICv3 group-0
1869 kvm-arm.vgic_v3_group1_trap=
1870 [KVM,ARM] Trap guest accesses to GICv3 group-1
1873 kvm-arm.vgic_v3_common_trap=
1874 [KVM,ARM] Trap guest accesses to GICv3 common
1877 kvm-intel.ept= [KVM,Intel] Disable extended page tables
1878 (virtualized MMU) support on capable Intel chips.
1879 Default is 1 (enabled)
1881 kvm-intel.emulate_invalid_guest_state=
1882 [KVM,Intel] Enable emulation of invalid guest states
1883 Default is 0 (disabled)
1885 kvm-intel.flexpriority=
1886 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
1887 Default is 1 (enabled)
1890 [KVM,Intel] Enable VMX nesting (nVMX).
1891 Default is 0 (disabled)
1893 kvm-intel.unrestricted_guest=
1894 [KVM,Intel] Disable unrestricted guest feature
1895 (virtualized real and unpaged mode) on capable
1896 Intel chips. Default is 1 (enabled)
1898 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
1901 Valid arguments: never, cond, always
1903 always: L1D cache flush on every VMENTER.
1904 cond: Flush L1D on VMENTER only when the code between
1905 VMEXIT and VMENTER can leak host memory.
1906 never: Disables the mitigation
1908 Default is cond (do L1 cache flush in specific instances)
1910 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
1911 feature (tagged TLBs) on capable Intel chips.
1912 Default is 1 (enabled)
1914 l1tf= [X86] Control mitigation of the L1TF vulnerability on
1917 The kernel PTE inversion protection is unconditionally
1918 enabled and cannot be disabled.
1921 Provides all available mitigations for the
1922 L1TF vulnerability. Disables SMT and
1923 enables all mitigations in the
1924 hypervisors, i.e. unconditional L1D flush.
1926 SMT control and L1D flush control via the
1927 sysfs interface is still possible after
1928 boot. Hypervisors will issue a warning
1929 when the first VM is started in a
1930 potentially insecure configuration,
1931 i.e. SMT enabled or L1D flush disabled.
1934 Same as 'full', but disables SMT and L1D
1935 flush runtime control. Implies the
1936 'nosmt=force' command line option.
1937 (i.e. sysfs control of SMT is disabled.)
1940 Leaves SMT enabled and enables the default
1941 hypervisor mitigation, i.e. conditional
1944 SMT control and L1D flush control via the
1945 sysfs interface is still possible after
1946 boot. Hypervisors will issue a warning
1947 when the first VM is started in a
1948 potentially insecure configuration,
1949 i.e. SMT enabled or L1D flush disabled.
1953 Disables SMT and enables the default
1954 hypervisor mitigation.
1956 SMT control and L1D flush control via the
1957 sysfs interface is still possible after
1958 boot. Hypervisors will issue a warning
1959 when the first VM is started in a
1960 potentially insecure configuration,
1961 i.e. SMT enabled or L1D flush disabled.
1964 Same as 'flush', but hypervisors will not
1965 warn when a VM is started in a potentially
1966 insecure configuration.
1969 Disables hypervisor mitigations and doesn't
1971 It also drops the swap size and available
1972 RAM limit restriction on both hypervisor and
1977 For details see: Documentation/admin-guide/l1tf.rst
1983 lapic [X86-32,APIC] Enable the local APIC even if BIOS
1986 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
1987 value for LAPIC timer one-shot implementation. Default
1988 back to the programmable timer unit in the LAPIC.
1990 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
1993 libata.dma= [LIBATA] DMA control
1994 libata.dma=0 Disable all PATA and SATA DMA
1995 libata.dma=1 PATA and SATA Disk DMA only
1996 libata.dma=2 ATAPI (CDROM) DMA only
1997 libata.dma=4 Compact Flash DMA only
1998 Combinations also work, so libata.dma=3 enables DMA
1999 for disks and CDROMs, but not CFs.
2001 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2002 libata.ignore_hpa=0 keep BIOS limits (default)
2003 libata.ignore_hpa=1 ignore limits, using full disk
2005 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2009 libata.force= [LIBATA] Force configurations. The format is comma
2010 separated list of "[ID:]VAL" where ID is
2011 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2012 matching port, link or device. Basically, it matches
2013 the ATA ID string printed on console by libata. If
2014 the whole ID part is omitted, the last PORT and DEVICE
2015 values are used. If ID hasn't been specified yet, the
2016 configuration applies to all ports, links and devices.
2018 If only DEVICE is omitted, the parameter applies to
2019 the port and all links and devices behind it. DEVICE
2020 number of 0 either selects the first device or the
2021 first fan-out link behind PMP device. It does not
2022 select the host link. DEVICE number of 15 selects the
2023 host link and device attached to it.
2025 The VAL specifies the configuration to force. As long
2026 as there's no ambiguity shortcut notation is allowed.
2027 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2028 The following configurations can be forced.
2030 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2031 Any ID with matching PORT is used.
2033 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2035 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2036 udma[/][16,25,33,44,66,100,133] notation is also
2039 * [no]ncq: Turn on or off NCQ.
2041 * [no]ncqtrim: Turn off queued DSM TRIM.
2043 * nohrst, nosrst, norst: suppress hard, soft
2046 * rstonce: only attempt one reset during
2047 hot-unplug link recovery
2049 * dump_id: dump IDENTIFY data.
2051 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2053 * disable: Disable this device.
2055 If there are multiple matching configurations changing
2056 the same attribute, the last one is used.
2058 memblock=debug [KNL] Enable memblock debug messages.
2060 load_ramdisk= [RAM] List of ramdisks to load from floppy
2061 See Documentation/blockdev/ramdisk.txt.
2063 lockd.nlm_grace_period=P [NFS] Assign grace period.
2066 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2069 lockd.nlm_timeout=T [NFS] Assign timeout value.
2072 lockd.nlm_udpport=M [NFS] Assign UDP port.
2075 locktorture.nreaders_stress= [KNL]
2076 Set the number of locking read-acquisition kthreads.
2077 Defaults to being automatically set based on the
2078 number of online CPUs.
2080 locktorture.nwriters_stress= [KNL]
2081 Set the number of locking write-acquisition kthreads.
2083 locktorture.onoff_holdoff= [KNL]
2084 Set time (s) after boot for CPU-hotplug testing.
2086 locktorture.onoff_interval= [KNL]
2087 Set time (s) between CPU-hotplug operations, or
2088 zero to disable CPU-hotplug testing.
2090 locktorture.shuffle_interval= [KNL]
2091 Set task-shuffle interval (jiffies). Shuffling
2092 tasks allows some CPUs to go into dyntick-idle
2093 mode during the locktorture test.
2095 locktorture.shutdown_secs= [KNL]
2096 Set time (s) after boot system shutdown. This
2097 is useful for hands-off automated testing.
2099 locktorture.stat_interval= [KNL]
2100 Time (s) between statistics printk()s.
2102 locktorture.stutter= [KNL]
2103 Time (s) to stutter testing, for example,
2104 specifying five seconds causes the test to run for
2105 five seconds, wait for five seconds, and so on.
2106 This tests the locking primitive's ability to
2107 transition abruptly to and from idle.
2109 locktorture.torture_runnable= [BOOT]
2110 Start locktorture running at boot time.
2112 locktorture.torture_type= [KNL]
2113 Specify the locking implementation to test.
2115 locktorture.verbose= [KNL]
2116 Enable additional printk() statements.
2118 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2121 loglevel= All Kernel Messages with a loglevel smaller than the
2122 console loglevel will be printed to the console. It can
2123 also be changed with klogd or other programs. The
2124 loglevels are defined as follows:
2126 0 (KERN_EMERG) system is unusable
2127 1 (KERN_ALERT) action must be taken immediately
2128 2 (KERN_CRIT) critical conditions
2129 3 (KERN_ERR) error conditions
2130 4 (KERN_WARNING) warning conditions
2131 5 (KERN_NOTICE) normal but significant condition
2132 6 (KERN_INFO) informational
2133 7 (KERN_DEBUG) debug-level messages
2135 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2136 in bytes. n must be a power of two and greater
2137 than the minimal size. The minimal size is defined
2138 by LOG_BUF_SHIFT kernel config parameter. There is
2139 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2140 that allows to increase the default size depending on
2141 the number of CPUs. See init/Kconfig for more details.
2143 logo.nologo [FB] Disables display of the built-in Linux logo.
2144 This may be used to provide more screen space for
2145 kernel log messages and is useful when debugging
2146 kernel boot problems.
2148 lp=0 [LP] Specify parallel ports to use, e.g,
2149 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2150 lp=reset first parallel port). 'lp=0' disables the
2151 lp=auto printer driver. 'lp=reset' (which can be
2152 specified in addition to the ports) causes
2153 attached printers to be reset. Using
2154 lp=port1,port2,... specifies the parallel ports
2155 to associate lp devices with, starting with
2156 lp0. A port specification may be 'none' to skip
2157 that lp device, or a parport name such as
2158 'parport0'. Specifying 'lp=auto' instead of a
2159 port specification list means that device IDs
2160 from each port should be examined, to see if
2161 an IEEE 1284-compliant printer is attached; if
2162 so, the driver will manage that printer.
2163 See also header of drivers/char/lp.c.
2166 Sets loops_per_jiffy to given constant, thus avoiding
2167 time-consuming boot-time autodetection (up to 250 ms per
2168 CPU). 0 enables autodetection (default). To determine
2169 the correct value for your kernel, boot with normal
2170 autodetection and see what value is printed. Note that
2171 on SMP systems the preset will be applied to all CPUs,
2172 which is likely to cause problems if your CPUs need
2173 significantly divergent settings. An incorrect value
2174 will cause delays in the kernel to be wrong, leading to
2175 unpredictable I/O errors and other breakage. Although
2176 unlikely, in the extreme case this might damage your
2180 Format: <io>,<irq>,<dma>
2182 machvec= [IA-64] Force the use of a particular machine-vector
2183 (machvec) in a generic kernel.
2184 Example: machvec=hpzx1_swiotlb
2186 machtype= [Loongson] Share the same kernel image file between different
2188 Example: machtype=lemote-yeeloong-2f-7inch
2190 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2191 than or equal to this physical address is ignored.
2193 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2194 will bring up during bootup. maxcpus=n : n >= 0 limits
2195 the kernel to bring up 'n' processors. Surely after
2196 bootup you can bring up the other plugged cpu by executing
2197 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2198 only takes effect during system bootup.
2199 While n=0 is a special case, it is equivalent to "nosmp",
2200 which also disables the IO APIC.
2202 max_loop= [LOOP] The number of loop block devices that get
2203 (loop.max_loop) unconditionally pre-created at init time. The default
2204 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2205 of statically allocating a predefined number, loop
2206 devices can be requested on-demand with the
2207 /dev/loop-control interface.
2209 mce [X86-32] Machine Check Exception
2211 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
2213 md= [HW] RAID subsystems devices and level
2214 See Documentation/admin-guide/md.rst.
2217 Format: <first>,<last>
2218 Specifies range of consoles to be captured by the MDA.
2220 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2221 Amount of memory to be used when the kernel is not able
2222 to see the whole system memory or for test.
2223 [X86] Work as limiting max address. Use together
2224 with memmap= to avoid physical address space collisions.
2225 Without memmap= PCI devices could be placed at addresses
2226 belonging to unused RAM.
2228 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2232 [KNL,SH] Allow user to override the default size for
2233 per-device physically contiguous DMA buffers.
2235 memhp_default_state=online/offline
2236 [KNL] Set the initial state for the memory hotplug
2237 onlining policy. If not specified, the default value is
2238 set according to the
2239 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2241 See Documentation/memory-hotplug.txt.
2243 memmap=exactmap [KNL,X86] Enable setting of an exact
2244 E820 memory map, as specified by the user.
2245 Such memmap=exactmap lines can be constructed based on
2246 BIOS output or other requirements. See the memmap=nn@ss
2249 memmap=nn[KMG]@ss[KMG]
2250 [KNL] Force usage of a specific region of memory.
2251 Region of memory to be used is from ss to ss+nn.
2252 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2253 which limits max address to nn[KMG].
2254 Multiple different regions can be specified,
2257 memmap=100M@2G,100M#3G,1G!1024G
2259 memmap=nn[KMG]#ss[KMG]
2260 [KNL,ACPI] Mark specific memory as ACPI data.
2261 Region of memory to be marked is from ss to ss+nn.
2263 memmap=nn[KMG]$ss[KMG]
2264 [KNL,ACPI] Mark specific memory as reserved.
2265 Region of memory to be reserved is from ss to ss+nn.
2266 Example: Exclude memory from 0x18690000-0x1869ffff
2267 memmap=64K$0x18690000
2269 memmap=0x10000$0x18690000
2270 Some bootloaders may need an escape character before '$',
2271 like Grub2, otherwise '$' and the following number
2274 memmap=nn[KMG]!ss[KMG]
2275 [KNL,X86] Mark specific memory as protected.
2276 Region of memory to be used, from ss to ss+nn.
2277 The memory region may be marked as e820 type 12 (0xc)
2278 and is NVDIMM or ADR memory.
2280 memory_corruption_check=0/1 [X86]
2281 Some BIOSes seem to corrupt the first 64k of
2282 memory when doing things like suspend/resume.
2283 Setting this option will scan the memory
2284 looking for corruption. Enabling this will
2285 both detect corruption and prevent the kernel
2286 from using the memory being corrupted.
2287 However, its intended as a diagnostic tool; if
2288 repeatable BIOS-originated corruption always
2289 affects the same memory, you can use memmap=
2290 to prevent the kernel from using that memory.
2292 memory_corruption_check_size=size [X86]
2293 By default it checks for corruption in the low
2294 64k, making this memory unavailable for normal
2295 use. Use this parameter to scan for
2296 corruption in more or less memory.
2298 memory_corruption_check_period=seconds [X86]
2299 By default it checks for corruption every 60
2300 seconds. Use this parameter to check at some
2301 other rate. 0 disables periodic checking.
2303 memtest= [KNL,X86,ARM] Enable memtest
2305 default : 0 <disable>
2306 Specifies the number of memtest passes to be
2307 performed. Each pass selects another test
2308 pattern from a given set of patterns. Memtest
2309 fills the memory with this pattern, validates
2310 memory contents and reserves bad memory
2311 regions that are detected.
2313 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2314 Valid arguments: on, off
2315 Default (depends on kernel configuration option):
2316 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2317 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2318 mem_encrypt=on: Activate SME
2319 mem_encrypt=off: Do not activate SME
2321 Refer to Documentation/x86/amd-memory-encryption.txt
2322 for details on when memory encryption can be activated.
2324 mem_sleep_default= [SUSPEND] Default system suspend mode:
2325 s2idle - Suspend-To-Idle
2326 shallow - Power-On Suspend or equivalent (if supported)
2327 deep - Suspend-To-RAM or equivalent (if supported)
2328 See Documentation/power/states.txt.
2330 meye.*= [HW] Set MotionEye Camera parameters
2331 See Documentation/video4linux/meye.txt.
2333 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2334 Multi-Function General Purpose Timers on AMD Geode
2337 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2338 the BIOS has incorrectly applied a workaround. TinyBIOS
2339 version 0.98 is known to be affected, 0.99 fixes the
2340 problem by letting the user disable the workaround.
2344 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2345 physical address is ignored.
2347 mini2440= [ARM,HW,KNL]
2348 Format:[0..2][b][c][t]
2350 MINI2440 configuration specification:
2351 0 - The attached screen is the 3.5" TFT
2352 1 - The attached screen is the 7" TFT
2353 2 - The VGA Shield is attached (1024x768)
2354 Leaving out the screen size parameter will not load
2355 the TFT driver, and the framebuffer will be left
2357 b - Enable backlight. The TFT backlight pin will be
2358 linked to the kernel VESA blanking code and a GPIO
2359 LED. This parameter is not necessary when using the
2361 c - Enable the s3c camera interface.
2362 t - Reserved for enabling touchscreen support. The
2363 touchscreen support is not enabled in the mainstream
2364 kernel as of 2.6.30, a preliminary port can be found
2365 in the "bleeding edge" mini2440 support kernel at
2366 http://repo.or.cz/w/linux-2.6/mini2440.git
2369 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2370 parameter allows control of the logging verbosity for
2371 the additional memory initialisation checks. A value
2372 of 0 disables mminit logging and a level of 4 will
2373 log everything. Information is printed at KERN_DEBUG
2374 so loglevel=8 may also need to be specified.
2377 [KNL] When CONFIG_MODULE_SIG is set, this means that
2378 modules without (valid) signatures will fail to load.
2379 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2380 is always true, so this option does nothing.
2382 module_blacklist= [KNL] Do not load a comma-separated list of
2383 modules. Useful for debugging problem modules.
2386 [MOUSE] Maximum time between finger touching and
2387 leaving touchpad surface for touch to be considered
2388 a tap and be reported as a left button click (for
2389 touchpads working in absolute mode only).
2391 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2392 reporting absolute coordinates, such as tablets
2393 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2394 reporting absolute coordinates, such as tablets
2396 movablecore=nn[KMG] [KNL,X86,IA-64,PPC] This parameter
2397 is similar to kernelcore except it specifies the
2398 amount of memory used for migratable allocations.
2399 If both kernelcore and movablecore is specified,
2400 then kernelcore will be at *least* the specified
2401 value but may be more. If movablecore on its own
2402 is specified, the administrator must be careful
2403 that the amount of memory usable for all allocations
2406 movable_node [KNL] Boot-time switch to make hotplugable memory
2407 NUMA nodes to be movable. This means that the memory
2408 of such nodes will be usable only for movable
2409 allocations which rules out almost all kernel
2410 allocations. Use with caution!
2412 MTD_Partition= [MTD]
2413 Format: <name>,<region-number>,<size>,<offset>
2415 MTD_Region= [MTD] Format:
2416 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2419 See drivers/mtd/cmdlinepart.c.
2421 multitce=off [PPC] This parameter disables the use of the pSeries
2422 firmware feature for updating multiple TCE entries
2425 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2427 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2429 boundary - index of last SLC block on Flex-OneNAND.
2430 The remaining blocks are configured as MLC blocks.
2431 lock - Configure if Flex-OneNAND boundary should be locked.
2432 Once locked, the boundary cannot be changed.
2433 1 indicates lock status, 0 indicates unlock status.
2436 ARM/S3C2412 JIVE boot control
2438 See arch/arm/mach-s3c2412/mach-jive.c
2440 mtouchusb.raw_coordinates=
2441 [HW] Make the MicroTouch USB driver use raw coordinates
2442 ('y', default) or cooked coordinates ('n')
2444 mtrr_chunk_size=nn[KMG] [X86]
2445 used for mtrr cleanup. It is largest continuous chunk
2446 that could hold holes aka. UC entries.
2448 mtrr_gran_size=nn[KMG] [X86]
2449 Used for mtrr cleanup. It is granularity of mtrr block.
2451 Large value could prevent small alignment from
2454 mtrr_spare_reg_nr=n [X86]
2456 Range: 0,7 : spare reg number
2458 Used for mtrr cleanup. It is spare mtrr entries number.
2459 Set to 2 or more if your graphical card needs more.
2461 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2463 netdev= [NET] Network devices parameters
2464 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2465 Note that mem_start is often overloaded to mean
2466 something different and driver-specific.
2467 This usage is only documented in each driver source
2471 [NETFILTER] Enable connection tracking flow accounting
2472 0 to disable accounting
2473 1 to enable accounting
2476 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2477 See Documentation/filesystems/nfs/nfsroot.txt.
2479 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2480 See Documentation/filesystems/nfs/nfsroot.txt.
2482 nfsrootdebug [NFS] enable nfsroot debugging messages.
2483 See Documentation/filesystems/nfs/nfsroot.txt.
2485 nfs.callback_nr_threads=
2486 [NFSv4] set the total number of threads that the
2487 NFS client will assign to service NFSv4 callback
2490 nfs.callback_tcpport=
2491 [NFS] set the TCP port on which the NFSv4 callback
2492 channel should listen.
2495 [NFS] sets the pathname to the program which is used
2496 to update the NFS client cache entries.
2498 nfs.cache_getent_timeout=
2499 [NFS] sets the timeout after which an attempt to
2500 update a cache entry is deemed to have failed.
2502 nfs.idmap_cache_timeout=
2503 [NFS] set the maximum lifetime for idmapper cache
2507 [NFS] enable 64-bit inode numbers.
2508 If zero, the NFS client will fake up a 32-bit inode
2509 number for the readdir() and stat() syscalls instead
2510 of returning the full 64-bit number.
2511 The default is to return 64-bit inode numbers.
2513 nfs.max_session_cb_slots=
2514 [NFSv4.1] Sets the maximum number of session
2515 slots the client will assign to the callback
2516 channel. This determines the maximum number of
2517 callbacks the client will process in parallel for
2518 a particular server.
2520 nfs.max_session_slots=
2521 [NFSv4.1] Sets the maximum number of session slots
2522 the client will attempt to negotiate with the server.
2523 This limits the number of simultaneous RPC requests
2524 that the client can send to the NFSv4.1 server.
2525 Note that there is little point in setting this
2526 value higher than the max_tcp_slot_table_limit.
2528 nfs.nfs4_disable_idmapping=
2529 [NFSv4] When set to the default of '1', this option
2530 ensures that both the RPC level authentication
2531 scheme and the NFS level operations agree to use
2532 numeric uids/gids if the mount is using the
2533 'sec=sys' security flavour. In effect it is
2534 disabling idmapping, which can make migration from
2535 legacy NFSv2/v3 systems to NFSv4 easier.
2536 Servers that do not support this mode of operation
2537 will be autodetected by the client, and it will fall
2538 back to using the idmapper.
2539 To turn off this behaviour, set the value to '0'.
2541 [NFS4] Specify an additional fixed unique ident-
2542 ification string that NFSv4 clients can insert into
2543 their nfs_client_id4 string. This is typically a
2544 UUID that is generated at system install time.
2546 nfs.send_implementation_id =
2547 [NFSv4.1] Send client implementation identification
2548 information in exchange_id requests.
2549 If zero, no implementation identification information
2551 The default is to send the implementation identification
2554 nfs.recover_lost_locks =
2555 [NFSv4] Attempt to recover locks that were lost due
2556 to a lease timeout on the server. Please note that
2557 doing this risks data corruption, since there are
2558 no guarantees that the file will remain unchanged
2559 after the locks are lost.
2560 If you want to enable the kernel legacy behaviour of
2561 attempting to recover these locks, then set this
2563 The default parameter value of '0' causes the kernel
2564 not to attempt recovery of lost locks.
2566 nfs4.layoutstats_timer =
2567 [NFSv4.2] Change the rate at which the kernel sends
2568 layoutstats to the pNFS metadata server.
2570 Setting this to value to 0 causes the kernel to use
2571 whatever value is the default set by the layout
2572 driver. A non-zero value sets the minimum interval
2573 in seconds between layoutstats transmissions.
2575 nfsd.nfs4_disable_idmapping=
2576 [NFSv4] When set to the default of '1', the NFSv4
2577 server will return only numeric uids and gids to
2578 clients using auth_sys, and will accept numeric uids
2579 and gids from such clients. This is intended to ease
2580 migration from NFSv2/v3.
2582 nmi_debug= [KNL,SH] Specify one or more actions to take
2583 when a NMI is triggered.
2584 Format: [state][,regs][,debounce][,die]
2586 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2587 Format: [panic,][nopanic,][num]
2589 0 - turn hardlockup detector in nmi_watchdog off
2590 1 - turn hardlockup detector in nmi_watchdog on
2591 When panic is specified, panic when an NMI watchdog
2592 timeout occurs (or 'nopanic' to override the opposite
2593 default). To disable both hard and soft lockup detectors,
2594 please see 'nowatchdog'.
2595 This is useful when you use a panic=... timeout and
2596 need the box quickly up again.
2598 netpoll.carrier_timeout=
2599 [NET] Specifies amount of time (in seconds) that
2600 netpoll should wait for a carrier. By default netpoll
2603 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2604 emulation library even if a 387 maths coprocessor
2608 [HW] Never suspend the console
2609 Disable suspending of consoles during suspend and
2610 hibernate operations. Once disabled, debugging
2611 messages can reach various consoles while the rest
2612 of the system is being put to sleep (ie, while
2613 debugging driver suspend/resume hooks). This may
2614 not work reliably with all consoles, but is known
2615 to work with serial and VGA consoles.
2616 To facilitate more flexible debugging, we also add
2617 console_suspend, a printk module parameter to control
2618 it. Users could use console_suspend (usually
2619 /sys/module/printk/parameters/console_suspend) to
2620 turn on/off it dynamically.
2622 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2623 caches in the slab allocator. Saves per-node memory,
2624 but will impact performance.
2628 noaltinstr [S390] Disables alternative instructions patching
2629 (CPU alternatives feature).
2631 noapic [SMP,APIC] Tells the kernel to not make use of any
2632 IOAPICs that may be present in the system.
2634 noautogroup Disable scheduler automatic task group creation.
2636 nobats [PPC] Do not use BATs for mapping kernel lowmem
2637 on "Classic" PPC cores.
2641 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2643 nodelayacct [KNL] Disable per-task delay accounting
2645 nodsp [SH] Disable hardware DSP at boot time.
2647 noefi Disable EFI runtime services support.
2652 On X86-32 available only on PAE configured kernels.
2653 noexec=on: enable non-executable mappings (default)
2654 noexec=off: disable non-executable mappings
2657 Disable SMAP (Supervisor Mode Access Prevention)
2658 even if it is supported by processor.
2661 Disable SMEP (Supervisor Mode Execution Prevention)
2662 even if it is supported by processor.
2665 This affects only 32-bit executables.
2666 noexec32=on: enable non-executable mappings (default)
2667 read doesn't imply executable mappings
2668 noexec32=off: disable non-executable mappings
2669 read implies executable mappings
2671 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2673 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2674 register save and restore. The kernel will only save
2675 legacy floating-point registers on task switch.
2677 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2679 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
2680 Equivalent to smt=1.
2682 [KNL,x86] Disable symmetric multithreading (SMT).
2683 nosmt=force: Force disable SMT, cannot be undone
2684 via the sysfs control file.
2686 nospectre_v2 [X86] Disable all mitigations for the Spectre variant 2
2687 (indirect branch prediction) vulnerability. System may
2688 allow data leaks with this option, which is equivalent
2691 nospec_store_bypass_disable
2692 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
2694 noxsave [BUGS=X86] Disables x86 extended register state save
2695 and restore using xsave. The kernel will fallback to
2696 enabling legacy floating-point and sse state.
2698 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2699 register states. The kernel will fall back to use
2700 xsave to save the states. By using this parameter,
2701 performance of saving the states is degraded because
2702 xsave doesn't support modified optimization while
2703 xsaveopt supports it on xsaveopt enabled systems.
2705 noxsaves [X86] Disables xsaves and xrstors used in saving and
2706 restoring x86 extended register state in compacted
2707 form of xsave area. The kernel will fall back to use
2708 xsaveopt and xrstor to save and restore the states
2709 in standard form of xsave area. By using this
2710 parameter, xsave area per process might occupy more
2711 memory on xsaves enabled systems.
2713 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2714 wfi(ARM) instruction doesn't work correctly and not to
2715 use it. This is also useful when using JTAG debugger.
2717 no_file_caps Tells the kernel not to honor file capabilities. The
2718 only way then for a file to be executed with privilege
2719 is to be setuid root or executed by root.
2721 nohalt [IA-64] Tells the kernel not to use the power saving
2722 function PAL_HALT_LIGHT when idle. This increases
2723 power-consumption. On the positive side, it reduces
2724 interrupt wake-up latency, which may improve performance
2725 in certain environments such as networked servers or
2728 nohibernate [HIBERNATION] Disable hibernation and resume.
2730 nohz= [KNL] Boottime enable/disable dynamic ticks
2731 Valid arguments: on, off
2734 nohz_full= [KNL,BOOT]
2735 The argument is a cpu list, as described above.
2736 In kernels built with CONFIG_NO_HZ_FULL=y, set
2737 the specified list of CPUs whose tick will be stopped
2738 whenever possible. The boot CPU will be forced outside
2739 the range to maintain the timekeeping. Any CPUs
2740 in this list will have their RCU callbacks offloaded,
2741 just as if they had also been called out in the
2742 rcu_nocbs= boot parameter.
2744 noiotrap [SH] Disables trapped I/O port accesses.
2746 noirqdebug [X86-32] Disables the code which attempts to detect and
2747 disable unhandled interrupt sources.
2749 no_timer_check [X86,APIC] Disables the code which tests for
2750 broken timer IRQ sources.
2752 noisapnp [ISAPNP] Disables ISA PnP code.
2754 noinitrd [RAM] Tells the kernel not to load any configured
2757 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
2759 [Deprecated - use intremap=off]
2763 noinvpcid [X86] Disable the INVPCID cpu feature.
2765 nojitter [IA-64] Disables jitter checking for ITC timers.
2767 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
2769 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
2773 [X86,PV_OPS] Disable paravirtualized VMware scheduler
2774 clock and use the default one.
2776 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
2777 steal time is computed, but won't influence scheduler
2780 nolapic [X86-32,APIC] Do not enable or use the local APIC.
2782 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
2784 noltlbs [PPC] Do not use large page/tlb entries for kernel
2785 lowmem mapping on PPC40x and PPC8xx
2787 nomca [IA-64] Disable machine check abort handling
2789 nomce [X86-32] Disable Machine Check Exception
2791 nomfgpt [X86-32] Disable Multi-Function General Purpose
2792 Timer usage (for AMD Geode machines).
2794 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
2795 shutdown the other cpus. Instead use the REBOOT_VECTOR
2798 nomodule Disable module load
2800 nopat [X86] Disable PAT (page attribute table extension of
2801 pagetables) support.
2803 nopcid [X86-64] Disable the PCID cpu feature.
2805 norandmaps Don't use address space randomization. Equivalent to
2806 echo 0 > /proc/sys/kernel/randomize_va_space
2808 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
2809 with UP alternatives
2811 nordrand [X86] Disable kernel use of the RDRAND and
2812 RDSEED instructions even if they are supported
2813 by the processor. RDRAND and RDSEED are still
2814 available to user space applications.
2816 noresume [SWSUSP] Disables resume and restores original swap
2819 no-scroll [VGA] Disables scrollback.
2820 This is required for the Braillex ib80-piezo Braille
2821 reader made by F.H. Papenmeier (Germany).
2825 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
2827 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
2828 and disable the IO APIC. legacy for "maxcpus=0".
2830 nosoftlockup [KNL] Disable the soft-lockup detector.
2832 nosync [HW,M68K] Disables sync negotiation for all devices.
2834 notsc [BUGS=X86-32] Disable Time Stamp Counter
2836 nowatchdog [KNL] Disable both lockup detectors, i.e.
2837 soft-lockup and NMI watchdog (hard-lockup).
2841 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
2843 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
2844 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
2845 Some features depend on CPU0. Known dependencies are:
2846 1. Resume from suspend/hibernate depends on CPU0.
2847 Suspend/hibernate will fail if CPU0 is offline and you
2848 need to online CPU0 before suspend/hibernate.
2849 2. PIC interrupts also depend on CPU0. CPU0 can't be
2850 removed if a PIC interrupt is detected.
2851 It's said poweroff/reboot may depend on CPU0 on some
2852 machines although I haven't seen such issues so far
2853 after CPU0 is offline on a few tested machines.
2854 If the dependencies are under your control, you can
2855 turn on cpu0_hotplug.
2857 nps_mtm_hs_ctr= [KNL,ARC]
2858 This parameter sets the maximum duration, in
2859 cycles, each HW thread of the CTOP can run
2860 without interruptions, before HW switches it.
2861 The actual maximum duration is 16 times this
2863 Format: integer between 1 and 255
2866 nptcg= [IA-64] Override max number of concurrent global TLB
2867 purges which is reported from either PAL_VM_SUMMARY or
2870 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
2871 could support. nr_cpus=n : n >= 1 limits the kernel to
2872 support 'n' processors. It could be larger than the
2873 number of already plugged CPU during bootup, later in
2874 runtime you can physically add extra cpu until it reaches
2875 n. So during boot up some boot time memory for per-cpu
2876 variables need be pre-allocated for later physical cpu
2879 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
2881 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
2882 Allowed values are enable and disable
2884 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
2885 'node', 'default' can be specified
2886 This can be set from sysctl after boot.
2887 See Documentation/sysctl/vm.txt for details.
2889 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
2890 See Documentation/debugging-via-ohci1394.txt for more
2893 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
2894 Rather than timing out after 20 ms if an EC
2895 command is not properly ACKed, override the length
2896 of the timeout. We have interrupts disabled while
2897 waiting for the ACK, so if this is set too high
2898 interrupts *may* be lost!
2900 omap_mux= [OMAP] Override bootloader pin multiplexing.
2901 Format: <mux_mode0.mode_name=value>...
2902 For example, to override I2C bus2:
2903 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
2905 oprofile.timer= [HW]
2906 Use timer interrupt instead of performance counters
2908 oprofile.cpu_type= Force an oprofile cpu type
2909 This might be useful if you have an older oprofile
2910 userland or if you want common events.
2911 Format: { arch_perfmon }
2912 arch_perfmon: [X86] Force use of architectural
2913 perfmon on Intel CPUs instead of the
2914 CPU specific event set.
2915 timer: [X86] Force use of architectural NMI
2916 timer mode (see also oprofile.timer
2917 for generic hr timer mode)
2919 oops=panic Always panic on oopses. Default is to just kill the
2920 process, but there is a small probability of
2921 deadlocking the machine.
2922 This will also cause panics on machine check exceptions.
2923 Useful together with panic=30 to trigger a reboot.
2926 See Documentation/sound/oss/oss-parameters.txt
2928 page_owner= [KNL] Boot-time page_owner enabling option.
2929 Storage of the information about who allocated
2930 each page is disabled in default. With this switch,
2932 on: enable the feature
2934 page_poison= [KNL] Boot-time parameter changing the state of
2935 poisoning on the buddy allocator.
2936 off: turn off poisoning
2937 on: turn on poisoning
2939 panic= [KNL] Kernel behaviour on panic: delay <timeout>
2940 timeout > 0: seconds before rebooting
2941 timeout = 0: wait forever
2942 timeout < 0: reboot immediately
2945 panic_on_warn panic() instead of WARN(). Useful to cause kdump
2948 crash_kexec_post_notifiers
2949 Run kdump after running panic-notifiers and dumping
2950 kmsg. This only for the users who doubt kdump always
2951 succeeds in any situation.
2952 Note that this also increases risks of kdump failure,
2953 because some panic notifiers can make the crashed
2954 kernel more unstable.
2956 parkbd.port= [HW] Parallel port number the keyboard adapter is
2957 connected to, default is 0.
2959 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
2960 0 for XT, 1 for AT (default is AT).
2963 parport= [HW,PPT] Specify parallel ports. 0 disables.
2964 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
2965 Use 'auto' to force the driver to use any
2966 IRQ/DMA settings detected (the default is to
2967 ignore detected IRQ/DMA settings because of
2968 possible conflicts). You can specify the base
2969 address, IRQ, and DMA settings; IRQ and DMA
2970 should be numbers, or 'auto' (for using detected
2971 settings on that particular port), or 'nofifo'
2972 (to avoid using a FIFO even if it is detected).
2973 Parallel ports are assigned in the order they
2974 are specified on the command line, starting
2977 parport_init_mode= [HW,PPT]
2978 Configure VIA parallel port to operate in
2979 a specific mode. This is necessary on Pegasos
2980 computer where firmware has no options for setting
2981 up parallel port mode and sets it to spp.
2982 Currently this function knows 686a and 8231 chips.
2983 Format: [spp|ps2|epp|ecp|ecpepp]
2986 Halt all CPUs after the first oops has been printed for
2987 the specified number of seconds. This is to be used if
2988 your oopses keep scrolling off the screen.
2993 See header of drivers/block/paride/pcd.c.
2994 See also Documentation/blockdev/paride.txt.
2996 pci=option[,option...] [PCI] various PCI subsystem options:
2997 earlydump [X86] dump PCI config space before the kernel
2999 off [X86] don't probe for the PCI bus
3000 bios [X86-32] force use of PCI BIOS, don't access
3001 the hardware directly. Use this if your machine
3002 has a non-standard PCI host bridge.
3003 nobios [X86-32] disallow use of PCI BIOS, only direct
3004 hardware access methods are allowed. Use this
3005 if you experience crashes upon bootup and you
3006 suspect they are caused by the BIOS.
3007 conf1 [X86] Force use of PCI Configuration Access
3008 Mechanism 1 (config address in IO port 0xCF8,
3009 data in IO port 0xCFC, both 32-bit).
3010 conf2 [X86] Force use of PCI Configuration Access
3011 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3012 the function, IO port 0xCFA, also 8-bit, sets
3013 bus number. The config space is then accessed
3014 through ports 0xC000-0xCFFF).
3015 See http://wiki.osdev.org/PCI for more info
3016 on the configuration access mechanisms.
3017 noaer [PCIE] If the PCIEAER kernel config parameter is
3018 enabled, this kernel boot option can be used to
3019 disable the use of PCIE advanced error reporting.
3020 nodomains [PCI] Disable support for multiple PCI
3021 root domains (aka PCI segments, in ACPI-speak).
3022 nommconf [X86] Disable use of MMCONFIG for PCI
3024 check_enable_amd_mmconf [X86] check for and enable
3025 properly configured MMIO access to PCI
3026 config space on AMD family 10h CPU
3027 nomsi [MSI] If the PCI_MSI kernel config parameter is
3028 enabled, this kernel boot option can be used to
3029 disable the use of MSI interrupts system-wide.
3030 noioapicquirk [APIC] Disable all boot interrupt quirks.
3031 Safety option to keep boot IRQs enabled. This
3032 should never be necessary.
3033 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3034 primary IO-APIC for bridges that cannot disable
3035 boot IRQs. This fixes a source of spurious IRQs
3036 when the system masks IRQs.
3037 noioapicreroute [APIC] Disable workaround that uses the
3038 boot IRQ equivalent of an IRQ that connects to
3039 a chipset where boot IRQs cannot be disabled.
3040 The opposite of ioapicreroute.
3041 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3042 routing table. These calls are known to be buggy
3043 on several machines and they hang the machine
3044 when used, but on other computers it's the only
3045 way to get the interrupt routing table. Try
3046 this option if the kernel is unable to allocate
3047 IRQs or discover secondary PCI buses on your
3049 rom [X86] Assign address space to expansion ROMs.
3050 Use with caution as certain devices share
3051 address decoders between ROMs and other
3053 norom [X86] Do not assign address space to
3054 expansion ROMs that do not already have
3055 BIOS assigned address ranges.
3056 nobar [X86] Do not assign address space to the
3057 BARs that weren't assigned by the BIOS.
3058 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3059 assigned automatically to PCI devices. You can
3060 make the kernel exclude IRQs of your ISA cards
3062 pirqaddr=0xAAAAA [X86] Specify the physical address
3063 of the PIRQ table (normally generated
3064 by the BIOS) if it is outside the
3065 F0000h-100000h range.
3066 lastbus=N [X86] Scan all buses thru bus #N. Can be
3067 useful if the kernel is unable to find your
3068 secondary buses and you want to tell it
3069 explicitly which ones they are.
3070 assign-busses [X86] Always assign all PCI bus
3071 numbers ourselves, overriding
3072 whatever the firmware may have done.
3073 usepirqmask [X86] Honor the possible IRQ mask stored
3074 in the BIOS $PIR table. This is needed on
3075 some systems with broken BIOSes, notably
3076 some HP Pavilion N5400 and Omnibook XE3
3077 notebooks. This will have no effect if ACPI
3078 IRQ routing is enabled.
3079 noacpi [X86] Do not use ACPI for IRQ routing
3080 or for PCI scanning.
3081 use_crs [X86] Use PCI host bridge window information
3082 from ACPI. On BIOSes from 2008 or later, this
3083 is enabled by default. If you need to use this,
3084 please report a bug.
3085 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3086 If you need to use this, please report a bug.
3087 routeirq Do IRQ routing for all PCI devices.
3088 This is normally done in pci_enable_device(),
3089 so this option is a temporary workaround
3090 for broken drivers that don't call it.
3091 skip_isa_align [X86] do not align io start addr, so can
3092 handle more pci cards
3093 noearly [X86] Don't do any early type 1 scanning.
3094 This might help on some broken boards which
3095 machine check when some devices' config space
3096 is read. But various workarounds are disabled
3097 and some IOMMU drivers will not work.
3098 bfsort Sort PCI devices into breadth-first order.
3099 This sorting is done to get a device
3100 order compatible with older (<= 2.4) kernels.
3101 nobfsort Don't sort PCI devices into breadth-first order.
3102 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3103 tuning and use the BIOS-configured MPS defaults.
3104 pcie_bus_safe Set every device's MPS to the largest value
3105 supported by all devices below the root complex.
3106 pcie_bus_perf Set device MPS to the largest allowable MPS
3107 based on its parent bus. Also set MRRS (Max
3108 Read Request Size) to the largest supported
3109 value (no larger than the MPS that the device
3110 or bus can support) for best performance.
3111 pcie_bus_peer2peer Set every device's MPS to 128B, which
3112 every device is guaranteed to support. This
3113 configuration allows peer-to-peer DMA between
3114 any pair of devices, possibly at the cost of
3115 reduced performance. This also guarantees
3116 that hot-added devices will work.
3117 cbiosize=nn[KMG] The fixed amount of bus space which is
3118 reserved for the CardBus bridge's IO window.
3119 The default value is 256 bytes.
3120 cbmemsize=nn[KMG] The fixed amount of bus space which is
3121 reserved for the CardBus bridge's memory
3122 window. The default value is 64 megabytes.
3125 [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
3126 [<order of align>@]pci:<vendor>:<device>\
3127 [:<subvendor>:<subdevice>][; ...]
3128 Specifies alignment and device to reassign
3129 aligned memory resources.
3130 If <order of align> is not specified,
3131 PAGE_SIZE is used as alignment.
3132 PCI-PCI bridge can be specified, if resource
3133 windows need to be expanded.
3134 To specify the alignment for several
3135 instances of a device, the PCI vendor,
3136 device, subvendor, and subdevice may be
3137 specified, e.g., 4096@pci:8086:9c22:103c:198f
3138 ecrc= Enable/disable PCIe ECRC (transaction layer
3139 end-to-end CRC checking).
3140 bios: Use BIOS/firmware settings. This is the
3144 hpiosize=nn[KMG] The fixed amount of bus space which is
3145 reserved for hotplug bridge's IO window.
3146 Default size is 256 bytes.
3147 hpmemsize=nn[KMG] The fixed amount of bus space which is
3148 reserved for hotplug bridge's memory window.
3149 Default size is 2 megabytes.
3150 hpbussize=nn The minimum amount of additional bus numbers
3151 reserved for buses below a hotplug bridge.
3153 realloc= Enable/disable reallocating PCI bridge resources
3154 if allocations done by BIOS are too small to
3155 accommodate resources required by all child
3157 off: Turn realloc off
3159 realloc same as realloc=on
3160 noari do not use PCIe ARI.
3161 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3162 only look for one device below a PCIe downstream
3165 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3168 force Enable ASPM even on devices that claim not to support it.
3169 WARNING: Forcing ASPM on may cause system lockups.
3171 pcie_hp= [PCIE] PCI Express Hotplug driver options:
3172 nomsi Do not use MSI for PCI Express Native Hotplug (this
3173 makes all PCIe ports use INTx for hotplug services).
3175 pcie_ports= [PCIE] PCIe ports handling:
3176 auto Ask the BIOS whether or not to use native PCIe services
3177 associated with PCIe ports (PME, hot-plug, AER). Use
3178 them only if that is allowed by the BIOS.
3179 native Use native PCIe services associated with PCIe ports
3181 compat Treat PCIe ports as PCI-to-PCI bridges, disable the PCIe
3184 pcie_port_pm= [PCIE] PCIe port power management handling:
3185 off Disable power management of all PCIe ports
3186 force Forcibly enable power management of all PCIe ports
3188 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3189 nomsi Do not use MSI for native PCIe PME signaling (this makes
3190 all PCIe root ports use INTx for all services).
3192 pcmv= [HW,PCMCIA] BadgePAD 4
3196 Keep all power-domains already enabled by bootloader on,
3197 even if no driver has claimed them. This is useful
3198 for debug and development, but should not be
3199 needed on a platform with proper driver support.
3202 See Documentation/blockdev/paride.txt.
3204 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3207 See arch/parisc/kernel/pdc_chassis.c
3209 percpu_alloc= Select which percpu first chunk allocator to use.
3210 Currently supported values are "embed" and "page".
3211 Archs may support subset or none of the selections.
3212 See comments in mm/percpu.c for details on each
3213 allocator. This parameter is primarily for debugging
3214 and performance comparison.
3217 See Documentation/blockdev/paride.txt.
3220 See Documentation/blockdev/paride.txt.
3222 pirq= [SMP,APIC] Manual mp-table setup
3223 See Documentation/x86/i386/IO-APIC.txt.
3225 plip= [PPT,NET] Parallel port network link
3226 Format: { parport<nr> | timid | 0 }
3227 See also Documentation/parport.txt.
3229 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3230 Override pmtimer IOPort with a hex value.
3234 Enable PNP debug messages (depends on the
3235 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3236 via /sys/module/pnp/parameters/debug. We always show
3237 current resource usage; turning this on also shows
3238 possible settings and some assignment information.
3244 { on | off | curr | res | no-curr | no-res }
3247 [ISAPNP] Exclude IRQs for the autoconfiguration
3250 [ISAPNP] Exclude DMAs for the autoconfiguration
3252 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3253 Ranges are in pairs (I/O port base and size).
3256 [ISAPNP] Exclude memory regions for the
3258 Ranges are in pairs (memory base and size).
3260 ports= [IP_VS_FTP] IPVS ftp helper module
3262 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3264 Format: <port>,<port>....
3266 powersave=off [PPC] This option disables power saving features.
3267 It specifically disables cpuidle and sets the
3268 platform machine description specific power_save
3269 function to NULL. On Idle the CPU just reduces
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.
3278 print-fatal-signals=
3279 [KNL] debug: print fatal signals
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
3286 If you hit the warning due to signal overflow,
3287 you might want to try "ulimit -i unlimited".
3291 printk.always_kmsg_dump=
3292 Trigger kmsg_dump for cases other than kernel oops or
3294 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
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
3304 printk.time= Show timing data prefixed to each printk message line
3305 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3307 processor.max_cstate= [HW,ACPI]
3308 Limit processor to maximum C-state
3309 max_cstate=9 overrides any DMI blacklist limit.
3311 processor.nocst [HW,ACPI]
3312 Ignore the _CST method to determine C-states,
3313 instead using the legacy FADT method
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.
3324 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3326 See Documentation/blockdev/ramdisk.txt.
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
3332 psmouse.resetafter= [HW,MOUSE]
3333 Try to reset the device after so many bad packets
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).
3341 pstore.backend= Specify the name of the pstore backend to use
3344 See Documentation/blockdev/paride.txt.
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.
3351 on - unconditionally enable
3352 off - unconditionally disable
3353 auto - kernel detects whether your CPU model is
3354 vulnerable to issues that PTI mitigates
3356 Not specifying this option is equivalent to pti=auto.
3359 Equivalent to pti=off
3362 [KNL] Number of legacy pty's. Overwrites compiled-in
3365 quiet [KNL] Disable most log messages
3370 See Documentation/admin-guide/md.rst.
3372 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3373 See Documentation/blockdev/ramdisk.txt.
3375 ras=option[,option,...] [KNL] RAS-specific options
3378 Disable the Correctable Errors Collector,
3379 see CONFIG_RAS_CEC help text.
3382 The argument is a cpu list, as described above.
3384 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3385 the specified list of CPUs to be no-callback CPUs.
3386 Invocation of these CPUs' RCU callbacks will
3387 be offloaded to "rcuox/N" kthreads created for
3388 that purpose, where "x" is "b" for RCU-bh, "p"
3389 for RCU-preempt, and "s" for RCU-sched, and "N"
3390 is the CPU number. This reduces OS jitter on the
3391 offloaded CPUs, which can be useful for HPC and
3392 real-time workloads. It can also improve energy
3393 efficiency for asymmetric multiprocessors.
3396 Rather than requiring that offloaded CPUs
3397 (specified by rcu_nocbs= above) explicitly
3398 awaken the corresponding "rcuoN" kthreads,
3399 make these kthreads poll for callbacks.
3400 This improves the real-time response for the
3401 offloaded CPUs by relieving them of the need to
3402 wake up the corresponding kthread, but degrades
3403 energy efficiency by requiring that the kthreads
3404 periodically wake up to do the polling.
3406 rcutree.blimit= [KNL]
3407 Set maximum number of finished RCU callbacks to
3408 process in one batch.
3410 rcutree.dump_tree= [KNL]
3411 Dump the structure of the rcu_node combining tree
3412 out at early boot. This is used for diagnostic
3413 purposes, to verify correct tree setup.
3415 rcutree.gp_cleanup_delay= [KNL]
3416 Set the number of jiffies to delay each step of
3417 RCU grace-period cleanup.
3419 rcutree.gp_init_delay= [KNL]
3420 Set the number of jiffies to delay each step of
3421 RCU grace-period initialization.
3423 rcutree.gp_preinit_delay= [KNL]
3424 Set the number of jiffies to delay each step of
3425 RCU grace-period pre-initialization, that is,
3426 the propagation of recent CPU-hotplug changes up
3427 the rcu_node combining tree.
3429 rcutree.rcu_fanout_exact= [KNL]
3430 Disable autobalancing of the rcu_node combining
3431 tree. This is used by rcutorture, and might
3432 possibly be useful for architectures having high
3433 cache-to-cache transfer latencies.
3435 rcutree.rcu_fanout_leaf= [KNL]
3436 Change the number of CPUs assigned to each
3437 leaf rcu_node structure. Useful for very
3438 large systems, which will choose the value 64,
3439 and for NUMA systems with large remote-access
3440 latencies, which will choose a value aligned
3441 with the appropriate hardware boundaries.
3443 rcutree.jiffies_till_sched_qs= [KNL]
3444 Set required age in jiffies for a
3445 given grace period before RCU starts
3446 soliciting quiescent-state help from
3447 rcu_note_context_switch().
3449 rcutree.jiffies_till_first_fqs= [KNL]
3450 Set delay from grace-period initialization to
3451 first attempt to force quiescent states.
3452 Units are jiffies, minimum value is zero,
3453 and maximum value is HZ.
3455 rcutree.jiffies_till_next_fqs= [KNL]
3456 Set delay between subsequent attempts to force
3457 quiescent states. Units are jiffies, minimum
3458 value is one, and maximum value is HZ.
3460 rcutree.kthread_prio= [KNL,BOOT]
3461 Set the SCHED_FIFO priority of the RCU per-CPU
3462 kthreads (rcuc/N). This value is also used for
3463 the priority of the RCU boost threads (rcub/N)
3464 and for the RCU grace-period kthreads (rcu_bh,
3465 rcu_preempt, and rcu_sched). If RCU_BOOST is
3466 set, valid values are 1-99 and the default is 1
3467 (the least-favored priority). Otherwise, when
3468 RCU_BOOST is not set, valid values are 0-99 and
3469 the default is zero (non-realtime operation).
3471 rcutree.rcu_nocb_leader_stride= [KNL]
3472 Set the number of NOCB kthread groups, which
3473 defaults to the square root of the number of
3474 CPUs. Larger numbers reduces the wakeup overhead
3475 on the per-CPU grace-period kthreads, but increases
3476 that same overhead on each group's leader.
3478 rcutree.qhimark= [KNL]
3479 Set threshold of queued RCU callbacks beyond which
3480 batch limiting is disabled.
3482 rcutree.qlowmark= [KNL]
3483 Set threshold of queued RCU callbacks below which
3484 batch limiting is re-enabled.
3486 rcutree.rcu_idle_gp_delay= [KNL]
3487 Set wakeup interval for idle CPUs that have
3488 RCU callbacks (RCU_FAST_NO_HZ=y).
3490 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3491 Set wakeup interval for idle CPUs that have
3492 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3493 Lazy RCU callbacks are those which RCU can
3494 prove do nothing more than free memory.
3496 rcutree.rcu_kick_kthreads= [KNL]
3497 Cause the grace-period kthread to get an extra
3498 wake_up() if it sleeps three times longer than
3499 it should at force-quiescent-state time.
3500 This wake_up() will be accompanied by a
3501 WARN_ONCE() splat and an ftrace_dump().
3503 rcuperf.gp_async= [KNL]
3504 Measure performance of asynchronous
3505 grace-period primitives such as call_rcu().
3507 rcuperf.gp_async_max= [KNL]
3508 Specify the maximum number of outstanding
3509 callbacks per writer thread. When a writer
3510 thread exceeds this limit, it invokes the
3511 corresponding flavor of rcu_barrier() to allow
3512 previously posted callbacks to drain.
3514 rcuperf.gp_exp= [KNL]
3515 Measure performance of expedited synchronous
3516 grace-period primitives.
3518 rcuperf.holdoff= [KNL]
3519 Set test-start holdoff period. The purpose of
3520 this parameter is to delay the start of the
3521 test until boot completes in order to avoid
3524 rcuperf.nreaders= [KNL]
3525 Set number of RCU readers. The value -1 selects
3526 N, where N is the number of CPUs. A value
3527 "n" less than -1 selects N-n+1, where N is again
3528 the number of CPUs. For example, -2 selects N
3529 (the number of CPUs), -3 selects N+1, and so on.
3530 A value of "n" less than or equal to -N selects
3533 rcuperf.nwriters= [KNL]
3534 Set number of RCU writers. The values operate
3535 the same as for rcuperf.nreaders.
3536 N, where N is the number of CPUs
3538 rcuperf.perf_runnable= [BOOT]
3539 Start rcuperf running at boot time.
3541 rcuperf.perf_type= [KNL]
3542 Specify the RCU implementation to test.
3544 rcuperf.shutdown= [KNL]
3545 Shut the system down after performance tests
3546 complete. This is useful for hands-off automated
3549 rcuperf.verbose= [KNL]
3550 Enable additional printk() statements.
3552 rcuperf.writer_holdoff= [KNL]
3553 Write-side holdoff between grace periods,
3554 in microseconds. The default of zero says
3557 rcutorture.cbflood_inter_holdoff= [KNL]
3558 Set holdoff time (jiffies) between successive
3559 callback-flood tests.
3561 rcutorture.cbflood_intra_holdoff= [KNL]
3562 Set holdoff time (jiffies) between successive
3563 bursts of callbacks within a given callback-flood
3566 rcutorture.cbflood_n_burst= [KNL]
3567 Set the number of bursts making up a given
3568 callback-flood test. Set this to zero to
3569 disable callback-flood testing.
3571 rcutorture.cbflood_n_per_burst= [KNL]
3572 Set the number of callbacks to be registered
3573 in a given burst of a callback-flood test.
3575 rcutorture.fqs_duration= [KNL]
3576 Set duration of force_quiescent_state bursts
3579 rcutorture.fqs_holdoff= [KNL]
3580 Set holdoff time within force_quiescent_state bursts
3583 rcutorture.fqs_stutter= [KNL]
3584 Set wait time between force_quiescent_state bursts
3587 rcutorture.gp_cond= [KNL]
3588 Use conditional/asynchronous update-side
3589 primitives, if available.
3591 rcutorture.gp_exp= [KNL]
3592 Use expedited update-side primitives, if available.
3594 rcutorture.gp_normal= [KNL]
3595 Use normal (non-expedited) asynchronous
3596 update-side primitives, if available.
3598 rcutorture.gp_sync= [KNL]
3599 Use normal (non-expedited) synchronous
3600 update-side primitives, if available. If all
3601 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3602 rcutorture.gp_normal=, and rcutorture.gp_sync=
3603 are zero, rcutorture acts as if is interpreted
3604 they are all non-zero.
3606 rcutorture.n_barrier_cbs= [KNL]
3607 Set callbacks/threads for rcu_barrier() testing.
3609 rcutorture.nfakewriters= [KNL]
3610 Set number of concurrent RCU writers. These just
3611 stress RCU, they don't participate in the actual
3612 test, hence the "fake".
3614 rcutorture.nreaders= [KNL]
3615 Set number of RCU readers. The value -1 selects
3616 N-1, where N is the number of CPUs. A value
3617 "n" less than -1 selects N-n-2, where N is again
3618 the number of CPUs. For example, -2 selects N
3619 (the number of CPUs), -3 selects N+1, and so on.
3621 rcutorture.object_debug= [KNL]
3622 Enable debug-object double-call_rcu() testing.
3624 rcutorture.onoff_holdoff= [KNL]
3625 Set time (s) after boot for CPU-hotplug testing.
3627 rcutorture.onoff_interval= [KNL]
3628 Set time (s) between CPU-hotplug operations, or
3629 zero to disable CPU-hotplug testing.
3631 rcutorture.shuffle_interval= [KNL]
3632 Set task-shuffle interval (s). Shuffling tasks
3633 allows some CPUs to go into dyntick-idle mode
3634 during the rcutorture test.
3636 rcutorture.shutdown_secs= [KNL]
3637 Set time (s) after boot system shutdown. This
3638 is useful for hands-off automated testing.
3640 rcutorture.stall_cpu= [KNL]
3641 Duration of CPU stall (s) to test RCU CPU stall
3642 warnings, zero to disable.
3644 rcutorture.stall_cpu_holdoff= [KNL]
3645 Time to wait (s) after boot before inducing stall.
3647 rcutorture.stat_interval= [KNL]
3648 Time (s) between statistics printk()s.
3650 rcutorture.stutter= [KNL]
3651 Time (s) to stutter testing, for example, specifying
3652 five seconds causes the test to run for five seconds,
3653 wait for five seconds, and so on. This tests RCU's
3654 ability to transition abruptly to and from idle.
3656 rcutorture.test_boost= [KNL]
3657 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3658 "Maybe" means test if the RCU implementation
3659 under test support RCU priority boosting.
3661 rcutorture.test_boost_duration= [KNL]
3662 Duration (s) of each individual boost test.
3664 rcutorture.test_boost_interval= [KNL]
3665 Interval (s) between each boost test.
3667 rcutorture.test_no_idle_hz= [KNL]
3668 Test RCU's dyntick-idle handling. See also the
3669 rcutorture.shuffle_interval parameter.
3671 rcutorture.torture_runnable= [BOOT]
3672 Start rcutorture running at boot time.
3674 rcutorture.torture_type= [KNL]
3675 Specify the RCU implementation to test.
3677 rcutorture.verbose= [KNL]
3678 Enable additional printk() statements.
3680 rcupdate.rcu_cpu_stall_suppress= [KNL]
3681 Suppress RCU CPU stall warning messages.
3683 rcupdate.rcu_cpu_stall_timeout= [KNL]
3684 Set timeout for RCU CPU stall warning messages.
3686 rcupdate.rcu_expedited= [KNL]
3687 Use expedited grace-period primitives, for
3688 example, synchronize_rcu_expedited() instead
3689 of synchronize_rcu(). This reduces latency,
3690 but can increase CPU utilization, degrade
3691 real-time latency, and degrade energy efficiency.
3692 No effect on CONFIG_TINY_RCU kernels.
3694 rcupdate.rcu_normal= [KNL]
3695 Use only normal grace-period primitives,
3696 for example, synchronize_rcu() instead of
3697 synchronize_rcu_expedited(). This improves
3698 real-time latency, CPU utilization, and
3699 energy efficiency, but can expose users to
3700 increased grace-period latency. This parameter
3701 overrides rcupdate.rcu_expedited. No effect on
3702 CONFIG_TINY_RCU kernels.
3704 rcupdate.rcu_normal_after_boot= [KNL]
3705 Once boot has completed (that is, after
3706 rcu_end_inkernel_boot() has been invoked), use
3707 only normal grace-period primitives. No effect
3708 on CONFIG_TINY_RCU kernels.
3710 rcupdate.rcu_task_stall_timeout= [KNL]
3711 Set timeout in jiffies for RCU task stall warning
3712 messages. Disable with a value less than or equal
3715 rcupdate.rcu_self_test= [KNL]
3716 Run the RCU early boot self tests
3718 rcupdate.rcu_self_test_bh= [KNL]
3719 Run the RCU bh early boot self tests
3721 rcupdate.rcu_self_test_sched= [KNL]
3722 Run the RCU sched early boot self tests
3726 Run specified binary instead of /init from the ramdisk,
3727 used for early userspace startup. See initrd.
3730 Turn on/off individual RDT features. List is:
3731 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, mba.
3732 E.g. to turn on cmt and turn off mba use:
3736 Format (x86 or x86_64):
3737 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
3739 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
3741 Where reboot_mode is one of warm (soft) or cold (hard) or gpio,
3742 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
3743 reboot_force is either force or not specified,
3744 reboot_cpu is s[mp]#### with #### being the processor
3745 to be used for rebooting.
3748 [KNL, SMP] Set scheduler's default relax_domain_level.
3749 See Documentation/cgroup-v1/cpusets.txt.
3751 reserve= [KNL,BUGS] Force the kernel to ignore some iomem area
3753 reservetop= [X86-32]
3755 Reserves a hole at the top of the kernel virtual
3760 Set the amount of memory to reserve for BIOS at
3761 the bottom of the address space.
3763 reset_devices [KNL] Force drivers to reset the underlying device
3764 during initialization.
3767 Specify the partition device for software suspend
3769 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
3771 resume_offset= [SWSUSP]
3772 Specify the offset from the beginning of the partition
3773 given by "resume=" at which the swap header is located,
3774 in <PAGE_SIZE> units (needed only for swap files).
3775 See Documentation/power/swsusp-and-swap-files.txt
3777 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
3778 read the resume files
3780 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
3781 Useful for devices that are detected asynchronously
3782 (e.g. USB and MMC devices).
3784 hibernate= [HIBERNATION]
3785 noresume Don't check if there's a hibernation image
3786 present during boot.
3787 nocompress Don't compress/decompress hibernation images.
3788 no Disable hibernation and resume.
3789 protect_image Turn on image protection during restoration
3790 (that will set all pages holding image data
3791 during restoration read-only).
3793 retain_initrd [RAM] Keep initrd memory after extraction
3795 rfkill.default_state=
3796 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
3797 etc. communication is blocked by default.
3800 rfkill.master_switch_mode=
3801 0 The "airplane mode" button does nothing.
3802 1 The "airplane mode" button toggles between everything
3803 blocked and the previous configuration.
3804 2 The "airplane mode" button toggles between everything
3805 blocked and everything unblocked.
3807 rhash_entries= [KNL,NET]
3808 Set number of hash buckets for route cache
3811 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
3814 ro [KNL] Mount root device read-only on boot
3817 on Mark read-only kernel memory as read-only (default).
3818 off Leave read-only kernel memory writable for debugging.
3821 Enable the uart passthrough on the designated usb port
3822 on Rockchip SoCs. When active, the signals of the
3823 debug-uart get routed to the D+ and D- pins of the usb
3824 port and the regular usb controller gets disabled.
3826 root= [KNL] Root filesystem
3827 See name_to_dev_t comment in init/do_mounts.c.
3829 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
3830 mount the root filesystem
3832 rootflags= [KNL] Set root filesystem mount option string
3834 rootfstype= [KNL] Set root filesystem type
3836 rootwait [KNL] Wait (indefinitely) for root device to show up.
3837 Useful for devices that are detected asynchronously
3838 (e.g. USB and MMC devices).
3840 rproc_mem=nn[KMG][@address]
3841 [KNL,ARM,CMA] Remoteproc physical memory block.
3842 Memory area to be used by remote processor image,
3845 rw [KNL] Mount root device read-write on boot
3847 S [KNL] Run init in single mode
3849 s390_iommu= [HW,S390]
3850 Set s390 IOTLB flushing mode
3852 With strict flushing every unmap operation will result in
3853 an IOTLB flush. Default is lazy flushing before reuse,
3857 See drivers/net/irda/sa1100_ir.c.
3859 sbni= [NET] Granch SBNI12 leased line adapter
3861 sched_debug [KNL] Enables verbose scheduler debug messages.
3863 schedstats= [KNL,X86] Enable or disable scheduled statistics.
3864 Allowed values are enable and disable. This feature
3865 incurs a small amount of overhead in the scheduler
3866 but is useful for debugging and performance tuning.
3868 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
3869 xtime_lock contention on larger systems, and/or RCU lock
3870 contention on all systems with CONFIG_MAXSMP set.
3871 Format: { "0" | "1" }
3872 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
3874 Note: increases power consumption, thus should only be
3875 enabled if running jitter sensitive (HPC/RT) workloads.
3877 security= [SECURITY] Choose a security module to enable at boot.
3878 If this boot parameter is not specified, only the first
3879 security module asking for security registration will be
3880 loaded. An invalid security module name will be treated
3881 as if no module has been chosen.
3883 selinux= [SELINUX] Disable or enable SELinux at boot time.
3884 Format: { "0" | "1" }
3885 See security/selinux/Kconfig help text.
3888 Default value is set via kernel config option.
3889 If enabled at boot time, /selinux/disable can be used
3890 later to disable prior to initial policy load.
3892 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
3893 Format: { "0" | "1" }
3894 See security/apparmor/Kconfig help text
3897 Default value is set via kernel config option.
3899 serialnumber [BUGS=X86-32]
3902 Maximal number of shapers.
3910 Disable merging of slabs with similar size. May be
3911 necessary if there is some reason to distinguish
3912 allocs to different slabs, especially in hardened
3913 environments where the risk of heap overflows and
3914 layout control by attackers can usually be
3915 frustrated by disabling merging. This will reduce
3916 most of the exposure of a heap attack to a single
3917 cache (risks via metadata attacks are mostly
3918 unchanged). Debug options disable merging on their
3920 For more information see Documentation/vm/slub.txt.
3922 slab_max_order= [MM, SLAB]
3923 Determines the maximum allowed order for slabs.
3924 A high setting may cause OOMs due to memory
3925 fragmentation. Defaults to 1 for systems with
3926 more than 32MB of RAM, 0 otherwise.
3928 slub_debug[=options[,slabs]] [MM, SLUB]
3929 Enabling slub_debug allows one to determine the
3930 culprit if slab objects become corrupted. Enabling
3931 slub_debug can create guard zones around objects and
3932 may poison objects when not in use. Also tracks the
3933 last alloc / free. For more information see
3934 Documentation/vm/slub.txt.
3936 slub_memcg_sysfs= [MM, SLUB]
3937 Determines whether to enable sysfs directories for
3938 memory cgroup sub-caches. 1 to enable, 0 to disable.
3939 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
3940 Enabling this can lead to a very high number of debug
3941 directories and files being created under
3944 slub_max_order= [MM, SLUB]
3945 Determines the maximum allowed order for slabs.
3946 A high setting may cause OOMs due to memory
3947 fragmentation. For more information see
3948 Documentation/vm/slub.txt.
3950 slub_min_objects= [MM, SLUB]
3951 The minimum number of objects per slab. SLUB will
3952 increase the slab order up to slub_max_order to
3953 generate a sufficiently large slab able to contain
3954 the number of objects indicated. The higher the number
3955 of objects the smaller the overhead of tracking slabs
3956 and the less frequently locks need to be acquired.
3957 For more information see Documentation/vm/slub.txt.
3959 slub_min_order= [MM, SLUB]
3960 Determines the minimum page order for slabs. Must be
3961 lower than slub_max_order.
3962 For more information see Documentation/vm/slub.txt.
3964 slub_nomerge [MM, SLUB]
3965 Same with slab_nomerge. This is supported for legacy.
3966 See slab_nomerge for more information.
3969 Format: <io1>[,<io2>[,...,<io8>]]
3971 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
3972 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
3973 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
3974 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
3975 smsc-ircc2.ircc_irq= [HW] IRQ line
3976 smsc-ircc2.ircc_dma= [HW] DMA channel
3977 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
3978 0: Toshiba Satellite 1800 (GP data pin select)
3979 1: Fast pin select (default)
3982 smt [KNL,S390] Set the maximum number of threads (logical
3983 CPUs) to use per physical CPU on systems capable of
3984 symmetric multithreading (SMT). Will be capped to the
3985 actual hardware limit.
3987 Default: -1 (no limit)
3990 [KNL] Should the soft-lockup detector generate panics.
3993 softlockup_all_cpu_backtrace=
3994 [KNL] Should the soft-lockup detector generate
3995 backtraces on all cpus.
3998 sonypi.*= [HW] Sony Programmable I/O Control Device driver
3999 See Documentation/laptops/sonypi.txt
4001 spectre_v2= [X86] Control mitigation of Spectre variant 2
4002 (indirect branch speculation) vulnerability.
4003 The default operation protects the kernel from
4006 on - unconditionally enable, implies
4008 off - unconditionally disable, implies
4010 auto - kernel detects whether your CPU model is
4013 Selecting 'on' will, and 'auto' may, choose a
4014 mitigation method at run time according to the
4015 CPU, the available microcode, the setting of the
4016 CONFIG_RETPOLINE configuration option, and the
4017 compiler with which the kernel was built.
4019 Selecting 'on' will also enable the mitigation
4020 against user space to user space task attacks.
4022 Selecting 'off' will disable both the kernel and
4023 the user space protections.
4025 Specific mitigations can also be selected manually:
4027 retpoline - replace indirect branches
4028 retpoline,generic - google's original retpoline
4029 retpoline,amd - AMD-specific minimal thunk
4031 Not specifying this option is equivalent to
4035 [X86] Control mitigation of Spectre variant 2
4036 (indirect branch speculation) vulnerability between
4039 on - Unconditionally enable mitigations. Is
4040 enforced by spectre_v2=on
4042 off - Unconditionally disable mitigations. Is
4043 enforced by spectre_v2=off
4045 prctl - Indirect branch speculation is enabled,
4046 but mitigation can be enabled via prctl
4047 per thread. The mitigation control state
4048 is inherited on fork.
4051 - Like "prctl" above, but only STIBP is
4052 controlled per thread. IBPB is issued
4053 always when switching between different user
4057 - Same as "prctl" above, but all seccomp
4058 threads will enable the mitigation unless
4059 they explicitly opt out.
4062 - Like "seccomp" above, but only STIBP is
4063 controlled per thread. IBPB is issued
4064 always when switching between different
4065 user space processes.
4067 auto - Kernel selects the mitigation depending on
4068 the available CPU features and vulnerability.
4071 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4073 Not specifying this option is equivalent to
4074 spectre_v2_user=auto.
4076 spec_store_bypass_disable=
4077 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4078 (Speculative Store Bypass vulnerability)
4080 Certain CPUs are vulnerable to an exploit against a
4081 a common industry wide performance optimization known
4082 as "Speculative Store Bypass" in which recent stores
4083 to the same memory location may not be observed by
4084 later loads during speculative execution. The idea
4085 is that such stores are unlikely and that they can
4086 be detected prior to instruction retirement at the
4087 end of a particular speculation execution window.
4089 In vulnerable processors, the speculatively forwarded
4090 store can be used in a cache side channel attack, for
4091 example to read memory to which the attacker does not
4092 directly have access (e.g. inside sandboxed code).
4094 This parameter controls whether the Speculative Store
4095 Bypass optimization is used.
4097 on - Unconditionally disable Speculative Store Bypass
4098 off - Unconditionally enable Speculative Store Bypass
4099 auto - Kernel detects whether the CPU model contains an
4100 implementation of Speculative Store Bypass and
4101 picks the most appropriate mitigation. If the
4102 CPU is not vulnerable, "off" is selected. If the
4103 CPU is vulnerable the default mitigation is
4104 architecture and Kconfig dependent. See below.
4105 prctl - Control Speculative Store Bypass per thread
4106 via prctl. Speculative Store Bypass is enabled
4107 for a process by default. The state of the control
4108 is inherited on fork.
4109 seccomp - Same as "prctl" above, but all seccomp threads
4110 will disable SSB unless they explicitly opt out.
4112 Not specifying this option is equivalent to
4113 spec_store_bypass_disable=auto.
4115 Default mitigations:
4116 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4118 spia_io_base= [HW,MTD]
4123 srcutree.counter_wrap_check [KNL]
4124 Specifies how frequently to check for
4125 grace-period sequence counter wrap for the
4126 srcu_data structure's ->srcu_gp_seq_needed field.
4127 The greater the number of bits set in this kernel
4128 parameter, the less frequently counter wrap will
4129 be checked for. Note that the bottom two bits
4132 srcutree.exp_holdoff [KNL]
4133 Specifies how many nanoseconds must elapse
4134 since the end of the last SRCU grace period for
4135 a given srcu_struct until the next normal SRCU
4136 grace period will be considered for automatic
4137 expediting. Set to zero to disable automatic
4141 Speculative Store Bypass Disable control
4143 On CPUs that are vulnerable to the Speculative
4144 Store Bypass vulnerability and offer a
4145 firmware based mitigation, this parameter
4146 indicates how the mitigation should be used:
4148 force-on: Unconditionally enable mitigation for
4149 for both kernel and userspace
4150 force-off: Unconditionally disable mitigation for
4151 for both kernel and userspace
4152 kernel: Always enable mitigation in the
4153 kernel, and offer a prctl interface
4154 to allow userspace to register its
4155 interest in being mitigated too.
4157 stack_guard_gap= [MM]
4158 override the default stack gap protection. The value
4159 is in page units and it defines how many pages prior
4160 to (for stacks growing down) resp. after (for stacks
4161 growing up) the main stack are reserved for no other
4162 mapping. Default value is 256 pages.
4165 Enabled the stack tracer on boot up.
4167 stacktrace_filter=[function-list]
4168 [FTRACE] Limit the functions that the stack tracer
4169 will trace at boot up. function-list is a comma separated
4170 list of functions. This list can be changed at run
4171 time by the stack_trace_filter file in the debugfs
4172 tracing directory. Note, this enables stack tracing
4173 and the stacktrace above is not needed.
4177 Set the STI (builtin display/keyboard on the HP-PARISC
4178 machines) console (graphic card) which should be used
4179 as the initial boot-console.
4180 See also comment in drivers/video/console/sticore.c.
4183 See comment in drivers/video/console/sticore.c.
4186 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4188 sunrpc.min_resvport=
4189 sunrpc.max_resvport=
4191 SunRPC servers often require that client requests
4192 originate from a privileged port (i.e. a port in the
4193 range 0 < portnr < 1024).
4194 An administrator who wishes to reserve some of these
4195 ports for other uses may adjust the range that the
4196 kernel's sunrpc client considers to be privileged
4197 using these two parameters to set the minimum and
4198 maximum port values.
4200 sunrpc.svc_rpc_per_connection_limit=
4202 Limit the number of requests that the server will
4203 process in parallel from a single connection.
4204 The default value is 0 (no limit).
4208 Control how the NFS server code allocates CPUs to
4209 service thread pools. Depending on how many NICs
4210 you have and where their interrupts are bound, this
4211 option will affect which CPUs will do NFS serving.
4212 Note: this parameter cannot be changed while the
4213 NFS server is running.
4215 auto the server chooses an appropriate mode
4216 automatically using heuristics
4217 global a single global pool contains all CPUs
4218 percpu one pool for each CPU
4219 pernode one pool for each NUMA node (equivalent
4220 to global on non-NUMA machines)
4222 sunrpc.tcp_slot_table_entries=
4223 sunrpc.udp_slot_table_entries=
4225 Sets the upper limit on the number of simultaneous
4226 RPC calls that can be sent from the client to a
4227 server. Increasing these values may allow you to
4228 improve throughput, but will also increase the
4229 amount of memory reserved for use by the client.
4231 suspend.pm_test_delay=
4233 Sets the number of seconds to remain in a suspend test
4234 mode before resuming the system (see
4235 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4236 is set. Default value is 5.
4239 [KNL] Enable accounting of swap in memory resource
4240 controller if no parameter or 1 is given or disable
4241 it if 0 is given (See Documentation/cgroup-v1/memory.txt)
4243 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4244 Format: { <int> | force | noforce }
4245 <int> -- Number of I/O TLB slabs
4246 force -- force using of bounce buffers even if they
4247 wouldn't be automatically used by the kernel
4248 noforce -- Never use bounce buffers (for debugging)
4252 sysfs.deprecated=0|1 [KNL]
4253 Enable/disable old style sysfs layout for old udev
4254 on older distributions. When this option is enabled
4255 very new udev will not work anymore. When this option
4256 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4257 in older udev will not work anymore.
4258 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4259 the kernel configuration.
4261 sysrq_always_enabled
4263 Ignore sysrq setting - this boot parameter will
4264 neutralize any effect of /proc/sys/kernel/sysrq.
4265 Useful for debugging.
4267 tcpmhash_entries= [KNL,NET]
4268 Set the number of tcp_metrics_hash slots.
4269 Default value is 8192 or 16384 depending on total
4270 ram pages. This is used to specify the TCP metrics
4271 cache size. See Documentation/networking/ip-sysctl.txt
4272 "tcp_no_metrics_save" section for more details.
4276 test_suspend= [SUSPEND][,N]
4277 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4278 standby suspend) or "freeze" (for suspend type freeze)
4279 as the system sleep state during system startup with
4280 the optional capability to repeat N number of times.
4281 The system is woken from this state using a
4282 wakeup-capable RTC alarm.
4284 thash_entries= [KNL,NET]
4285 Set number of hash buckets for TCP connection
4287 thermal.act= [HW,ACPI]
4288 -1: disable all active trip points in all thermal zones
4289 <degrees C>: override all lowest active trip points
4291 thermal.crt= [HW,ACPI]
4292 -1: disable all critical trip points in all thermal zones
4293 <degrees C>: override all critical trip points
4295 thermal.nocrt= [HW,ACPI]
4296 Set to disable actions on ACPI thermal zone
4297 critical and hot trip points.
4299 thermal.off= [HW,ACPI]
4300 1: disable ACPI thermal control
4302 thermal.psv= [HW,ACPI]
4303 -1: disable all passive trip points
4304 <degrees C>: override all passive trip points to this
4307 thermal.tzp= [HW,ACPI]
4308 Specify global default ACPI thermal zone polling rate
4309 <deci-seconds>: poll all this frequency
4310 0: no polling (default)
4313 Force threading of all interrupt handlers except those
4314 marked explicitly IRQF_NO_THREAD.
4317 Enable the Transcendent memory driver if built-in.
4319 tmem.cleancache=0|1 [KNL, XEN]
4320 Default is on (1). Disable the usage of the cleancache
4321 API to send anonymous pages to the hypervisor.
4323 tmem.frontswap=0|1 [KNL, XEN]
4324 Default is on (1). Disable the usage of the frontswap
4325 API to send swap pages to the hypervisor. If disabled
4326 the selfballooning and selfshrinking are force disabled.
4328 tmem.selfballooning=0|1 [KNL, XEN]
4329 Default is on (1). Disable the driving of swap pages
4332 tmem.selfshrinking=0|1 [KNL, XEN]
4333 Default is on (1). Partial swapoff that immediately
4334 transfers pages from Xen hypervisor back to the
4335 kernel based on different criteria.
4339 Specify if the kernel should make use of the cpu
4340 topology information if the hardware supports this.
4341 The scheduler will make use of this information and
4342 e.g. base its process migration decisions on it.
4345 topology_updates= [KNL, PPC, NUMA]
4347 Specify if the kernel should ignore (off)
4348 topology updates sent by the hypervisor to this
4353 tpm_suspend_pcr=[HW,TPM]
4354 Format: integer pcr id
4355 Specify that at suspend time, the tpm driver
4356 should extend the specified pcr with zeros,
4357 as a workaround for some chips which fail to
4358 flush the last written pcr on TPM_SaveState.
4359 This will guarantee that all the other pcrs
4362 trace_buf_size=nn[KMG]
4363 [FTRACE] will set tracing buffer size on each cpu.
4365 trace_event=[event-list]
4366 [FTRACE] Set and start specified trace events in order
4367 to facilitate early boot debugging. The event-list is a
4368 comma separated list of trace events to enable. See
4369 also Documentation/trace/events.txt
4371 trace_options=[option-list]
4372 [FTRACE] Enable or disable tracer options at boot.
4373 The option-list is a comma delimited list of options
4374 that can be enabled or disabled just as if you were
4375 to echo the option name into
4377 /sys/kernel/debug/tracing/trace_options
4379 For example, to enable stacktrace option (to dump the
4380 stack trace of each event), add to the command line:
4382 trace_options=stacktrace
4384 See also Documentation/trace/ftrace.txt "trace options"
4388 Have the tracepoints sent to printk as well as the
4389 tracing ring buffer. This is useful for early boot up
4390 where the system hangs or reboots and does not give the
4391 option for reading the tracing buffer or performing a
4392 ftrace_dump_on_oops.
4394 To turn off having tracepoints sent to printk,
4395 echo 0 > /proc/sys/kernel/tracepoint_printk
4396 Note, echoing 1 into this file without the
4397 tracepoint_printk kernel cmdline option has no effect.
4401 Having tracepoints sent to printk() and activating high
4402 frequency tracepoints such as irq or sched, can cause
4403 the system to live lock.
4406 [FTRACE] enable this option to disable tracing when a
4407 warning is hit. This turns off "tracing_on". Tracing can
4408 be enabled again by echoing '1' into the "tracing_on"
4409 file located in /sys/kernel/debug/tracing/
4411 This option is useful, as it disables the trace before
4412 the WARNING dump is called, which prevents the trace to
4413 be filled with content caused by the warning output.
4415 This option can also be set at run time via the sysctl
4416 option: kernel/traceoff_on_warning
4418 transparent_hugepage=
4420 Format: [always|madvise|never]
4421 Can be used to control the default behavior of the system
4422 with respect to transparent hugepages.
4423 See Documentation/vm/transhuge.txt for more details.
4425 tsc= Disable clocksource stability checks for TSC.
4427 [x86] reliable: mark tsc clocksource as reliable, this
4428 disables clocksource verification at runtime, as well
4429 as the stability checks done at bootup. Used to enable
4430 high-resolution timer mode on older hardware, and in
4431 virtualized environment.
4432 [x86] noirqtime: Do not use TSC to do irq accounting.
4433 Used to run time disable IRQ_TIME_ACCOUNTING on any
4434 platforms where RDTSC is slow and this accounting
4437 turbografx.map[2|3]= [HW,JOY]
4438 TurboGraFX parallel port interface
4440 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4441 See also Documentation/input/joystick-parport.txt
4443 udbg-immortal [PPC] When debugging early kernel crashes that
4444 happen after console_init() and before a proper
4445 console driver takes over, this boot options might
4446 help "seeing" what's going on.
4448 uhash_entries= [KNL,NET]
4449 Set number of hash buckets for UDP/UDP-Lite connections
4452 [USB] Ignore overcurrent events (default N).
4453 Some badly-designed motherboards generate lots of
4454 bogus events, for ports that aren't wired to
4455 anything. Set this parameter to avoid log spamming.
4456 Note that genuine overcurrent events won't be
4460 [X86] Cause panic on unknown NMI.
4462 usbcore.authorized_default=
4463 [USB] Default USB device authorization:
4464 (default -1 = authorized except for wireless USB,
4465 0 = not authorized, 1 = authorized)
4467 usbcore.autosuspend=
4468 [USB] The autosuspend time delay (in seconds) used
4469 for newly-detected USB devices (default 2). This
4470 is the time required before an idle device will be
4471 autosuspended. Devices for which the delay is set
4472 to a negative value won't be autosuspended at all.
4474 usbcore.usbfs_snoop=
4475 [USB] Set to log all usbfs traffic (default 0 = off).
4477 usbcore.usbfs_snoop_max=
4478 [USB] Maximum number of bytes to snoop in each URB
4481 usbcore.blinkenlights=
4482 [USB] Set to cycle leds on hubs (default 0 = off).
4484 usbcore.old_scheme_first=
4485 [USB] Start with the old device initialization
4486 scheme (default 0 = off).
4488 usbcore.usbfs_memory_mb=
4489 [USB] Memory limit (in MB) for buffers allocated by
4490 usbfs (default = 16, 0 = max = 2047).
4492 usbcore.use_both_schemes=
4493 [USB] Try the other device initialization scheme
4494 if the first one fails (default 1 = enabled).
4496 usbcore.initial_descriptor_timeout=
4497 [USB] Specifies timeout for the initial 64-byte
4498 USB_REQ_GET_DESCRIPTOR request in milliseconds
4499 (default 5000 = 5.0 seconds).
4501 usbcore.nousb [USB] Disable the USB subsystem
4504 [USBHID] The interval which mice are to be polled at.
4507 [USBHID] The interval which joysticks are to be polled at.
4509 usb-storage.delay_use=
4510 [UMS] The delay in seconds before a new device is
4511 scanned for Logical Units (default 1).
4514 [UMS] A list of quirks entries to supplement or
4515 override the built-in unusual_devs list. List
4516 entries are separated by commas. Each entry has
4517 the form VID:PID:Flags where VID and PID are Vendor
4518 and Product ID values (4-digit hex numbers) and
4519 Flags is a set of characters, each corresponding
4520 to a common usb-storage quirk flag as follows:
4521 a = SANE_SENSE (collect more than 18 bytes
4523 b = BAD_SENSE (don't collect more than 18
4524 bytes of sense data);
4525 c = FIX_CAPACITY (decrease the reported
4526 device capacity by one sector);
4527 d = NO_READ_DISC_INFO (don't use
4528 READ_DISC_INFO command);
4529 e = NO_READ_CAPACITY_16 (don't use
4530 READ_CAPACITY_16 command);
4531 f = NO_REPORT_OPCODES (don't use report opcodes
4533 g = MAX_SECTORS_240 (don't transfer more than
4534 240 sectors at a time, uas only);
4535 h = CAPACITY_HEURISTICS (decrease the
4536 reported device capacity by one
4537 sector if the number is odd);
4538 i = IGNORE_DEVICE (don't bind to this
4540 j = NO_REPORT_LUNS (don't use report luns
4542 l = NOT_LOCKABLE (don't try to lock and
4543 unlock ejectable media);
4544 m = MAX_SECTORS_64 (don't transfer more
4545 than 64 sectors = 32 KB at a time);
4546 n = INITIAL_READ10 (force a retry of the
4547 initial READ(10) command);
4548 o = CAPACITY_OK (accept the capacity
4549 reported by the device);
4550 p = WRITE_CACHE (the device cache is ON
4552 r = IGNORE_RESIDUE (the device reports
4553 bogus residue values);
4554 s = SINGLE_LUN (the device has only one
4556 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4557 commands, uas only);
4558 u = IGNORE_UAS (don't bind to the uas driver);
4559 w = NO_WP_DETECT (don't test whether the
4560 medium is write-protected).
4561 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4562 even if the device claims no cache)
4563 Example: quirks=0419:aaf5:rl,0421:0433:rc
4565 user_debug= [KNL,ARM]
4567 See arch/arm/Kconfig.debug help text.
4568 1 - undefined instruction events
4570 4 - invalid data aborts
4573 Example: user_debug=31
4576 [X86] Flags controlling user PTE allocations.
4578 nohigh = do not allocate PTE pages in
4579 HIGHMEM regardless of setting
4583 On X86_32, this is an alias for vdso32=. Otherwise:
4585 vdso=1: enable VDSO (the default)
4586 vdso=0: disable VDSO mapping
4588 vdso32= [X86] Control the 32-bit vDSO
4589 vdso32=1: enable 32-bit VDSO
4590 vdso32=0 or vdso32=2: disable 32-bit VDSO
4592 See the help text for CONFIG_COMPAT_VDSO for more
4593 details. If CONFIG_COMPAT_VDSO is set, the default is
4594 vdso32=0; otherwise, the default is vdso32=1.
4596 For compatibility with older kernels, vdso32=2 is an
4599 Try vdso32=0 if you encounter an error that says:
4600 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
4603 vector=percpu: enable percpu vector domain
4605 video= [FB] Frame buffer configuration
4606 See Documentation/fb/modedb.txt.
4608 video.brightness_switch_enabled= [0,1]
4609 If set to 1, on receiving an ACPI notify event
4610 generated by hotkey, video driver will adjust brightness
4611 level and then send out the event to user space through
4612 the allocated input device; If set to 0, video driver
4613 will only send out the event without touching backlight
4618 [VMMIO] Memory mapped virtio (platform) device.
4620 <size>@<baseaddr>:<irq>[:<id>]
4622 <size> := size (can use standard suffixes
4624 <baseaddr> := physical base address
4625 <irq> := interrupt number (as passed to
4627 <id> := (optional) platform device id
4629 virtio_mmio.device=1K@0x100b0000:48:7
4631 Can be used multiple times for multiple devices.
4633 vga= [BOOT,X86-32] Select a particular video mode
4634 See Documentation/x86/boot.txt and
4635 Documentation/svga.txt.
4636 Use vga=ask for menu.
4637 This is actually a boot loader parameter; the value is
4638 passed to the kernel using a special protocol.
4640 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
4641 size of <nn>. This can be used to increase the
4642 minimum size (128MB on x86). It can also be used to
4643 decrease the size and leave more room for directly
4646 vmcp_cma=nn[MG] [KNL,S390]
4647 Sets the memory size reserved for contiguous memory
4648 allocations for the vmcp device driver.
4650 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
4653 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
4656 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
4660 Controls the behavior of vsyscalls (i.e. calls to
4661 fixed addresses of 0xffffffffff600x00 from legacy
4662 code). Most statically-linked binaries and older
4663 versions of glibc use these calls. Because these
4664 functions are at fixed addresses, they make nice
4665 targets for exploits that can control RIP.
4667 emulate [default] Vsyscalls turn into traps and are
4668 emulated reasonably safely.
4670 native Vsyscalls are native syscall instructions.
4671 This is a little bit faster than trapping
4672 and makes a few dynamic recompilers work
4673 better than they would in emulation mode.
4674 It also makes exploits much easier to write.
4676 none Vsyscalls don't work at all. This makes
4677 them quite hard to use for exploits but
4678 might break your system.
4680 vt.color= [VT] Default text color.
4681 Format: 0xYX, X = foreground, Y = background.
4682 Default: 0x07 = light gray on black.
4684 vt.cur_default= [VT] Default cursor shape.
4685 Format: 0xCCBBAA, where AA, BB, and CC are the same as
4686 the parameters of the <Esc>[?A;B;Cc escape sequence;
4687 see VGA-softcursor.txt. Default: 2 = underline.
4689 vt.default_blu= [VT]
4690 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
4691 Change the default blue palette of the console.
4692 This is a 16-member array composed of values
4695 vt.default_grn= [VT]
4696 Format: <green0>,<green1>,<green2>,...,<green15>
4697 Change the default green palette of the console.
4698 This is a 16-member array composed of values
4701 vt.default_red= [VT]
4702 Format: <red0>,<red1>,<red2>,...,<red15>
4703 Change the default red palette of the console.
4704 This is a 16-member array composed of values
4710 Set system-wide default UTF-8 mode for all tty's.
4711 Default is 1, i.e. UTF-8 mode is enabled for all
4712 newly opened terminals.
4714 vt.global_cursor_default=
4717 Set system-wide default for whether a cursor
4718 is shown on new VTs. Default is -1,
4719 i.e. cursors will be created by default unless
4720 overridden by individual drivers. 0 will hide
4721 cursors, 1 will display them.
4723 vt.italic= [VT] Default color for italic text; 0-15.
4726 vt.underline= [VT] Default color for underlined text; 0-15.
4729 watchdog timers [HW,WDT] For information on watchdog timers,
4730 see Documentation/watchdog/watchdog-parameters.txt
4731 or other driver-specific files in the
4732 Documentation/watchdog/ directory.
4734 workqueue.watchdog_thresh=
4735 If CONFIG_WQ_WATCHDOG is configured, workqueue can
4736 warn stall conditions and dump internal state to
4737 help debugging. 0 disables workqueue stall
4738 detection; otherwise, it's the stall threshold
4739 duration in seconds. The default value is 30 and
4740 it can be updated at runtime by writing to the
4741 corresponding sysfs file.
4743 workqueue.disable_numa
4744 By default, all work items queued to unbound
4745 workqueues are affine to the NUMA nodes they're
4746 issued on, which results in better behavior in
4747 general. If NUMA affinity needs to be disabled for
4748 whatever reason, this option can be used. Note
4749 that this also can be controlled per-workqueue for
4750 workqueues visible under /sys/bus/workqueue/.
4752 workqueue.power_efficient
4753 Per-cpu workqueues are generally preferred because
4754 they show better performance thanks to cache
4755 locality; unfortunately, per-cpu workqueues tend to
4756 be more power hungry than unbound workqueues.
4758 Enabling this makes the per-cpu workqueues which
4759 were observed to contribute significantly to power
4760 consumption unbound, leading to measurably lower
4761 power usage at the cost of small performance
4764 The default value of this parameter is determined by
4765 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
4767 workqueue.debug_force_rr_cpu
4768 Workqueue used to implicitly guarantee that work
4769 items queued without explicit CPU specified are put
4770 on the local CPU. This guarantee is no longer true
4771 and while local CPU is still preferred work items
4772 may be put on foreign CPUs. This debug option
4773 forces round-robin CPU selection to flush out
4774 usages which depend on the now broken guarantee.
4775 When enabled, memory and cache locality will be
4778 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
4779 default x2apic cluster mode on platforms
4782 x86_intel_mid_timer= [X86-32,APBT]
4783 Choose timer option for x86 Intel MID platform.
4784 Two valid options are apbt timer only and lapic timer
4785 plus one apbt timer for broadcast timer.
4786 x86_intel_mid_timer=apbt_only | lapic_and_apbt
4788 xen_512gb_limit [KNL,X86-64,XEN]
4789 Restricts the kernel running paravirtualized under Xen
4790 to use only up to 512 GB of RAM. The reason to do so is
4791 crash analysis tools and Xen tools for doing domain
4792 save/restore/migration must be enabled to handle larger
4795 xen_emul_unplug= [HW,X86,XEN]
4796 Unplug Xen emulated devices
4797 Format: [unplug0,][unplug1]
4798 ide-disks -- unplug primary master IDE devices
4799 aux-ide-disks -- unplug non-primary-master IDE devices
4800 nics -- unplug network devices
4801 all -- unplug all emulated devices (NICs and IDE disks)
4802 unnecessary -- unplugging emulated devices is
4803 unnecessary even if the host did not respond to
4805 never -- do not unplug even if version check succeeds
4807 xen_nopvspin [X86,XEN]
4808 Disables the ticketlock slowpath using Xen PV
4812 Disables the PV optimizations forcing the HVM guest to
4813 run as generic HVM guest with no PV drivers.
4815 xirc2ps_cs= [NET,PCMCIA]
4817 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]