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
118 acpi_no_auto_serialize [HW,ACPI]
119 Disable auto-serialization of AML methods
120 AML control methods that contain the opcodes to create
121 named objects will be marked as "Serialized" by the
122 auto-serialization feature.
123 This feature is enabled by default.
124 This option allows to turn off the feature.
126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
129 acpi_no_static_ssdt [HW,ACPI]
130 Disable installation of static SSDTs at early boot time
131 By default, SSDTs contained in the RSDT/XSDT will be
132 installed automatically and they will appear under
133 /sys/firmware/acpi/tables.
134 This option turns off this feature.
135 Note that specifying this option does not affect
136 dynamic table installation which will install SSDT
137 tables to /sys/firmware/acpi/tables/dynamic.
139 acpi_rsdp= [ACPI,EFI,KEXEC]
140 Pass the RSDP address to the kernel, mostly used
141 on machines running EFI runtime service to boot the
142 second kernel for kdump.
144 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
145 Format: To spoof as Windows 98: ="Microsoft Windows"
147 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
148 of 2 which is mandated by ACPI 6) as the supported ACPI
149 specification revision (when using this switch, it may
150 be necessary to carry out a cold reboot _twice_ in a
151 row to make it take effect on the platform firmware).
153 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
154 acpi_osi="string1" # add string1
155 acpi_osi="!string2" # remove string2
156 acpi_osi=!* # remove all strings
157 acpi_osi=! # disable all built-in OS vendor
159 acpi_osi=!! # enable all built-in OS vendor
161 acpi_osi= # disable all strings
163 'acpi_osi=!' can be used in combination with single or
164 multiple 'acpi_osi="string1"' to support specific OS
165 vendor string(s). Note that such command can only
166 affect the default state of the OS vendor strings, thus
167 it cannot affect the default state of the feature group
168 strings and the current state of the OS vendor strings,
169 specifying it multiple times through kernel command line
170 is meaningless. This command is useful when one do not
171 care about the state of the feature group strings which
172 should be controlled by the OSPM.
174 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
175 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
176 can make '_OSI("Windows 2000")' TRUE.
178 'acpi_osi=' cannot be used in combination with other
179 'acpi_osi=' command lines, the _OSI method will not
180 exist in the ACPI namespace. NOTE that such command can
181 only affect the _OSI support state, thus specifying it
182 multiple times through kernel command line is also
185 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
188 'acpi_osi=!*' can be used in combination with single or
189 multiple 'acpi_osi="string1"' to support specific
190 string(s). Note that such command can affect the
191 current state of both the OS vendor strings and the
192 feature group strings, thus specifying it multiple times
193 through kernel command line is meaningful. But it may
194 still not able to affect the final state of a string if
195 there are quirks related to this string. This command
196 is useful when one want to control the state of the
197 feature group strings to debug BIOS issues related to
200 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
201 '_OSI("Module Device")' FALSE.
202 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
203 '_OSI("Module Device")' TRUE.
204 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
206 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
208 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
209 they all will make '_OSI("Windows 2000")' TRUE.
212 Override the pmtimer bug detection: force the kernel
213 to assume that this machine's pmtimer latches its value
214 and always returns good values.
216 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
217 Format: { level | edge | high | low }
219 acpi_skip_timer_override [HW,ACPI]
220 Recognize and ignore IRQ0/pin2 Interrupt Override.
221 For broken nForce2 BIOS resulting in XT-PIC timer.
223 acpi_sleep= [HW,ACPI] Sleep options
224 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
225 old_ordering, nonvs, sci_force_enable, nobl }
226 See Documentation/power/video.txt for information on
228 s3_beep is for debugging; it makes the PC's speaker beep
229 as soon as the kernel's real-mode entry point is called.
230 s4_nohwsig prevents ACPI hardware signature from being
231 used during resume from hibernation.
232 old_ordering causes the ACPI 1.0 ordering of the _PTS
233 control method, with respect to putting devices into
234 low power states, to be enforced (the ACPI 2.0 ordering
235 of _PTS is used by default).
236 nonvs prevents the kernel from saving/restoring the
237 ACPI NVS memory during suspend/hibernation and resume.
238 sci_force_enable causes the kernel to set SCI_EN directly
239 on resume from S1/S3 (which is against the ACPI spec,
240 but some broken systems don't work without it).
241 nobl causes the internal blacklist of systems known to
242 behave incorrectly in some ways with respect to system
243 suspend and resume to be ignored (use wisely).
245 acpi_use_timer_override [HW,ACPI]
246 Use timer override. For some broken Nvidia NF5 boards
247 that require a timer override, but don't have HPET
249 add_efi_memmap [EFI; X86] Include EFI memory map in
250 kernel's map of available physical RAM.
253 { off | try_unsupported }
254 off: disable AGP support
255 try_unsupported: try to drive unsupported chipsets
256 (may crash computer or cause data corruption)
259 See Documentation/sound/alsa-configuration.rst
262 Allow the default userspace alignment fault handler
263 behaviour to be specified. Bit 0 enables warnings,
264 bit 1 enables fixups, and bit 2 sends a segfault.
266 align_va_addr= [X86-64]
267 Align virtual addresses by clearing slice [14:12] when
268 allocating a VMA at process creation time. This option
269 gives you up to 3% performance improvement on AMD F15h
270 machines (where it is enabled by default) for a
271 CPU-intensive style benchmark, and it can vary highly in
272 a microbenchmark depending on workload and compiler.
274 32: only for 32-bit processes
275 64: only for 64-bit processes
276 on: enable for both 32- and 64-bit processes
277 off: disable for both 32- and 64-bit processes
279 alloc_snapshot [FTRACE]
280 Allocate the ftrace snapshot buffer on boot up when the
281 main buffer is allocated. This is handy if debugging
282 and you need to use tracing_snapshot() on boot up, and
283 do not want to use tracing_snapshot_alloc() as it needs
284 to be done where GFP_KERNEL allocations are allowed.
286 amd_iommu= [HW,X86-64]
287 Pass parameters to the AMD IOMMU driver in the system.
289 fullflush - enable flushing of IO/TLB entries when
290 they are unmapped. Otherwise they are
291 flushed before they will be reused, which
293 off - do not initialize any AMD IOMMU found in
295 force_isolation - Force device isolation for all
296 devices. The IOMMU driver is not
297 allowed anymore to lift isolation
298 requirements as needed. This option
299 does not override iommu=pt
301 amd_iommu_dump= [HW,X86-64]
302 Enable AMD IOMMU driver option to dump the ACPI table
303 for AMD IOMMU. With this option enabled, AMD IOMMU
304 driver will print ACPI tables for AMD IOMMU during
305 IOMMU initialization.
307 amd_iommu_intr= [HW,X86-64]
308 Specifies one of the following AMD IOMMU interrupt
310 legacy - Use legacy interrupt remapping mode.
311 vapic - Use virtual APIC mode, which allows IOMMU
312 to inject interrupts directly into guest.
313 This mode requires kvm-amd.avic=1.
314 (Default when IOMMU HW support is present.)
316 amijoy.map= [HW,JOY] Amiga joystick support
317 Map of devices attached to JOY0DAT and JOY1DAT
319 See also Documentation/input/joydev/joystick.rst
321 analog.map= [HW,JOY] Analog joystick and gamepad support
322 Specifies type or capabilities of an analog joystick
323 connected to one of 16 gameports
324 Format: <type1>,<type2>,..<type16>
327 Power management functions (SPARCstation-4/5 + deriv.)
329 Disable APC CPU standby support. SPARCstation-Fox does
330 not play well with APC CPU idle - disable it if you have
331 APC and your system crashes randomly.
333 apic= [APIC,X86] Advanced Programmable Interrupt Controller
334 Change the output verbosity while booting
335 Format: { quiet (default) | verbose | debug }
336 Change the amount of debugging information output
337 when initialising the APIC and IO-APIC components.
338 For X86-32, this can also be used to specify an APIC
340 Format: apic=driver_name
341 Examples: apic=bigsmp
343 apic_extnmi= [APIC,X86] External NMI delivery setting
344 Format: { bsp (default) | all | none }
345 bsp: External NMI is delivered only to CPU 0
346 all: External NMIs are broadcast to all CPUs as a
348 none: External NMI is masked for all CPUs. This is
349 useful so that a dump capture kernel won't be
353 See Documentation/networking/ipv6.txt.
355 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
356 Limit apic dumping. The parameter defines the maximal
357 number of local apics being dumped. Also it is possible
358 to set it to "all" by meaning -- no limit here.
359 Format: { 1 (default) | 2 | ... | all }.
360 The parameter valid if only apic=debug or
361 apic=verbose is specified.
362 Example: apic=debug show_lapic=all
364 apm= [APM] Advanced Power Management
365 See header of arch/x86/kernel/apm_32.c.
367 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
368 Format: <io>,<irq>,<nodeID>
372 atarimouse= [HW,MOUSE] Atari Mouse
374 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
375 EzKey and similar keyboards
377 atkbd.reset= [HW] Reset keyboard during initialization
379 atkbd.set= [HW] Select keyboard code set
380 Format: <int> (2 = AT (default), 3 = PS/2)
382 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
385 atkbd.softraw= [HW] Choose between synthetic and real raw mode
386 Format: <bool> (0 = real, 1 = synthetic (default))
388 atkbd.softrepeat= [HW]
389 Use software keyboard repeat
391 audit= [KNL] Enable the audit sub-system
392 Format: { "0" | "1" | "off" | "on" }
393 0 | off - kernel audit is disabled and can not be
394 enabled until the next reboot
395 unset - kernel audit is initialized but disabled and
396 will be fully enabled by the userspace auditd.
397 1 | on - kernel audit is initialized and partially
398 enabled, storing at most audit_backlog_limit
399 messages in RAM until it is fully enabled by the
403 audit_backlog_limit= [KNL] Set the audit queue size limit.
404 Format: <int> (must be >=0)
407 bau= [X86_UV] Enable the BAU on SGI UV. The default
408 behavior is to disable the BAU (i.e. bau=0).
409 Format: { "0" | "1" }
412 unset - Disable the BAU.
414 baycom_epp= [HW,AX25]
417 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
419 See header of drivers/net/hamradio/baycom_par.c.
421 baycom_ser_fdx= [HW,AX25]
422 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
423 Format: <io>,<irq>,<mode>[,<baud>]
424 See header of drivers/net/hamradio/baycom_ser_fdx.c.
426 baycom_ser_hdx= [HW,AX25]
427 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
428 Format: <io>,<irq>,<mode>
429 See header of drivers/net/hamradio/baycom_ser_hdx.c.
431 blkdevparts= Manual partition parsing of block device(s) for
432 embedded devices based on command line input.
433 See Documentation/block/cmdline-partition.txt
435 boot_delay= Milliseconds to delay each printk during boot.
436 Values larger than 10 seconds (10000) are changed to
440 bootmem_debug [KNL] Enable bootmem allocator debug messages.
443 Disable BERT OS support on buggy BIOSes.
445 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
446 bttv.radio= Most important insmod options are available as
448 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst
451 bulk_remove=off [PPC] This parameter disables the use of the pSeries
452 firmware feature for flushing multiple hpte entries
455 c101= [NET] Moxa C101 synchronous serial card
457 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
458 Sometimes CPU hardware bugs make them report the cache
459 size incorrectly. The kernel will attempt work arounds
460 to fix known problems, but for some CPUs it is not
461 possible to determine what the correct size should be.
462 This option provides an override for these situations.
465 [NET] Specifies amount of time (in seconds) that
466 the kernel should wait for a network carrier. By default
467 it waits 120 seconds.
469 ca_keys= [KEYS] This parameter identifies a specific key(s) on
470 the system trusted keyring to be used for certificate
472 format: { id:<keyid> | builtin }
474 cca= [MIPS] Override the kernel pages' cache coherency
475 algorithm. Accepted values range from 0 to 7
476 inclusive. See arch/mips/include/asm/pgtable-bits.h
477 for platform specific values (SB1, Loongson3 and
480 ccw_timeout_log [S390]
481 See Documentation/s390/CommonIO for details.
483 cgroup_disable= [KNL] Disable a particular controller
484 Format: {name of the controller(s) to disable}
485 The effects of cgroup_disable=foo are:
486 - foo isn't auto-mounted if you mount all cgroups in
488 - foo isn't visible as an individually mountable
490 {Currently only "memory" controller deal with this and
491 cut the overhead, others just disable the usage. So
492 only cgroup_disable=memory is actually worthy}
494 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
495 Format: { { controller | "all" | "named" }
496 [,{ controller | "all" | "named" }...] }
497 Like cgroup_disable, but only applies to cgroup v1;
498 the blacklisted controllers remain available in cgroup2.
499 "all" blacklists all controllers and "named" disables
500 named mounts. Specifying both "all" and "named" disables
503 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
505 nosocket -- Disable socket memory accounting.
506 nokmem -- Disable kernel memory accounting.
508 checkreqprot [SELINUX] Set initial checkreqprot flag value.
509 Format: { "0" | "1" }
510 See security/selinux/Kconfig help text.
511 0 -- check protection applied by kernel (includes
512 any implied execute protection).
513 1 -- check protection requested by application.
514 Default value is set via a kernel config option.
515 Value can be changed at runtime via
516 /selinux/checkreqprot.
519 See Documentation/s390/CommonIO for details.
522 Prevents the clock framework from automatically gating
523 clocks that have not been explicitly enabled by a Linux
524 device driver but are enabled in hardware at reset or
525 by the bootloader/firmware. Note that this does not
526 force such clocks to be always-on nor does it reserve
527 those clocks in any way. This parameter is useful for
528 debug and development, but should not be needed on a
529 platform with proper driver support. For more
530 information, see Documentation/driver-api/clk.rst.
532 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
534 Forces specified clocksource (if available) to be used
535 when calculating gettimeofday(). If specified
536 clocksource is not available, it defaults to PIT.
537 Format: { pit | tsc | cyclone | pmtmr }
539 clocksource= Override the default clocksource
541 Override the default clocksource and use the clocksource
542 with the name specified.
543 Some clocksource names to choose from, depending on
545 [all] jiffies (this is the base, fallback clocksource)
547 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
548 pxa_timer,timer3,32k_counter,timer0_1
549 [X86-32] pit,hpet,tsc;
550 scx200_hrt on Geode; cyclone on IBM x440
558 clocksource.arm_arch_timer.evtstrm=
561 Enable/disable the eventstream feature of the ARM
562 architected timer so that code using WFE-based polling
563 loops can be debugged more effectively on production
566 clearcpuid=BITNUM [X86]
567 Disable CPUID feature X for the kernel. See
568 arch/x86/include/asm/cpufeatures.h for the valid bit
569 numbers. Note the Linux specific bits are not necessarily
570 stable over kernel options, but the vendor specific
572 Also note that user programs calling CPUID directly
573 or using the feature without checking anything
574 will still see it. This just prevents it from
575 being used by the kernel or shown in /proc/cpuinfo.
576 Also note the kernel might malfunction if you disable
579 cma=nn[MG]@[start[MG][-end[MG]]]
581 Sets the size of kernel global memory area for
582 contiguous memory allocations and optionally the
583 placement constraint by the physical address range of
584 memory allocations. A value of 0 disables CMA
585 altogether. For more information, see
586 include/linux/dma-contiguous.h
588 cmo_free_hint= [PPC] Format: { yes | no }
589 Specify whether pages are marked as being inactive
590 when they are freed. This is used in CMO environments
591 to determine OS memory pressure for page stealing by
595 coherent_pool=nn[KMG] [ARM,KNL]
596 Sets the size of memory pool for coherent, atomic dma
597 allocations, by default set to 256K.
599 com20020= [HW,NET] ARCnet - COM20020 chipset
601 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
603 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
607 ARCnet - COM90xx chipset (memory-mapped buffers)
608 Format: <io>[,<irq>[,<memstart>]]
610 condev= [HW,S390] console device
613 console= [KNL] Output console device and options.
615 tty<n> Use the virtual console device <n>.
619 Use the specified serial port. The options are of
620 the form "bbbbpnf", where "bbbb" is the baud rate,
621 "p" is parity ("n", "o", or "e"), "n" is number of
622 bits, and "f" is flow control ("r" for RTS or
623 omit it). Default is "9600n8".
625 See Documentation/admin-guide/serial-console.rst for more
627 Documentation/networking/netconsole.txt for an
630 uart[8250],io,<addr>[,options]
631 uart[8250],mmio,<addr>[,options]
632 uart[8250],mmio16,<addr>[,options]
633 uart[8250],mmio32,<addr>[,options]
634 uart[8250],0x<addr>[,options]
635 Start an early, polled-mode console on the 8250/16550
636 UART at the specified I/O port or MMIO address,
637 switching to the matching ttyS device later.
638 MMIO inter-register address stride is either 8-bit
639 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
640 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
641 to be equivalent to 'mmio'. 'options' are specified in
642 the same format described for ttyS above; if unspecified,
643 the h/w is not re-initialized.
645 hvc<n> Use the hypervisor console device <n>. This is for
646 both Xen and PowerPC hypervisors.
648 If the device connected to the port is not a TTY but a braille
649 device, prepend "brl," before the device type, for instance
651 For now, only VisioBraille is supported.
654 [KNL] Change console messages format
656 By default we print messages on consoles in
657 "[time stamp] text\n" format (time stamp may not be
658 printed, depending on CONFIG_PRINTK_TIME or
659 `printk_time' param).
661 Switch to syslog format: "<%u>[time stamp] text\n"
662 IOW, each message will have a facility and loglevel
663 prefix. The format is similar to one used by syslog()
664 syscall, or to executing "dmesg -S --raw" or to reading
667 consoleblank= [KNL] The console blank (screen saver) timeout in
668 seconds. A value of 0 disables the blank timer.
672 [KNL] Change the default value for
673 /proc/<pid>/coredump_filter.
674 See also Documentation/filesystems/proc.txt.
676 coresight_cpu_debug.enable
679 Enable/disable the CPU sampling based debugging.
680 0: default value, disable debugging
681 1: enable debugging at boot time
683 cpuidle.off=1 [CPU_IDLE]
684 disable the cpuidle sub-system
687 [CPU_IDLE] Name of the cpuidle governor to use.
689 cpufreq.off=1 [CPU_FREQ]
690 disable the cpufreq sub-system
693 [X86] Delay for N microsec between assert and de-assert
694 of APIC INIT to start processors. This delay occurs
695 on every CPU online, such as boot, and resume from suspend.
698 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
700 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
702 crashkernel=size[KMG][@offset[KMG]]
703 [KNL] Using kexec, Linux can switch to a 'crash kernel'
704 upon panic. This parameter reserves the physical
705 memory region [offset, offset + size] for that kernel
706 image. If '@offset' is omitted, then a suitable offset
707 is selected automatically.
708 [KNL, x86_64] select a region under 4G first, and
709 fall back to reserve region above 4G when '@offset'
710 hasn't been specified.
711 See Documentation/kdump/kdump.txt for further details.
713 crashkernel=range1:size1[,range2:size2,...][@offset]
714 [KNL] Same as above, but depends on the memory
715 in the running system. The syntax of range is
716 start-[end] where start and end are both
717 a memory unit (amount[KMG]). See also
718 Documentation/kdump/kdump.txt for an example.
720 crashkernel=size[KMG],high
721 [KNL, x86_64] range could be above 4G. Allow kernel
722 to allocate physical memory region from top, so could
723 be above 4G if system have more than 4G ram installed.
724 Otherwise memory region will be allocated below 4G, if
726 It will be ignored if crashkernel=X is specified.
727 crashkernel=size[KMG],low
728 [KNL, x86_64] range under 4G. When crashkernel=X,high
729 is passed, kernel could allocate physical memory region
730 above 4G, that cause second kernel crash on system
731 that require some amount of low memory, e.g. swiotlb
732 requires at least 64M+32K low memory, also enough extra
733 low memory is needed to make sure DMA buffers for 32-bit
734 devices won't run out. Kernel would try to allocate at
735 at least 256M below 4G automatically.
736 This one let user to specify own low range under 4G
737 for second kernel instead.
738 0: to disable low allocation.
739 It will be ignored when crashkernel=X,high is not used
740 or memory reserved is below 4G.
743 [KNL] Disable crypto self-tests
748 cs89x0_media= [HW,NET]
749 Format: { rj45 | aui | bnc }
752 See header of drivers/s390/block/dasd_devmap.c.
754 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
755 (one device per port)
756 Format: <port#>,<type>
757 See also Documentation/input/devices/joystick-parport.rst
759 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
761 Documentation/admin-guide/dynamic-debug-howto.rst for
762 details. Deprecated, see dyndbg.
764 debug [KNL] Enable kernel debugging (events log level).
767 [KNL] Enable printing [hashed] pointers early in the
768 boot sequence. If enabled, we use a weak hash instead
769 of siphash to hash pointers. Use this option if you are
770 seeing instances of '(___ptrval___)') and need to see a
771 value (hashed pointer) instead. Cryptographically
772 insecure, please do not use on production kernels.
775 [KNL] verbose self-tests
777 Print debugging info while doing the locking API
779 We default to 0 (no extra messages), setting it to
780 1 will print _a lot_ more information - normally
781 only useful to kernel developers.
783 debug_objects [KNL] Enable object debugging
786 [KNL] Disable object debugging
788 debug_guardpage_minorder=
789 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
790 parameter allows control of the order of pages that will
791 be intentionally kept free (and hence protected) by the
792 buddy allocator. Bigger value increase the probability
793 of catching random memory corruption, but reduce the
794 amount of memory for normal system use. The maximum
795 possible value is MAX_ORDER/2. Setting this parameter
796 to 1 or 2 should be enough to identify most random
797 memory corruption problems caused by bugs in kernel or
798 driver code when a CPU writes to (or reads from) a
799 random memory location. Note that there exists a class
800 of memory corruptions problems caused by buggy H/W or
801 F/W or by drivers badly programing DMA (basically when
802 memory is written at bus level and the CPU MMU is
803 bypassed) which are not detectable by
804 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
805 tracking down these problems.
808 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
809 parameter enables the feature at boot time. In
810 default, it is disabled. We can avoid allocating huge
811 chunk of memory for debug pagealloc if we don't enable
812 it at boot time and the system will work mostly same
813 with the kernel built without CONFIG_DEBUG_PAGEALLOC.
814 on: enable the feature
816 debugpat [X86] Enable PAT debugging
818 decnet.addr= [HW,NET]
819 Format: <area>[,<node>]
820 See also Documentation/networking/decnet.txt.
823 [same as hugepagesz=] The size of the default
824 HugeTLB page size. This is the size represented by
825 the legacy /proc/ hugepages APIs, used for SHM, and
826 default size when mounting hugetlbfs filesystems.
827 Defaults to the default architecture's huge page size
830 deferred_probe_timeout=
831 [KNL] Debugging option to set a timeout in seconds for
832 deferred probe to give up waiting on dependencies to
833 probe. Only specific dependencies (subsystems or
834 drivers) that have opted in will be ignored. A timeout of 0
835 will timeout at the end of initcalls. This option will also
836 dump out devices still on the deferred probe list after
840 Set number of hash buckets for dentry cache.
842 disable_1tb_segments [PPC]
843 Disables the use of 1TB hash page table segments. This
844 causes the kernel to fall back to 256MB segments which
845 can be useful when debugging issues that require an SLB
849 See Documentation/networking/ipv6.txt.
852 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
853 hardening is enabled for this boot. Hardened
854 usercopy checking is used to protect the kernel
855 from reading or writing beyond known memory
856 allocation boundaries as a proactive defense
857 against bounds-checking flaws in the kernel's
858 copy_to_user()/copy_from_user() interface.
859 on Perform hardened usercopy checks (default).
860 off Disable hardened usercopy checks.
863 Disable RADIX MMU mode on POWER9
865 disable_cpu_apicid= [X86,APIC,SMP]
867 The number of initial APIC ID for the
868 corresponding CPU to be disabled at boot,
869 mostly used for the kdump 2nd kernel to
870 disable BSP to wake up multiple CPUs without
871 causing system reset or hang due to sending
874 perf_v4_pmi= [X86,INTEL]
876 Disable Intel PMU counter freezing feature.
877 The feature only exists starting from
878 Arch Perfmon v4 (Skylake and newer).
880 disable_ddw [PPC/PSERIES]
881 Disable Dynamic DMA Window support. Use this if
882 to workaround buggy firmware.
885 See Documentation/networking/ipv6.txt.
887 disable_mtrr_cleanup [X86]
888 The kernel tries to adjust MTRR layout from continuous
889 to discrete, to make X server driver able to add WB
890 entry later. This parameter disables that.
892 disable_mtrr_trim [X86, Intel and AMD only]
893 By default the kernel will trim any uncacheable
894 memory out of your available memory pool based on
895 MTRR settings. This parameter disables that behavior,
896 possibly causing your machine to run very slowly.
898 disable_timer_pin_1 [X86]
899 Disable PIN 1 of APIC timer
900 Can be useful to work around chipset bugs.
902 dis_ucode_ldr [X86] Disable the microcode loader.
904 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
905 this option disables the debugging code at boot.
907 dma_debug_entries=<number>
908 This option allows to tune the number of preallocated
909 entries for DMA-API debugging code. One entry is
910 required per DMA-API allocation. Use this if the
911 DMA-API debugging code disables itself because the
912 architectural default is too low.
914 dma_debug_driver=<driver_name>
915 With this option the DMA-API debugging driver
916 filter feature can be enabled at boot time. Just
917 pass the driver to filter for as the parameter.
918 The filter can be disabled or changed to another
919 driver later using sysfs.
921 driver_async_probe= [KNL]
922 List of driver names to be probed asynchronously.
923 Format: <driver_name1>,<driver_name2>...
925 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
926 Broken monitors, graphic adapters, KVMs and EDIDless
927 panels may send no or incorrect EDID data sets.
928 This parameter allows to specify an EDID data sets
929 in the /lib/firmware directory that are used instead.
930 Generic built-in EDID data sets are used, if one of
931 edid/1024x768.bin, edid/1280x1024.bin,
932 edid/1680x1050.bin, or edid/1920x1080.bin is given
933 and no file with the same name exists. Details and
934 instructions how to build your own EDID data are
935 available in Documentation/EDID/HOWTO.txt. An EDID
936 data set will only be used for a particular connector,
937 if its name and a colon are prepended to the EDID
938 name. Each connector may use a unique EDID data
939 set by separating the files with a comma. An EDID
940 data set with no connector name will be used for
941 any connectors not explicitly specified.
946 Format: {"off" | "known"}
947 Control how the dt_cpu_ftrs device-tree binding is
948 used for CPU feature discovery and setup (if it
950 off: Do not use it, fall back to legacy cpu table.
951 known: Do not pass through unknown features to guests
952 or userspace, only those that the kernel is aware of.
954 dump_apple_properties [X86]
955 Dump name and content of EFI device properties on
956 x86 Macs. Useful for driver authors to determine
957 what data is available or for reverse-engineering.
959 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
960 module.dyndbg[="val"]
961 Enable debug messages at boot time. See
962 Documentation/admin-guide/dynamic-debug-howto.rst
965 nompx [X86] Disables Intel Memory Protection Extensions.
966 See Documentation/x86/intel_mpx.txt for more
967 information about the feature.
969 nopku [X86] Disable Memory Protection Keys CPU feature found
972 module.async_probe [KNL]
973 Enable asynchronous probe on this module.
975 early_ioremap_debug [KNL]
976 Enable debug messages in early_ioremap support. This
977 is useful for tracking down temporary early mappings
978 which are not unmapped.
980 earlycon= [KNL] Output early console device and options.
982 [ARM64] The early console is determined by the
983 stdout-path property in device tree's chosen node,
984 or determined by the ACPI SPCR table.
986 [X86] When used with no options the early console is
987 determined by the ACPI SPCR table.
989 cdns,<addr>[,options]
990 Start an early, polled-mode console on a Cadence
991 (xuartps) serial port at the specified address. Only
992 supported option is baud rate. If baud rate is not
993 specified, the serial port must already be setup and
996 uart[8250],io,<addr>[,options]
997 uart[8250],mmio,<addr>[,options]
998 uart[8250],mmio32,<addr>[,options]
999 uart[8250],mmio32be,<addr>[,options]
1000 uart[8250],0x<addr>[,options]
1001 Start an early, polled-mode console on the 8250/16550
1002 UART at the specified I/O port or MMIO address.
1003 MMIO inter-register address stride is either 8-bit
1004 (mmio) or 32-bit (mmio32 or mmio32be).
1005 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1006 to be equivalent to 'mmio'. 'options' are specified
1007 in the same format described for "console=ttyS<n>"; if
1008 unspecified, the h/w is not initialized.
1012 Start an early, polled-mode console on a pl011 serial
1013 port at the specified address. The pl011 serial port
1014 must already be setup and configured. Options are not
1015 yet supported. If 'mmio32' is specified, then only
1016 the driver will use only 32-bit accessors to read/write
1017 the device registers.
1020 Start an early, polled-mode console on a meson serial
1021 port at the specified address. The serial port must
1022 already be setup and configured. Options are not yet
1026 Start an early, polled-mode console on an msm serial
1027 port at the specified address. The serial port
1028 must already be setup and configured. Options are not
1031 msm_serial_dm,<addr>
1032 Start an early, polled-mode console on an msm serial
1033 dm port at the specified address. The serial port
1034 must already be setup and configured. Options are not
1038 Start an early, polled-mode console on a serial port
1039 of an Actions Semi SoC, such as S500 or S900, at the
1040 specified address. The serial port must already be
1041 setup and configured. Options are not yet supported.
1044 Start an early, polled-mode console on a serial port
1045 of an RDA Micro SoC, such as RDA8810PL, at the
1046 specified address. The serial port must already be
1047 setup and configured. Options are not yet supported.
1049 smh Use ARM semihosting calls for early console.
1057 Use early console provided by serial driver available
1058 on Samsung SoCs, requires selecting proper type and
1059 a correct base address of the selected UART port. The
1060 serial port must already be setup and configured.
1061 Options are not yet supported.
1064 Start an early, polled-mode console on a lantiq serial
1065 (lqasc) port at the specified address. The serial port
1066 must already be setup and configured. Options are not
1071 Use early console provided by Freescale LP UART driver
1072 found on Freescale Vybrid and QorIQ LS1021A processors.
1073 A valid base address must be provided, and the serial
1074 port must already be setup and configured.
1077 Start an early, polled-mode console on the
1078 Armada 3700 serial port at the specified
1079 address. The serial port must already be setup
1080 and configured. Options are not yet supported.
1083 Start an early, polled-mode console on a Qualcomm
1084 Generic Interface (GENI) based serial port at the
1085 specified address. The serial port must already be
1086 setup and configured. Options are not yet supported.
1089 Start an early, unaccelerated console on the EFI
1090 memory mapped framebuffer (if available). On cache
1091 coherent non-x86 systems that use system memory for
1092 the framebuffer, pass the 'ram' option so that it is
1093 mapped with the correct attributes.
1095 earlyprintk= [X86,SH,ARM,M68k,S390]
1099 earlyprintk=serial[,ttySn[,baudrate]]
1100 earlyprintk=serial[,0x...[,baudrate]]
1101 earlyprintk=ttySn[,baudrate]
1102 earlyprintk=dbgp[debugController#]
1103 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1104 earlyprintk=xdbc[xhciController#]
1106 earlyprintk is useful when the kernel crashes before
1107 the normal console is initialized. It is not enabled by
1108 default because it has some cosmetic problems.
1110 Append ",keep" to not disable it when the real console
1113 Only one of vga, efi, serial, or usb debug port can
1116 Currently only ttyS0 and ttyS1 may be specified by
1117 name. Other I/O ports may be explicitly specified
1118 on some architectures (x86 and arm at least) by
1119 replacing ttySn with an I/O port address, like this:
1120 earlyprintk=serial,0x1008,115200
1121 You can find the port for a given device in
1122 /proc/tty/driver/serial:
1123 2: uart:ST16650V2 port:00001008 irq:18 ...
1125 Interaction with the standard serial driver is not
1128 The VGA and EFI output is eventually overwritten by
1131 The xen output can only be used by Xen PV guests.
1133 The sclp output can only be used on s390.
1135 The optional "force" to "pciserial" enables use of a
1136 PCI device even when its classcode is not of the
1139 edac_report= [HW,EDAC] Control how to report EDAC event
1140 Format: {"on" | "off" | "force"}
1141 on: enable EDAC to report H/W event. May be overridden
1142 by other higher priority error reporting module.
1143 off: disable H/W event reporting through EDAC.
1144 force: enforce the use of EDAC to report H/W event.
1147 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1150 This is designed to be used in conjunction with
1151 the boot argument: earlyprintk=vga
1154 Format: {"off" | "on" | "skip[mbr]"}
1157 Format: { "old_map", "nochunk", "noruntime", "debug" }
1158 old_map [X86-64]: switch to the old ioremap-based EFI
1159 runtime services mapping. 32-bit still uses this one by
1161 nochunk: disable reading files in "chunks" in the EFI
1162 boot stub, as chunking can cause problems with some
1163 firmware implementations.
1164 noruntime : disable EFI runtime services support
1165 debug: enable misc debug output
1167 efi_no_storage_paranoia [EFI; X86]
1168 Using this parameter you can use more than 50% of
1169 your efi variable storage. Use this parameter only if
1170 you are really sure that your UEFI does sane gc and
1171 fulfills the spec otherwise your board may brick.
1173 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1174 Add arbitrary attribute to specific memory range by
1175 updating original EFI memory map.
1176 Region of memory which aa attribute is added to is
1178 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1179 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1180 attribute is added to range 0x100000000-0x180000000 and
1181 0x10a0000000-0x1120000000.
1183 Using this parameter you can do debugging of EFI memmap
1184 related feature. For example, you can do debugging of
1185 Address Range Mirroring feature even if your box
1188 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1189 that is to be dynamically loaded by Linux. If there are
1190 multiple variables with the same name but with different
1191 vendor GUIDs, all of them will be loaded. See
1192 Documentation/acpi/ssdt-overlays.txt for details.
1195 eisa_irq_edge= [PARISC,HW]
1196 See header of drivers/parisc/eisa.c.
1199 See comment before function elanfreq_setup() in
1200 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1203 Format: { "mq-deadline" | "kyber" | "bfq" }
1204 See Documentation/block/deadline-iosched.txt,
1205 Documentation/block/kyber-iosched.txt and
1206 Documentation/block/bfq-iosched.txt for details.
1208 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1209 Specifies physical address of start of kernel core
1210 image elf header and optionally the size. Generally
1211 kexec loader will pass this option to capture kernel.
1212 See Documentation/kdump/kdump.txt for details.
1214 enable_mtrr_cleanup [X86]
1215 The kernel tries to adjust MTRR layout from continuous
1216 to discrete, to make X server driver able to add WB
1217 entry later. This parameter enables that.
1219 enable_timer_pin_1 [X86]
1220 Enable PIN 1 of APIC timer
1221 Can be useful to work around chipset bugs
1222 (in particular on some ATI chipsets).
1223 The kernel tries to set a reasonable default.
1225 enforcing [SELINUX] Set initial enforcing status.
1227 See security/selinux/Kconfig help text.
1228 0 -- permissive (log only, no denials).
1229 1 -- enforcing (deny and log).
1231 Value can be changed at runtime via /selinux/enforce.
1234 Disable Error Record Serialization Table (ERST)
1237 ether= [HW,NET] Ethernet cards parameters
1238 This option is obsoleted by the "netdev=" option, which
1239 has equivalent usage. See its documentation for details.
1243 Permit 'security.evm' to be updated regardless of
1244 current integrity status.
1248 fail_make_request=[KNL]
1249 General fault injection mechanism.
1250 Format: <interval>,<probability>,<space>,<times>
1251 See also Documentation/fault-injection/.
1254 See Documentation/blockdev/floppy.txt.
1256 force_pal_cache_flush
1257 [IA-64] Avoid check_sal_cache_flush which may hang on
1258 buggy SAL_CACHE_FLUSH implementations. Using this
1259 parameter will force ia64_sal_cache_flush to call
1260 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1263 Forcefully enable Physical Address Extension (PAE).
1264 Many Pentium M systems disable PAE but may have a
1265 functionally usable PAE implementation.
1266 Warning: use of this parameter will taint the kernel
1267 and may cause unknown problems.
1270 [FTRACE] will set and start the specified tracer
1271 as early as possible in order to facilitate early
1274 ftrace_dump_on_oops[=orig_cpu]
1275 [FTRACE] will dump the trace buffers on oops.
1276 If no parameter is passed, ftrace will dump
1277 buffers of all CPUs, but if you pass orig_cpu, it will
1278 dump only the buffer of the CPU that triggered the
1281 ftrace_filter=[function-list]
1282 [FTRACE] Limit the functions traced by the function
1283 tracer at boot up. function-list is a comma separated
1284 list of functions. This list can be changed at run
1285 time by the set_ftrace_filter file in the debugfs
1288 ftrace_notrace=[function-list]
1289 [FTRACE] Do not trace the functions specified in
1290 function-list. This list can be changed at run time
1291 by the set_ftrace_notrace file in the debugfs
1294 ftrace_graph_filter=[function-list]
1295 [FTRACE] Limit the top level callers functions traced
1296 by the function graph tracer at boot up.
1297 function-list is a comma separated list of functions
1298 that can be changed at run time by the
1299 set_graph_function file in the debugfs tracing directory.
1301 ftrace_graph_notrace=[function-list]
1302 [FTRACE] Do not trace from the functions specified in
1303 function-list. This list is a comma separated list of
1304 functions that can be changed at run time by the
1305 set_graph_notrace file in the debugfs tracing directory.
1307 ftrace_graph_max_depth=<uint>
1308 [FTRACE] Used with the function graph tracer. This is
1309 the max depth it will trace into a function. This value
1310 can be changed at run time by the max_graph_depth file
1311 in the tracefs tracing directory. default: 0 (no limit)
1314 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1315 support via parallel port (up to 5 devices per port)
1316 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1317 See also Documentation/input/devices/joystick-parport.rst
1321 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1325 gcov_persist= [GCOV] When non-zero (default), profiling data for
1326 kernel modules is saved and remains accessible via
1327 debugfs, even when the module is unloaded/reloaded.
1328 When zero, profiling data is discarded and associated
1329 debugfs files are removed at module unload time.
1331 goldfish [X86] Enable the goldfish android emulator platform.
1332 Don't use this when you are not running on the
1335 gpt [EFI] Forces disk with valid GPT signature but
1336 invalid Protective MBR to be treated as GPT. If the
1337 primary GPT is corrupted, it enables the backup/alternate
1338 GPT to be used instead.
1340 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1341 the "Enable 0" bit of the configuration register.
1344 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1345 the "Enable 0" bit of the configuration register.
1348 grcan.select= [HW] Select which physical interface to use.
1351 grcan.txsize= [HW] Sets the size of the tx buffer.
1352 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1354 grcan.rxsize= [HW] Sets the size of the rx buffer.
1355 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1358 gpio-mockup.gpio_mockup_ranges
1359 [HW] Sets the ranges of gpiochip of for this device.
1360 Format: <start1>,<end1>,<start2>,<end2>...
1362 hardlockup_all_cpu_backtrace=
1363 [KNL] Should the hard-lockup detector generate
1364 backtraces on all cpus.
1367 hashdist= [KNL,NUMA] Large hashes allocated during boot
1368 are distributed across NUMA nodes. Defaults on
1369 for 64-bit NUMA, off otherwise.
1370 Format: 0 | 1 (for off | on)
1372 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1374 hd= [EIDE] (E)IDE hard drive subsystem geometry
1375 Format: <cyl>,<head>,<sect>
1378 Disable Hardware Error Source Table (HEST) support;
1379 corresponding firmware-first mode error processing
1380 logic will be disabled.
1382 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1383 size of <nn>. This works even on boxes that have no
1384 highmem otherwise. This also works to reduce highmem
1385 size on bigger boxes.
1387 highres= [KNL] Enable/disable high resolution timer mode.
1388 Valid parameters: "on", "off"
1392 See Documentation/isdn/README.HiSax.
1396 hpet= [X86-32,HPET] option to control HPET usage
1397 Format: { enable (default) | disable | force |
1399 disable: disable HPET and use PIT instead
1400 force: allow force enabled of undocumented chips (ICH4,
1402 verbose: show contents of HPET registers during setup
1404 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1405 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1407 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1408 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1409 On x86-64 and powerpc, this option can be specified
1410 multiple times interleaved with hugepages= to reserve
1411 huge pages of different sizes. Valid pages sizes on
1412 x86-64 are 2M (when the CPU supports "pse") and 1G
1413 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1416 [KNL] Should the hung task detector generate panics.
1419 A nonzero value instructs the kernel to panic when a
1420 hung task is detected. The default value is controlled
1421 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1422 option. The value selected by this boot parameter can
1423 be changed later by the kernel.hung_task_panic sysctl.
1425 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1426 terminal devices. Valid values: 0..8
1427 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1428 If specified, z/VM IUCV HVC accepts connections
1429 from listed z/VM user IDs only.
1431 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1432 which allow the hypervisor to 'idle' the
1433 guest on lock contention.
1436 Do not unregister boot console at start. This is only
1437 useful for debugging when something happens in the window
1438 between unregistering the boot console and initializing
1441 i2c_bus= [HW] Override the default board specific I2C bus speed
1442 or register an additional I2C bus that is not
1443 registered from board initialization code.
1447 i8042.debug [HW] Toggle i8042 debug mode
1448 i8042.unmask_kbd_data
1449 [HW] Enable printing of interrupt data from the KBD port
1450 (disabled by default, and as a pre-condition
1451 requires that i8042.debug=1 be enabled)
1452 i8042.direct [HW] Put keyboard port into non-translated mode
1453 i8042.dumbkbd [HW] Pretend that controller can only read data from
1454 keyboard and cannot control its state
1455 (Don't attempt to blink the leds)
1456 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1457 i8042.nokbd [HW] Don't check/create keyboard port
1458 i8042.noloop [HW] Disable the AUX Loopback command while probing
1460 i8042.nomux [HW] Don't check presence of an active multiplexing
1462 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1464 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1465 i8042.reset [HW] Reset the controller during init, cleanup and
1466 suspend-to-ram transitions, only during s2r
1467 transitions, or never reset
1468 Format: { 1 | Y | y | 0 | N | n }
1469 1, Y, y: always reset controller
1470 0, N, n: don't ever reset controller
1471 Default: only on s2r transitions on x86; most other
1472 architectures force reset to be always executed
1473 i8042.unlock [HW] Unlock (ignore) the keylock
1474 i8042.kbdreset [HW] Reset device connected to KBD port
1478 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1479 indicates that the driver is running on unsupported
1481 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1482 does not match list of supported models.
1484 [HW] Report power status in /proc/i8k
1485 (disabled by default)
1486 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1489 i915.invert_brightness=
1490 [DRM] Invert the sense of the variable that is used to
1491 set the brightness of the panel backlight. Normally a
1492 brightness value of 0 indicates backlight switched off,
1493 and the maximum of the brightness value sets the backlight
1494 to maximum brightness. If this parameter is set to 0
1495 (default) and the machine requires it, or this parameter
1496 is set to 1, a brightness value of 0 sets the backlight
1497 to maximum brightness, and the maximum of the brightness
1498 value switches the backlight off.
1499 -1 -- never invert brightness
1500 0 -- machine default
1501 1 -- force brightness inversion
1504 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1506 ide-core.nodma= [HW] (E)IDE subsystem
1507 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1508 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1509 .cdrom .chs .ignore_cable are additional options
1510 See Documentation/ide/ide.txt.
1512 ide-generic.probe-mask= [HW] (E)IDE subsystem
1514 Probe mask for legacy ISA IDE ports. Depending on
1515 platform up to 6 ports are supported, enabled by
1516 setting corresponding bits in the mask to 1. The
1517 default value is 0x0, which has a special meaning.
1518 On systems that have PCI, it triggers scanning the
1519 PCI bus for the first and the second port, which
1520 are then probed. On systems without PCI the value
1521 of 0x0 enables probing the two first ports as if it
1524 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1525 Claim all unknown PCI IDE storage controllers.
1528 Format: idle=poll, idle=halt, idle=nomwait
1529 Poll forces a polling idle loop that can slightly
1530 improve the performance of waking up a idle CPU, but
1531 will use a lot of power and make the system run hot.
1533 idle=halt: Halt is forced to be used for CPU idle.
1534 In such case C2/C3 won't be used again.
1535 idle=nomwait: Disable mwait for CPU C-states
1537 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1538 Format: { strict | legacy | 2008 | relaxed }
1541 Choose which programs will be accepted for execution
1542 based on the IEEE 754 NaN encoding(s) supported by
1543 the FPU and the NaN encoding requested with the value
1544 of an ELF file header flag individually set by each
1545 binary. Hardware implementations are permitted to
1546 support either or both of the legacy and the 2008 NaN
1549 Available settings are as follows:
1550 strict accept binaries that request a NaN encoding
1551 supported by the FPU
1552 legacy only accept legacy-NaN binaries, if supported
1554 2008 only accept 2008-NaN binaries, if supported
1556 relaxed accept any binaries regardless of whether
1557 supported by the FPU
1559 The FPU emulator is always able to support both NaN
1560 encodings, so if no FPU hardware is present or it has
1561 been disabled with 'nofpu', then the settings of
1562 'legacy' and '2008' strap the emulator accordingly,
1563 'relaxed' straps the emulator for both legacy-NaN and
1564 2008-NaN, whereas 'strict' enables legacy-NaN only on
1565 legacy processors and both NaN encodings on MIPS32 or
1568 The setting for ABS.fmt/NEG.fmt instruction execution
1569 mode generally follows that for the NaN encoding,
1570 except where unsupported by hardware.
1572 ignore_loglevel [KNL]
1573 Ignore loglevel setting - this will print /all/
1574 kernel messages to the console. Useful for debugging.
1575 We also add it as printk module parameter, so users
1576 could change it dynamically, usually by
1577 /sys/module/printk/parameters/ignore_loglevel.
1580 Ignore RLIMIT_DATA setting for data mappings,
1581 print warning at first misuse. Can be changed via
1582 /sys/module/kernel/parameters/ignore_rlimit_data.
1584 ihash_entries= [KNL]
1585 Set number of hash buckets for inode cache.
1587 ima_appraise= [IMA] appraise integrity measurements
1588 Format: { "off" | "enforce" | "fix" | "log" }
1591 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1592 The builtin appraise policy appraises all files
1595 ima_canonical_fmt [IMA]
1596 Use the canonical format for the binary runtime
1597 measurements, instead of host native format.
1600 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1604 The list of supported hash algorithms is defined
1605 in crypto/hash_info.h.
1608 The builtin policies to load during IMA setup.
1609 Format: "tcb | appraise_tcb | secure_boot |
1612 The "tcb" policy measures all programs exec'd, files
1613 mmap'd for exec, and all files opened with the read
1614 mode bit set by either the effective uid (euid=0) or
1617 The "appraise_tcb" policy appraises the integrity of
1618 all files owned by root.
1620 The "secure_boot" policy appraises the integrity
1621 of files (eg. kexec kernel image, kernel modules,
1622 firmware, policy, etc) based on file signatures.
1624 The "fail_securely" policy forces file signature
1625 verification failure also on privileged mounted
1626 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1629 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1630 Load a policy which meets the needs of the Trusted
1631 Computing Base. This means IMA will measure all
1632 programs exec'd, files mmap'd for exec, and all files
1633 opened for read by uid=0.
1636 Select one of defined IMA measurements template formats.
1637 Formats: { "ima" | "ima-ng" | "ima-sig" }
1641 [IMA] Define a custom template format.
1642 Format: { "field1|...|fieldN" }
1644 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1645 Format: <min_file_size>
1646 Set the minimal file size for using asynchronous hash.
1647 If left unspecified, ahash usage is disabled.
1649 ahash performance varies for different data sizes on
1650 different crypto accelerators. This option can be used
1651 to achieve the best performance for a particular HW.
1653 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1655 Set hashing buffer size. Default: 4k.
1657 ahash performance varies for different chunk sizes on
1658 different crypto accelerators. This option can be used
1659 to achieve best performance for particular HW.
1663 Run specified binary instead of /sbin/init as init
1666 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1667 for working out where the kernel is dying during
1670 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1671 initcall functions. Useful for debugging built-in
1672 modules and initcalls.
1674 initrd= [BOOT] Specify the location of the initial ramdisk
1676 init_pkru= [x86] Specify the default memory protection keys rights
1677 register contents for all processes. 0x55555554 by
1678 default (disallow access to all but pkey 0). Can
1679 override in debugfs after boot.
1681 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1684 int_pln_enable [x86] Enable power limit notification interrupt
1686 integrity_audit=[IMA]
1687 Format: { "0" | "1" }
1688 0 -- basic integrity auditing messages. (Default)
1689 1 -- additional integrity auditing messages.
1691 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1693 Enable intel iommu driver.
1695 Disable intel iommu driver.
1696 igfx_off [Default Off]
1697 By default, gfx is mapped as normal device. If a gfx
1698 device has a dedicated DMAR unit, the DMAR unit is
1699 bypassed by not enabling DMAR with this option. In
1700 this case, gfx device will use physical address for
1703 With this option iommu will not optimize to look
1704 for io virtual address below 32-bit forcing dual
1705 address cycle on pci bus for cards supporting greater
1706 than 32-bit addressing. The default is to look
1707 for translation below 32-bit and if not available
1708 then look in the higher range.
1709 strict [Default Off]
1710 With this option on every unmap_single operation will
1711 result in a hardware IOTLB flush operation as opposed
1712 to batching them for performance.
1713 sp_off [Default Off]
1714 By default, super page will be supported if Intel IOMMU
1715 has the capability. With this option, super page will
1718 By default, scalable mode will be disabled even if the
1719 hardware advertises that it has support for the scalable
1720 mode translation. With this option set, scalable mode
1721 will be used on hardware which claims to support it.
1722 tboot_noforce [Default Off]
1723 Do not force the Intel IOMMU enabled under tboot.
1724 By default, tboot will force Intel IOMMU on, which
1725 could harm performance of some high-throughput
1726 devices like 40GBit network cards, even if identity
1728 Note that using this option lowers the security
1729 provided by tboot because it makes the system
1730 vulnerable to DMA attacks.
1732 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1733 0 disables intel_idle and fall back on acpi_idle.
1734 1 to 9 specify maximum depth of C-state.
1738 Do not enable intel_pstate as the default
1739 scaling driver for the supported processors
1741 Use intel_pstate as a scaling driver, but configure it
1742 to work with generic cpufreq governors (instead of
1743 enabling its internal governor). This mode cannot be
1744 used along with the hardware-managed P-states (HWP)
1747 Enable intel_pstate on systems that prohibit it by default
1748 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1749 instead of acpi-cpufreq may disable platform features, such
1750 as thermal controls and power capping, that rely on ACPI
1751 P-States information being indicated to OSPM and therefore
1752 should be used with caution. This option does not work with
1753 processors that aren't supported by the intel_pstate driver
1754 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1756 Do not enable hardware P state control (HWP)
1759 Only load intel_pstate on systems which support
1760 hardware P state control (HWP) if available.
1762 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1763 Description Table, specifies preferred power management
1764 profile as "Enterprise Server" or "Performance Server",
1765 then this feature is turned on by default.
1767 Allow per-logical-CPU P-State performance control limits using
1768 cpufreq sysfs interface
1770 intremap= [X86-64, Intel-IOMMU]
1771 on enable Interrupt Remapping (default)
1772 off disable Interrupt Remapping
1773 nosid disable Source ID checking
1775 BIOS x2APIC opt-out request will be ignored
1776 nopost disable Interrupt Posting
1778 iomem= Disable strict checking of access to MMIO memory
1779 strict regions from userspace.
1794 nobypass [PPC/POWERNV]
1795 Disable IOMMU bypass, using IOMMU for PCI devices.
1797 iommu.strict= [ARM64] Configure TLB invalidation behaviour
1798 Format: { "0" | "1" }
1800 Request that DMA unmap operations use deferred
1801 invalidation of hardware TLBs, for increased
1802 throughput at the cost of reduced device isolation.
1803 Will fall back to strict mode if not supported by
1804 the relevant IOMMU driver.
1805 1 - Strict mode (default).
1806 DMA unmap operations invalidate IOMMU hardware TLBs
1810 [ARM64] Configure DMA to bypass the IOMMU by default.
1811 Format: { "0" | "1" }
1812 0 - Use IOMMU translation for DMA.
1813 1 - Bypass the IOMMU for DMA.
1814 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
1816 io7= [HW] IO7 for Marvel based alpha systems
1817 See comment before marvel_specify_io7 in
1818 arch/alpha/kernel/core_marvel.c.
1820 io_delay= [X86] I/O delay method
1822 Standard port 0x80 based delay
1824 Alternate port 0xed based delay (needed on some systems)
1826 Simple two microseconds delay
1831 See Documentation/filesystems/nfs/nfsroot.txt.
1833 ipcmni_extend [KNL] Extend the maximum number of unique System V
1834 IPC identifiers from 32,768 to 16,777,216.
1836 irqaffinity= [SMP] Set the default irq affinity mask
1837 The argument is a cpu list, as described above.
1839 irqchip.gicv2_force_probe=
1842 Force the kernel to look for the second 4kB page
1843 of a GICv2 controller even if the memory range
1844 exposed by the device tree is too small.
1846 irqchip.gicv3_nolpi=
1848 Force the kernel to ignore the availability of
1849 LPIs (and by consequence ITSs). Intended for system
1850 that use the kernel as a bootloader, and thus want
1851 to let secondary kernels in charge of setting up
1854 irqchip.gicv3_pseudo_nmi= [ARM64]
1855 Enables support for pseudo-NMIs in the kernel. This
1856 requires the kernel to be built with
1857 CONFIG_ARM64_PSEUDO_NMI.
1860 When an interrupt is not handled search all handlers
1861 for it. Intended to get systems with badly broken
1865 When an interrupt is not handled search all handlers
1866 for it. Also check all handlers each timer
1867 interrupt. Intended to get systems with badly broken
1871 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1873 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
1874 [Deprecated - use cpusets instead]
1875 Format: [flag-list,]<cpu-list>
1877 Specify one or more CPUs to isolate from disturbances
1878 specified in the flag list (default: domain):
1881 Disable the tick when a single task runs.
1883 A residual 1Hz tick is offloaded to workqueues, which you
1884 need to affine to housekeeping through the global
1885 workqueue's affinity configured via the
1886 /sys/devices/virtual/workqueue/cpumask sysfs file, or
1887 by using the 'domain' flag described below.
1889 NOTE: by default the global workqueue runs on all CPUs,
1890 so to protect individual CPUs the 'cpumask' file has to
1891 be configured manually after bootup.
1894 Isolate from the general SMP balancing and scheduling
1895 algorithms. Note that performing domain isolation this way
1896 is irreversible: it's not possible to bring back a CPU to
1897 the domains once isolated through isolcpus. It's strongly
1898 advised to use cpusets instead to disable scheduler load
1899 balancing through the "cpuset.sched_load_balance" file.
1900 It offers a much more flexible interface where CPUs can
1901 move in and out of an isolated set anytime.
1903 You can move a process onto or off an "isolated" CPU via
1904 the CPU affinity syscalls or cpuset.
1905 <cpu number> begins at 0 and the maximum value is
1906 "number of CPUs in system - 1".
1908 The format of <cpu-list> is described above.
1914 ivrs_ioapic [HW,X86_64]
1915 Provide an override to the IOAPIC-ID<->DEVICE-ID
1916 mapping provided in the IVRS ACPI table. For
1917 example, to map IOAPIC-ID decimal 10 to
1918 PCI device 00:14.0 write the parameter as:
1919 ivrs_ioapic[10]=00:14.0
1921 ivrs_hpet [HW,X86_64]
1922 Provide an override to the HPET-ID<->DEVICE-ID
1923 mapping provided in the IVRS ACPI table. For
1924 example, to map HPET-ID decimal 0 to
1925 PCI device 00:14.0 write the parameter as:
1926 ivrs_hpet[0]=00:14.0
1928 ivrs_acpihid [HW,X86_64]
1929 Provide an override to the ACPI-HID:UID<->DEVICE-ID
1930 mapping provided in the IVRS ACPI table. For
1931 example, to map UART-HID:UID AMD0020:0 to
1932 PCI device 00:14.5 write the parameter as:
1933 ivrs_acpihid[00:14.5]=AMD0020:0
1935 js= [HW,JOY] Analog joystick
1936 See Documentation/input/joydev/joystick.rst.
1939 When CONFIG_RANDOMIZE_BASE is set, this disables
1940 kernel and module base offset ASLR (Address Space
1941 Layout Randomization).
1944 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
1945 report on every invalid memory access. Without this
1946 parameter KASAN will print report only for the first
1951 kernelcore= [KNL,X86,IA-64,PPC]
1952 Format: nn[KMGTPE] | nn% | "mirror"
1953 This parameter specifies the amount of memory usable by
1954 the kernel for non-movable allocations. The requested
1955 amount is spread evenly throughout all nodes in the
1956 system as ZONE_NORMAL. The remaining memory is used for
1957 movable memory in its own zone, ZONE_MOVABLE. In the
1958 event, a node is too small to have both ZONE_NORMAL and
1959 ZONE_MOVABLE, kernelcore memory will take priority and
1960 other nodes will have a larger ZONE_MOVABLE.
1962 ZONE_MOVABLE is used for the allocation of pages that
1963 may be reclaimed or moved by the page migration
1964 subsystem. Note that allocations like PTEs-from-HighMem
1965 still use the HighMem zone if it exists, and the Normal
1966 zone if it does not.
1968 It is possible to specify the exact amount of memory in
1969 the form of "nn[KMGTPE]", a percentage of total system
1970 memory in the form of "nn%", or "mirror". If "mirror"
1971 option is specified, mirrored (reliable) memory is used
1972 for non-movable allocations and remaining memory is used
1973 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
1974 are exclusive, so you cannot specify multiple forms.
1976 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1977 Format: <Controller#>[,poll interval]
1978 The controller # is the number of the ehci usb debug
1979 port as it is probed via PCI. The poll interval is
1980 optional and is the number seconds in between
1981 each poll cycle to the debug port in case you need
1982 the functionality for interrupting the kernel with
1983 gdb or control-c on the dbgp connection. When
1984 not using this parameter you use sysrq-g to break into
1985 the kernel debugger.
1987 kgdboc= [KGDB,HW] kgdb over consoles.
1988 Requires a tty driver that supports console polling,
1989 or a supported polling keyboard driver (non-usb).
1990 Serial only format: <serial_device>[,baud]
1991 keyboard only format: kbd
1992 keyboard and serial format: kbd,<serial_device>[,baud]
1993 Optional Kernel mode setting:
1994 kms, kbd format: kms,kbd
1995 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
1997 kgdbwait [KGDB] Stop kernel execution and enter the
1998 kernel debugger at the earliest opportunity.
2000 kmac= [MIPS] korina ethernet MAC address.
2001 Configure the RouterBoard 532 series on-chip
2002 Ethernet adapter MAC address.
2004 kmemleak= [KNL] Boot-time kmemleak enable/disable
2005 Valid arguments: on, off
2007 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2010 kpti= [ARM64] Control page table isolation of user
2011 and kernel address spaces.
2012 Default: enabled on cores which need mitigation.
2016 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2017 Default is 0 (don't ignore, but inject #GP)
2019 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2020 Default is false (don't support).
2022 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
2026 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2027 Default is 1 (enabled)
2029 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2031 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2033 kvm-arm.vgic_v3_group0_trap=
2034 [KVM,ARM] Trap guest accesses to GICv3 group-0
2037 kvm-arm.vgic_v3_group1_trap=
2038 [KVM,ARM] Trap guest accesses to GICv3 group-1
2041 kvm-arm.vgic_v3_common_trap=
2042 [KVM,ARM] Trap guest accesses to GICv3 common
2045 kvm-arm.vgic_v4_enable=
2046 [KVM,ARM] Allow use of GICv4 for direct injection of
2049 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2050 (virtualized MMU) support on capable Intel chips.
2051 Default is 1 (enabled)
2053 kvm-intel.emulate_invalid_guest_state=
2054 [KVM,Intel] Enable emulation of invalid guest states
2055 Default is 0 (disabled)
2057 kvm-intel.flexpriority=
2058 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2059 Default is 1 (enabled)
2062 [KVM,Intel] Enable VMX nesting (nVMX).
2063 Default is 0 (disabled)
2065 kvm-intel.unrestricted_guest=
2066 [KVM,Intel] Disable unrestricted guest feature
2067 (virtualized real and unpaged mode) on capable
2068 Intel chips. Default is 1 (enabled)
2070 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2073 Valid arguments: never, cond, always
2075 always: L1D cache flush on every VMENTER.
2076 cond: Flush L1D on VMENTER only when the code between
2077 VMEXIT and VMENTER can leak host memory.
2078 never: Disables the mitigation
2080 Default is cond (do L1 cache flush in specific instances)
2082 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2083 feature (tagged TLBs) on capable Intel chips.
2084 Default is 1 (enabled)
2086 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2089 The kernel PTE inversion protection is unconditionally
2090 enabled and cannot be disabled.
2093 Provides all available mitigations for the
2094 L1TF vulnerability. Disables SMT and
2095 enables all mitigations in the
2096 hypervisors, i.e. unconditional L1D flush.
2098 SMT control and L1D flush control via the
2099 sysfs interface is still possible after
2100 boot. Hypervisors will issue a warning
2101 when the first VM is started in a
2102 potentially insecure configuration,
2103 i.e. SMT enabled or L1D flush disabled.
2106 Same as 'full', but disables SMT and L1D
2107 flush runtime control. Implies the
2108 'nosmt=force' command line option.
2109 (i.e. sysfs control of SMT is disabled.)
2112 Leaves SMT enabled and enables the default
2113 hypervisor mitigation, i.e. conditional
2116 SMT control and L1D flush control via the
2117 sysfs interface is still possible after
2118 boot. Hypervisors will issue a warning
2119 when the first VM is started in a
2120 potentially insecure configuration,
2121 i.e. SMT enabled or L1D flush disabled.
2125 Disables SMT and enables the default
2126 hypervisor mitigation.
2128 SMT control and L1D flush control via the
2129 sysfs interface is still possible after
2130 boot. Hypervisors will issue a warning
2131 when the first VM is started in a
2132 potentially insecure configuration,
2133 i.e. SMT enabled or L1D flush disabled.
2136 Same as 'flush', but hypervisors will not
2137 warn when a VM is started in a potentially
2138 insecure configuration.
2141 Disables hypervisor mitigations and doesn't
2143 It also drops the swap size and available
2144 RAM limit restriction on both hypervisor and
2149 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2155 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2158 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2159 value for LAPIC timer one-shot implementation. Default
2160 back to the programmable timer unit in the LAPIC.
2162 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2165 libata.dma= [LIBATA] DMA control
2166 libata.dma=0 Disable all PATA and SATA DMA
2167 libata.dma=1 PATA and SATA Disk DMA only
2168 libata.dma=2 ATAPI (CDROM) DMA only
2169 libata.dma=4 Compact Flash DMA only
2170 Combinations also work, so libata.dma=3 enables DMA
2171 for disks and CDROMs, but not CFs.
2173 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2174 libata.ignore_hpa=0 keep BIOS limits (default)
2175 libata.ignore_hpa=1 ignore limits, using full disk
2177 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2181 libata.force= [LIBATA] Force configurations. The format is comma
2182 separated list of "[ID:]VAL" where ID is
2183 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2184 matching port, link or device. Basically, it matches
2185 the ATA ID string printed on console by libata. If
2186 the whole ID part is omitted, the last PORT and DEVICE
2187 values are used. If ID hasn't been specified yet, the
2188 configuration applies to all ports, links and devices.
2190 If only DEVICE is omitted, the parameter applies to
2191 the port and all links and devices behind it. DEVICE
2192 number of 0 either selects the first device or the
2193 first fan-out link behind PMP device. It does not
2194 select the host link. DEVICE number of 15 selects the
2195 host link and device attached to it.
2197 The VAL specifies the configuration to force. As long
2198 as there's no ambiguity shortcut notation is allowed.
2199 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2200 The following configurations can be forced.
2202 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2203 Any ID with matching PORT is used.
2205 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2207 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2208 udma[/][16,25,33,44,66,100,133] notation is also
2211 * [no]ncq: Turn on or off NCQ.
2213 * [no]ncqtrim: Turn off queued DSM TRIM.
2215 * nohrst, nosrst, norst: suppress hard, soft
2218 * rstonce: only attempt one reset during
2219 hot-unplug link recovery
2221 * dump_id: dump IDENTIFY data.
2223 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2225 * disable: Disable this device.
2227 If there are multiple matching configurations changing
2228 the same attribute, the last one is used.
2230 memblock=debug [KNL] Enable memblock debug messages.
2232 load_ramdisk= [RAM] List of ramdisks to load from floppy
2233 See Documentation/blockdev/ramdisk.txt.
2235 lockd.nlm_grace_period=P [NFS] Assign grace period.
2238 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2241 lockd.nlm_timeout=T [NFS] Assign timeout value.
2244 lockd.nlm_udpport=M [NFS] Assign UDP port.
2247 locktorture.nreaders_stress= [KNL]
2248 Set the number of locking read-acquisition kthreads.
2249 Defaults to being automatically set based on the
2250 number of online CPUs.
2252 locktorture.nwriters_stress= [KNL]
2253 Set the number of locking write-acquisition kthreads.
2255 locktorture.onoff_holdoff= [KNL]
2256 Set time (s) after boot for CPU-hotplug testing.
2258 locktorture.onoff_interval= [KNL]
2259 Set time (s) between CPU-hotplug operations, or
2260 zero to disable CPU-hotplug testing.
2262 locktorture.shuffle_interval= [KNL]
2263 Set task-shuffle interval (jiffies). Shuffling
2264 tasks allows some CPUs to go into dyntick-idle
2265 mode during the locktorture test.
2267 locktorture.shutdown_secs= [KNL]
2268 Set time (s) after boot system shutdown. This
2269 is useful for hands-off automated testing.
2271 locktorture.stat_interval= [KNL]
2272 Time (s) between statistics printk()s.
2274 locktorture.stutter= [KNL]
2275 Time (s) to stutter testing, for example,
2276 specifying five seconds causes the test to run for
2277 five seconds, wait for five seconds, and so on.
2278 This tests the locking primitive's ability to
2279 transition abruptly to and from idle.
2281 locktorture.torture_type= [KNL]
2282 Specify the locking implementation to test.
2284 locktorture.verbose= [KNL]
2285 Enable additional printk() statements.
2287 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2290 loglevel= All Kernel Messages with a loglevel smaller than the
2291 console loglevel will be printed to the console. It can
2292 also be changed with klogd or other programs. The
2293 loglevels are defined as follows:
2295 0 (KERN_EMERG) system is unusable
2296 1 (KERN_ALERT) action must be taken immediately
2297 2 (KERN_CRIT) critical conditions
2298 3 (KERN_ERR) error conditions
2299 4 (KERN_WARNING) warning conditions
2300 5 (KERN_NOTICE) normal but significant condition
2301 6 (KERN_INFO) informational
2302 7 (KERN_DEBUG) debug-level messages
2304 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2305 in bytes. n must be a power of two and greater
2306 than the minimal size. The minimal size is defined
2307 by LOG_BUF_SHIFT kernel config parameter. There is
2308 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2309 that allows to increase the default size depending on
2310 the number of CPUs. See init/Kconfig for more details.
2312 logo.nologo [FB] Disables display of the built-in Linux logo.
2313 This may be used to provide more screen space for
2314 kernel log messages and is useful when debugging
2315 kernel boot problems.
2317 lp=0 [LP] Specify parallel ports to use, e.g,
2318 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2319 lp=reset first parallel port). 'lp=0' disables the
2320 lp=auto printer driver. 'lp=reset' (which can be
2321 specified in addition to the ports) causes
2322 attached printers to be reset. Using
2323 lp=port1,port2,... specifies the parallel ports
2324 to associate lp devices with, starting with
2325 lp0. A port specification may be 'none' to skip
2326 that lp device, or a parport name such as
2327 'parport0'. Specifying 'lp=auto' instead of a
2328 port specification list means that device IDs
2329 from each port should be examined, to see if
2330 an IEEE 1284-compliant printer is attached; if
2331 so, the driver will manage that printer.
2332 See also header of drivers/char/lp.c.
2335 Sets loops_per_jiffy to given constant, thus avoiding
2336 time-consuming boot-time autodetection (up to 250 ms per
2337 CPU). 0 enables autodetection (default). To determine
2338 the correct value for your kernel, boot with normal
2339 autodetection and see what value is printed. Note that
2340 on SMP systems the preset will be applied to all CPUs,
2341 which is likely to cause problems if your CPUs need
2342 significantly divergent settings. An incorrect value
2343 will cause delays in the kernel to be wrong, leading to
2344 unpredictable I/O errors and other breakage. Although
2345 unlikely, in the extreme case this might damage your
2349 Format: <io>,<irq>,<dma>
2351 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2354 [SECURITY] Choose order of LSM initialization. This
2355 overrides CONFIG_LSM, and the "security=" parameter.
2357 machvec= [IA-64] Force the use of a particular machine-vector
2358 (machvec) in a generic kernel.
2359 Example: machvec=hpzx1_swiotlb
2361 machtype= [Loongson] Share the same kernel image file between different
2363 Example: machtype=lemote-yeeloong-2f-7inch
2365 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2366 than or equal to this physical address is ignored.
2368 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2369 will bring up during bootup. maxcpus=n : n >= 0 limits
2370 the kernel to bring up 'n' processors. Surely after
2371 bootup you can bring up the other plugged cpu by executing
2372 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2373 only takes effect during system bootup.
2374 While n=0 is a special case, it is equivalent to "nosmp",
2375 which also disables the IO APIC.
2377 max_loop= [LOOP] The number of loop block devices that get
2378 (loop.max_loop) unconditionally pre-created at init time. The default
2379 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2380 of statically allocating a predefined number, loop
2381 devices can be requested on-demand with the
2382 /dev/loop-control interface.
2384 mce [X86-32] Machine Check Exception
2386 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.txt
2388 md= [HW] RAID subsystems devices and level
2389 See Documentation/admin-guide/md.rst.
2392 Format: <first>,<last>
2393 Specifies range of consoles to be captured by the MDA.
2396 Control mitigation for the Micro-architectural Data
2397 Sampling (MDS) vulnerability.
2399 Certain CPUs are vulnerable to an exploit against CPU
2400 internal buffers which can forward information to a
2401 disclosure gadget under certain conditions.
2403 In vulnerable processors, the speculatively
2404 forwarded data can be used in a cache side channel
2405 attack, to access data to which the attacker does
2406 not have direct access.
2408 This parameter controls the MDS mitigation. The
2411 full - Enable MDS mitigation on vulnerable CPUs
2412 full,nosmt - Enable MDS mitigation and disable
2413 SMT on vulnerable CPUs
2414 off - Unconditionally disable MDS mitigation
2416 Not specifying this option is equivalent to
2419 For details see: Documentation/admin-guide/hw-vuln/mds.rst
2421 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2422 Amount of memory to be used when the kernel is not able
2423 to see the whole system memory or for test.
2424 [X86] Work as limiting max address. Use together
2425 with memmap= to avoid physical address space collisions.
2426 Without memmap= PCI devices could be placed at addresses
2427 belonging to unused RAM.
2429 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2433 [KNL,SH] Allow user to override the default size for
2434 per-device physically contiguous DMA buffers.
2436 memhp_default_state=online/offline
2437 [KNL] Set the initial state for the memory hotplug
2438 onlining policy. If not specified, the default value is
2439 set according to the
2440 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2442 See Documentation/memory-hotplug.txt.
2444 memmap=exactmap [KNL,X86] Enable setting of an exact
2445 E820 memory map, as specified by the user.
2446 Such memmap=exactmap lines can be constructed based on
2447 BIOS output or other requirements. See the memmap=nn@ss
2450 memmap=nn[KMG]@ss[KMG]
2451 [KNL] Force usage of a specific region of memory.
2452 Region of memory to be used is from ss to ss+nn.
2453 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2454 which limits max address to nn[KMG].
2455 Multiple different regions can be specified,
2458 memmap=100M@2G,100M#3G,1G!1024G
2460 memmap=nn[KMG]#ss[KMG]
2461 [KNL,ACPI] Mark specific memory as ACPI data.
2462 Region of memory to be marked is from ss to ss+nn.
2464 memmap=nn[KMG]$ss[KMG]
2465 [KNL,ACPI] Mark specific memory as reserved.
2466 Region of memory to be reserved is from ss to ss+nn.
2467 Example: Exclude memory from 0x18690000-0x1869ffff
2468 memmap=64K$0x18690000
2470 memmap=0x10000$0x18690000
2471 Some bootloaders may need an escape character before '$',
2472 like Grub2, otherwise '$' and the following number
2475 memmap=nn[KMG]!ss[KMG]
2476 [KNL,X86] Mark specific memory as protected.
2477 Region of memory to be used, from ss to ss+nn.
2478 The memory region may be marked as e820 type 12 (0xc)
2479 and is NVDIMM or ADR memory.
2481 memmap=<size>%<offset>-<oldtype>+<newtype>
2482 [KNL,ACPI] Convert memory within the specified region
2483 from <oldtype> to <newtype>. If "-<oldtype>" is left
2484 out, the whole region will be marked as <newtype>,
2485 even if previously unavailable. If "+<newtype>" is left
2486 out, matching memory will be removed. Types are
2487 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2488 3 = ACPI, 12 = PRAM.
2490 memory_corruption_check=0/1 [X86]
2491 Some BIOSes seem to corrupt the first 64k of
2492 memory when doing things like suspend/resume.
2493 Setting this option will scan the memory
2494 looking for corruption. Enabling this will
2495 both detect corruption and prevent the kernel
2496 from using the memory being corrupted.
2497 However, its intended as a diagnostic tool; if
2498 repeatable BIOS-originated corruption always
2499 affects the same memory, you can use memmap=
2500 to prevent the kernel from using that memory.
2502 memory_corruption_check_size=size [X86]
2503 By default it checks for corruption in the low
2504 64k, making this memory unavailable for normal
2505 use. Use this parameter to scan for
2506 corruption in more or less memory.
2508 memory_corruption_check_period=seconds [X86]
2509 By default it checks for corruption every 60
2510 seconds. Use this parameter to check at some
2511 other rate. 0 disables periodic checking.
2513 memtest= [KNL,X86,ARM,PPC] Enable memtest
2515 default : 0 <disable>
2516 Specifies the number of memtest passes to be
2517 performed. Each pass selects another test
2518 pattern from a given set of patterns. Memtest
2519 fills the memory with this pattern, validates
2520 memory contents and reserves bad memory
2521 regions that are detected.
2523 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2524 Valid arguments: on, off
2525 Default (depends on kernel configuration option):
2526 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2527 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2528 mem_encrypt=on: Activate SME
2529 mem_encrypt=off: Do not activate SME
2531 Refer to Documentation/x86/amd-memory-encryption.txt
2532 for details on when memory encryption can be activated.
2534 mem_sleep_default= [SUSPEND] Default system suspend mode:
2535 s2idle - Suspend-To-Idle
2536 shallow - Power-On Suspend or equivalent (if supported)
2537 deep - Suspend-To-RAM or equivalent (if supported)
2538 See Documentation/admin-guide/pm/sleep-states.rst.
2540 meye.*= [HW] Set MotionEye Camera parameters
2541 See Documentation/media/v4l-drivers/meye.rst.
2543 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2544 Multi-Function General Purpose Timers on AMD Geode
2547 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2548 the BIOS has incorrectly applied a workaround. TinyBIOS
2549 version 0.98 is known to be affected, 0.99 fixes the
2550 problem by letting the user disable the workaround.
2554 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2555 physical address is ignored.
2557 mini2440= [ARM,HW,KNL]
2558 Format:[0..2][b][c][t]
2560 MINI2440 configuration specification:
2561 0 - The attached screen is the 3.5" TFT
2562 1 - The attached screen is the 7" TFT
2563 2 - The VGA Shield is attached (1024x768)
2564 Leaving out the screen size parameter will not load
2565 the TFT driver, and the framebuffer will be left
2567 b - Enable backlight. The TFT backlight pin will be
2568 linked to the kernel VESA blanking code and a GPIO
2569 LED. This parameter is not necessary when using the
2571 c - Enable the s3c camera interface.
2572 t - Reserved for enabling touchscreen support. The
2573 touchscreen support is not enabled in the mainstream
2574 kernel as of 2.6.30, a preliminary port can be found
2575 in the "bleeding edge" mini2440 support kernel at
2576 http://repo.or.cz/w/linux-2.6/mini2440.git
2579 [X86,PPC,S390,ARM64] Control optional mitigations for
2580 CPU vulnerabilities. This is a set of curated,
2581 arch-independent options, each of which is an
2582 aggregation of existing arch-specific options.
2585 Disable all optional CPU mitigations. This
2586 improves system performance, but it may also
2587 expose users to several CPU vulnerabilities.
2588 Equivalent to: nopti [X86,PPC]
2592 nospectre_v2 [X86,PPC,S390,ARM64]
2593 spectre_v2_user=off [X86]
2594 spec_store_bypass_disable=off [X86,PPC]
2595 ssbd=force-off [ARM64]
2600 Mitigate all CPU vulnerabilities, but leave SMT
2601 enabled, even if it's vulnerable. This is for
2602 users who don't want to be surprised by SMT
2603 getting disabled across kernel upgrades, or who
2604 have other ways of avoiding SMT-based attacks.
2605 Equivalent to: (default behavior)
2608 Mitigate all CPU vulnerabilities, disabling SMT
2609 if needed. This is for users who always want to
2610 be fully mitigated, even if it means losing SMT.
2611 Equivalent to: l1tf=flush,nosmt [X86]
2612 mds=full,nosmt [X86]
2615 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2616 parameter allows control of the logging verbosity for
2617 the additional memory initialisation checks. A value
2618 of 0 disables mminit logging and a level of 4 will
2619 log everything. Information is printed at KERN_DEBUG
2620 so loglevel=8 may also need to be specified.
2623 [KNL] When CONFIG_MODULE_SIG is set, this means that
2624 modules without (valid) signatures will fail to load.
2625 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2626 is always true, so this option does nothing.
2628 module_blacklist= [KNL] Do not load a comma-separated list of
2629 modules. Useful for debugging problem modules.
2632 [MOUSE] Maximum time between finger touching and
2633 leaving touchpad surface for touch to be considered
2634 a tap and be reported as a left button click (for
2635 touchpads working in absolute mode only).
2637 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2638 reporting absolute coordinates, such as tablets
2639 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2640 reporting absolute coordinates, such as tablets
2642 movablecore= [KNL,X86,IA-64,PPC]
2643 Format: nn[KMGTPE] | nn%
2644 This parameter is the complement to kernelcore=, it
2645 specifies the amount of memory used for migratable
2646 allocations. If both kernelcore and movablecore is
2647 specified, then kernelcore will be at *least* the
2648 specified value but may be more. If movablecore on its
2649 own is specified, the administrator must be careful
2650 that the amount of memory usable for all allocations
2653 movable_node [KNL] Boot-time switch to make hotplugable memory
2654 NUMA nodes to be movable. This means that the memory
2655 of such nodes will be usable only for movable
2656 allocations which rules out almost all kernel
2657 allocations. Use with caution!
2659 MTD_Partition= [MTD]
2660 Format: <name>,<region-number>,<size>,<offset>
2662 MTD_Region= [MTD] Format:
2663 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2666 See drivers/mtd/cmdlinepart.c.
2668 multitce=off [PPC] This parameter disables the use of the pSeries
2669 firmware feature for updating multiple TCE entries
2672 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2674 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2676 boundary - index of last SLC block on Flex-OneNAND.
2677 The remaining blocks are configured as MLC blocks.
2678 lock - Configure if Flex-OneNAND boundary should be locked.
2679 Once locked, the boundary cannot be changed.
2680 1 indicates lock status, 0 indicates unlock status.
2683 ARM/S3C2412 JIVE boot control
2685 See arch/arm/mach-s3c2412/mach-jive.c
2687 mtouchusb.raw_coordinates=
2688 [HW] Make the MicroTouch USB driver use raw coordinates
2689 ('y', default) or cooked coordinates ('n')
2691 mtrr_chunk_size=nn[KMG] [X86]
2692 used for mtrr cleanup. It is largest continuous chunk
2693 that could hold holes aka. UC entries.
2695 mtrr_gran_size=nn[KMG] [X86]
2696 Used for mtrr cleanup. It is granularity of mtrr block.
2698 Large value could prevent small alignment from
2701 mtrr_spare_reg_nr=n [X86]
2703 Range: 0,7 : spare reg number
2705 Used for mtrr cleanup. It is spare mtrr entries number.
2706 Set to 2 or more if your graphical card needs more.
2708 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2710 netdev= [NET] Network devices parameters
2711 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2712 Note that mem_start is often overloaded to mean
2713 something different and driver-specific.
2714 This usage is only documented in each driver source
2718 [NETFILTER] Enable connection tracking flow accounting
2719 0 to disable accounting
2720 1 to enable accounting
2723 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2724 See Documentation/filesystems/nfs/nfsroot.txt.
2726 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2727 See Documentation/filesystems/nfs/nfsroot.txt.
2729 nfsrootdebug [NFS] enable nfsroot debugging messages.
2730 See Documentation/filesystems/nfs/nfsroot.txt.
2732 nfs.callback_nr_threads=
2733 [NFSv4] set the total number of threads that the
2734 NFS client will assign to service NFSv4 callback
2737 nfs.callback_tcpport=
2738 [NFS] set the TCP port on which the NFSv4 callback
2739 channel should listen.
2742 [NFS] sets the pathname to the program which is used
2743 to update the NFS client cache entries.
2745 nfs.cache_getent_timeout=
2746 [NFS] sets the timeout after which an attempt to
2747 update a cache entry is deemed to have failed.
2749 nfs.idmap_cache_timeout=
2750 [NFS] set the maximum lifetime for idmapper cache
2754 [NFS] enable 64-bit inode numbers.
2755 If zero, the NFS client will fake up a 32-bit inode
2756 number for the readdir() and stat() syscalls instead
2757 of returning the full 64-bit number.
2758 The default is to return 64-bit inode numbers.
2760 nfs.max_session_cb_slots=
2761 [NFSv4.1] Sets the maximum number of session
2762 slots the client will assign to the callback
2763 channel. This determines the maximum number of
2764 callbacks the client will process in parallel for
2765 a particular server.
2767 nfs.max_session_slots=
2768 [NFSv4.1] Sets the maximum number of session slots
2769 the client will attempt to negotiate with the server.
2770 This limits the number of simultaneous RPC requests
2771 that the client can send to the NFSv4.1 server.
2772 Note that there is little point in setting this
2773 value higher than the max_tcp_slot_table_limit.
2775 nfs.nfs4_disable_idmapping=
2776 [NFSv4] When set to the default of '1', this option
2777 ensures that both the RPC level authentication
2778 scheme and the NFS level operations agree to use
2779 numeric uids/gids if the mount is using the
2780 'sec=sys' security flavour. In effect it is
2781 disabling idmapping, which can make migration from
2782 legacy NFSv2/v3 systems to NFSv4 easier.
2783 Servers that do not support this mode of operation
2784 will be autodetected by the client, and it will fall
2785 back to using the idmapper.
2786 To turn off this behaviour, set the value to '0'.
2788 [NFS4] Specify an additional fixed unique ident-
2789 ification string that NFSv4 clients can insert into
2790 their nfs_client_id4 string. This is typically a
2791 UUID that is generated at system install time.
2793 nfs.send_implementation_id =
2794 [NFSv4.1] Send client implementation identification
2795 information in exchange_id requests.
2796 If zero, no implementation identification information
2798 The default is to send the implementation identification
2801 nfs.recover_lost_locks =
2802 [NFSv4] Attempt to recover locks that were lost due
2803 to a lease timeout on the server. Please note that
2804 doing this risks data corruption, since there are
2805 no guarantees that the file will remain unchanged
2806 after the locks are lost.
2807 If you want to enable the kernel legacy behaviour of
2808 attempting to recover these locks, then set this
2810 The default parameter value of '0' causes the kernel
2811 not to attempt recovery of lost locks.
2813 nfs4.layoutstats_timer =
2814 [NFSv4.2] Change the rate at which the kernel sends
2815 layoutstats to the pNFS metadata server.
2817 Setting this to value to 0 causes the kernel to use
2818 whatever value is the default set by the layout
2819 driver. A non-zero value sets the minimum interval
2820 in seconds between layoutstats transmissions.
2822 nfsd.nfs4_disable_idmapping=
2823 [NFSv4] When set to the default of '1', the NFSv4
2824 server will return only numeric uids and gids to
2825 clients using auth_sys, and will accept numeric uids
2826 and gids from such clients. This is intended to ease
2827 migration from NFSv2/v3.
2829 nmi_debug= [KNL,SH] Specify one or more actions to take
2830 when a NMI is triggered.
2831 Format: [state][,regs][,debounce][,die]
2833 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2834 Format: [panic,][nopanic,][num]
2836 0 - turn hardlockup detector in nmi_watchdog off
2837 1 - turn hardlockup detector in nmi_watchdog on
2838 When panic is specified, panic when an NMI watchdog
2839 timeout occurs (or 'nopanic' to override the opposite
2840 default). To disable both hard and soft lockup detectors,
2841 please see 'nowatchdog'.
2842 This is useful when you use a panic=... timeout and
2843 need the box quickly up again.
2845 These settings can be accessed at runtime via
2846 the nmi_watchdog and hardlockup_panic sysctls.
2848 netpoll.carrier_timeout=
2849 [NET] Specifies amount of time (in seconds) that
2850 netpoll should wait for a carrier. By default netpoll
2853 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2854 emulation library even if a 387 maths coprocessor
2857 no5lvl [X86-64] Disable 5-level paging mode. Forces
2858 kernel to use 4-level paging instead.
2861 [HW] Never suspend the console
2862 Disable suspending of consoles during suspend and
2863 hibernate operations. Once disabled, debugging
2864 messages can reach various consoles while the rest
2865 of the system is being put to sleep (ie, while
2866 debugging driver suspend/resume hooks). This may
2867 not work reliably with all consoles, but is known
2868 to work with serial and VGA consoles.
2869 To facilitate more flexible debugging, we also add
2870 console_suspend, a printk module parameter to control
2871 it. Users could use console_suspend (usually
2872 /sys/module/printk/parameters/console_suspend) to
2873 turn on/off it dynamically.
2875 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2876 caches in the slab allocator. Saves per-node memory,
2877 but will impact performance.
2881 noaltinstr [S390] Disables alternative instructions patching
2882 (CPU alternatives feature).
2884 noapic [SMP,APIC] Tells the kernel to not make use of any
2885 IOAPICs that may be present in the system.
2887 noautogroup Disable scheduler automatic task group creation.
2889 nobats [PPC] Do not use BATs for mapping kernel lowmem
2890 on "Classic" PPC cores.
2894 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2896 nodelayacct [KNL] Disable per-task delay accounting
2898 nodsp [SH] Disable hardware DSP at boot time.
2900 noefi Disable EFI runtime services support.
2905 On X86-32 available only on PAE configured kernels.
2906 noexec=on: enable non-executable mappings (default)
2907 noexec=off: disable non-executable mappings
2910 Disable SMAP (Supervisor Mode Access Prevention)
2911 even if it is supported by processor.
2914 Disable SMEP (Supervisor Mode Execution Prevention)
2915 even if it is supported by processor.
2918 This affects only 32-bit executables.
2919 noexec32=on: enable non-executable mappings (default)
2920 read doesn't imply executable mappings
2921 noexec32=off: disable non-executable mappings
2922 read implies executable mappings
2924 nofpu [MIPS,SH] Disable hardware FPU at boot time.
2926 nofxsr [BUGS=X86-32] Disables x86 floating point extended
2927 register save and restore. The kernel will only save
2928 legacy floating-point registers on task switch.
2930 nohugeiomap [KNL,x86] Disable kernel huge I/O mappings.
2932 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
2933 Equivalent to smt=1.
2935 [KNL,x86] Disable symmetric multithreading (SMT).
2936 nosmt=force: Force disable SMT, cannot be undone
2937 via the sysfs control file.
2939 nospectre_v1 [PPC] Disable mitigations for Spectre Variant 1 (bounds
2940 check bypass). With this option data leaks are possible
2943 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for
2944 the Spectre variant 2 (indirect branch prediction)
2945 vulnerability. System may allow data leaks with this
2948 nospec_store_bypass_disable
2949 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
2951 noxsave [BUGS=X86] Disables x86 extended register state save
2952 and restore using xsave. The kernel will fallback to
2953 enabling legacy floating-point and sse state.
2955 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
2956 register states. The kernel will fall back to use
2957 xsave to save the states. By using this parameter,
2958 performance of saving the states is degraded because
2959 xsave doesn't support modified optimization while
2960 xsaveopt supports it on xsaveopt enabled systems.
2962 noxsaves [X86] Disables xsaves and xrstors used in saving and
2963 restoring x86 extended register state in compacted
2964 form of xsave area. The kernel will fall back to use
2965 xsaveopt and xrstor to save and restore the states
2966 in standard form of xsave area. By using this
2967 parameter, xsave area per process might occupy more
2968 memory on xsaves enabled systems.
2970 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
2971 wfi(ARM) instruction doesn't work correctly and not to
2972 use it. This is also useful when using JTAG debugger.
2974 no_file_caps Tells the kernel not to honor file capabilities. The
2975 only way then for a file to be executed with privilege
2976 is to be setuid root or executed by root.
2978 nohalt [IA-64] Tells the kernel not to use the power saving
2979 function PAL_HALT_LIGHT when idle. This increases
2980 power-consumption. On the positive side, it reduces
2981 interrupt wake-up latency, which may improve performance
2982 in certain environments such as networked servers or
2985 nohibernate [HIBERNATION] Disable hibernation and resume.
2987 nohz= [KNL] Boottime enable/disable dynamic ticks
2988 Valid arguments: on, off
2991 nohz_full= [KNL,BOOT,SMP,ISOL]
2992 The argument is a cpu list, as described above.
2993 In kernels built with CONFIG_NO_HZ_FULL=y, set
2994 the specified list of CPUs whose tick will be stopped
2995 whenever possible. The boot CPU will be forced outside
2996 the range to maintain the timekeeping. Any CPUs
2997 in this list will have their RCU callbacks offloaded,
2998 just as if they had also been called out in the
2999 rcu_nocbs= boot parameter.
3001 noiotrap [SH] Disables trapped I/O port accesses.
3003 noirqdebug [X86-32] Disables the code which attempts to detect and
3004 disable unhandled interrupt sources.
3006 no_timer_check [X86,APIC] Disables the code which tests for
3007 broken timer IRQ sources.
3009 noisapnp [ISAPNP] Disables ISA PnP code.
3011 noinitrd [RAM] Tells the kernel not to load any configured
3014 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3016 [Deprecated - use intremap=off]
3020 noinvpcid [X86] Disable the INVPCID cpu feature.
3022 nojitter [IA-64] Disables jitter checking for ITC timers.
3024 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3026 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3030 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3031 clock and use the default one.
3033 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
3034 steal time is computed, but won't influence scheduler
3037 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3039 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3041 noltlbs [PPC] Do not use large page/tlb entries for kernel
3042 lowmem mapping on PPC40x and PPC8xx
3044 nomca [IA-64] Disable machine check abort handling
3046 nomce [X86-32] Disable Machine Check Exception
3048 nomfgpt [X86-32] Disable Multi-Function General Purpose
3049 Timer usage (for AMD Geode machines).
3051 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3052 shutdown the other cpus. Instead use the REBOOT_VECTOR
3055 nomodule Disable module load
3057 nopat [X86] Disable PAT (page attribute table extension of
3058 pagetables) support.
3060 nopcid [X86-64] Disable the PCID cpu feature.
3062 norandmaps Don't use address space randomization. Equivalent to
3063 echo 0 > /proc/sys/kernel/randomize_va_space
3065 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3066 with UP alternatives
3068 nordrand [X86] Disable kernel use of the RDRAND and
3069 RDSEED instructions even if they are supported
3070 by the processor. RDRAND and RDSEED are still
3071 available to user space applications.
3073 noresume [SWSUSP] Disables resume and restores original swap
3076 no-scroll [VGA] Disables scrollback.
3077 This is required for the Braillex ib80-piezo Braille
3078 reader made by F.H. Papenmeier (Germany).
3082 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
3084 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3085 and disable the IO APIC. legacy for "maxcpus=0".
3087 nosoftlockup [KNL] Disable the soft-lockup detector.
3089 nosync [HW,M68K] Disables sync negotiation for all devices.
3091 nowatchdog [KNL] Disable both lockup detectors, i.e.
3092 soft-lockup and NMI watchdog (hard-lockup).
3096 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3098 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
3099 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
3100 Some features depend on CPU0. Known dependencies are:
3101 1. Resume from suspend/hibernate depends on CPU0.
3102 Suspend/hibernate will fail if CPU0 is offline and you
3103 need to online CPU0 before suspend/hibernate.
3104 2. PIC interrupts also depend on CPU0. CPU0 can't be
3105 removed if a PIC interrupt is detected.
3106 It's said poweroff/reboot may depend on CPU0 on some
3107 machines although I haven't seen such issues so far
3108 after CPU0 is offline on a few tested machines.
3109 If the dependencies are under your control, you can
3110 turn on cpu0_hotplug.
3112 nps_mtm_hs_ctr= [KNL,ARC]
3113 This parameter sets the maximum duration, in
3114 cycles, each HW thread of the CTOP can run
3115 without interruptions, before HW switches it.
3116 The actual maximum duration is 16 times this
3118 Format: integer between 1 and 255
3121 nptcg= [IA-64] Override max number of concurrent global TLB
3122 purges which is reported from either PAL_VM_SUMMARY or
3125 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3126 could support. nr_cpus=n : n >= 1 limits the kernel to
3127 support 'n' processors. It could be larger than the
3128 number of already plugged CPU during bootup, later in
3129 runtime you can physically add extra cpu until it reaches
3130 n. So during boot up some boot time memory for per-cpu
3131 variables need be pre-allocated for later physical cpu
3134 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3136 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
3137 Allowed values are enable and disable
3139 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3140 'node', 'default' can be specified
3141 This can be set from sysctl after boot.
3142 See Documentation/sysctl/vm.txt for details.
3144 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3145 See Documentation/debugging-via-ohci1394.txt for more
3148 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3149 Rather than timing out after 20 ms if an EC
3150 command is not properly ACKed, override the length
3151 of the timeout. We have interrupts disabled while
3152 waiting for the ACK, so if this is set too high
3153 interrupts *may* be lost!
3155 omap_mux= [OMAP] Override bootloader pin multiplexing.
3156 Format: <mux_mode0.mode_name=value>...
3157 For example, to override I2C bus2:
3158 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3160 oprofile.timer= [HW]
3161 Use timer interrupt instead of performance counters
3163 oprofile.cpu_type= Force an oprofile cpu type
3164 This might be useful if you have an older oprofile
3165 userland or if you want common events.
3166 Format: { arch_perfmon }
3167 arch_perfmon: [X86] Force use of architectural
3168 perfmon on Intel CPUs instead of the
3169 CPU specific event set.
3170 timer: [X86] Force use of architectural NMI
3171 timer mode (see also oprofile.timer
3172 for generic hr timer mode)
3174 oops=panic Always panic on oopses. Default is to just kill the
3175 process, but there is a small probability of
3176 deadlocking the machine.
3177 This will also cause panics on machine check exceptions.
3178 Useful together with panic=30 to trigger a reboot.
3181 [KNL] Boolean flag to control whether the page allocator
3182 should randomize its free lists. The randomization may
3183 be automatically enabled if the kernel detects it is
3184 running on a platform with a direct-mapped memory-side
3185 cache, and this parameter can be used to
3186 override/disable that behavior. The state of the flag
3187 can be read from sysfs at:
3188 /sys/module/page_alloc/parameters/shuffle.
3190 page_owner= [KNL] Boot-time page_owner enabling option.
3191 Storage of the information about who allocated
3192 each page is disabled in default. With this switch,
3194 on: enable the feature
3196 page_poison= [KNL] Boot-time parameter changing the state of
3197 poisoning on the buddy allocator, available with
3198 CONFIG_PAGE_POISONING=y.
3199 off: turn off poisoning (default)
3200 on: turn on poisoning
3202 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3203 timeout > 0: seconds before rebooting
3204 timeout = 0: wait forever
3205 timeout < 0: reboot immediately
3208 panic_print= Bitmask for printing system info when panic happens.
3209 User can chose combination of the following bits:
3210 bit 0: print all tasks info
3211 bit 1: print system memory info
3212 bit 2: print timer info
3213 bit 3: print locks info if CONFIG_LOCKDEP is on
3214 bit 4: print ftrace buffer
3215 bit 5: print all printk messages in buffer
3217 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3220 crash_kexec_post_notifiers
3221 Run kdump after running panic-notifiers and dumping
3222 kmsg. This only for the users who doubt kdump always
3223 succeeds in any situation.
3224 Note that this also increases risks of kdump failure,
3225 because some panic notifiers can make the crashed
3226 kernel more unstable.
3228 parkbd.port= [HW] Parallel port number the keyboard adapter is
3229 connected to, default is 0.
3231 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3232 0 for XT, 1 for AT (default is AT).
3235 parport= [HW,PPT] Specify parallel ports. 0 disables.
3236 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3237 Use 'auto' to force the driver to use any
3238 IRQ/DMA settings detected (the default is to
3239 ignore detected IRQ/DMA settings because of
3240 possible conflicts). You can specify the base
3241 address, IRQ, and DMA settings; IRQ and DMA
3242 should be numbers, or 'auto' (for using detected
3243 settings on that particular port), or 'nofifo'
3244 (to avoid using a FIFO even if it is detected).
3245 Parallel ports are assigned in the order they
3246 are specified on the command line, starting
3249 parport_init_mode= [HW,PPT]
3250 Configure VIA parallel port to operate in
3251 a specific mode. This is necessary on Pegasos
3252 computer where firmware has no options for setting
3253 up parallel port mode and sets it to spp.
3254 Currently this function knows 686a and 8231 chips.
3255 Format: [spp|ps2|epp|ecp|ecpepp]
3258 Halt all CPUs after the first oops has been printed for
3259 the specified number of seconds. This is to be used if
3260 your oopses keep scrolling off the screen.
3265 See header of drivers/block/paride/pcd.c.
3266 See also Documentation/blockdev/paride.txt.
3268 pci=option[,option...] [PCI] various PCI subsystem options.
3270 Some options herein operate on a specific device
3271 or a set of devices (<pci_dev>). These are
3272 specified in one of the following formats:
3274 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3275 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3277 Note: the first format specifies a PCI
3278 bus/device/function address which may change
3279 if new hardware is inserted, if motherboard
3280 firmware changes, or due to changes caused
3281 by other kernel parameters. If the
3282 domain is left unspecified, it is
3283 taken to be zero. Optionally, a path
3284 to a device through multiple device/function
3285 addresses can be specified after the base
3286 address (this is more robust against
3287 renumbering issues). The second format
3288 selects devices using IDs from the
3289 configuration space which may match multiple
3290 devices in the system.
3292 earlydump dump PCI config space before the kernel
3294 off [X86] don't probe for the PCI bus
3295 bios [X86-32] force use of PCI BIOS, don't access
3296 the hardware directly. Use this if your machine
3297 has a non-standard PCI host bridge.
3298 nobios [X86-32] disallow use of PCI BIOS, only direct
3299 hardware access methods are allowed. Use this
3300 if you experience crashes upon bootup and you
3301 suspect they are caused by the BIOS.
3302 conf1 [X86] Force use of PCI Configuration Access
3303 Mechanism 1 (config address in IO port 0xCF8,
3304 data in IO port 0xCFC, both 32-bit).
3305 conf2 [X86] Force use of PCI Configuration Access
3306 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3307 the function, IO port 0xCFA, also 8-bit, sets
3308 bus number. The config space is then accessed
3309 through ports 0xC000-0xCFFF).
3310 See http://wiki.osdev.org/PCI for more info
3311 on the configuration access mechanisms.
3312 noaer [PCIE] If the PCIEAER kernel config parameter is
3313 enabled, this kernel boot option can be used to
3314 disable the use of PCIE advanced error reporting.
3315 nodomains [PCI] Disable support for multiple PCI
3316 root domains (aka PCI segments, in ACPI-speak).
3317 nommconf [X86] Disable use of MMCONFIG for PCI
3319 check_enable_amd_mmconf [X86] check for and enable
3320 properly configured MMIO access to PCI
3321 config space on AMD family 10h CPU
3322 nomsi [MSI] If the PCI_MSI kernel config parameter is
3323 enabled, this kernel boot option can be used to
3324 disable the use of MSI interrupts system-wide.
3325 noioapicquirk [APIC] Disable all boot interrupt quirks.
3326 Safety option to keep boot IRQs enabled. This
3327 should never be necessary.
3328 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3329 primary IO-APIC for bridges that cannot disable
3330 boot IRQs. This fixes a source of spurious IRQs
3331 when the system masks IRQs.
3332 noioapicreroute [APIC] Disable workaround that uses the
3333 boot IRQ equivalent of an IRQ that connects to
3334 a chipset where boot IRQs cannot be disabled.
3335 The opposite of ioapicreroute.
3336 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3337 routing table. These calls are known to be buggy
3338 on several machines and they hang the machine
3339 when used, but on other computers it's the only
3340 way to get the interrupt routing table. Try
3341 this option if the kernel is unable to allocate
3342 IRQs or discover secondary PCI buses on your
3344 rom [X86] Assign address space to expansion ROMs.
3345 Use with caution as certain devices share
3346 address decoders between ROMs and other
3348 norom [X86] Do not assign address space to
3349 expansion ROMs that do not already have
3350 BIOS assigned address ranges.
3351 nobar [X86] Do not assign address space to the
3352 BARs that weren't assigned by the BIOS.
3353 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3354 assigned automatically to PCI devices. You can
3355 make the kernel exclude IRQs of your ISA cards
3357 pirqaddr=0xAAAAA [X86] Specify the physical address
3358 of the PIRQ table (normally generated
3359 by the BIOS) if it is outside the
3360 F0000h-100000h range.
3361 lastbus=N [X86] Scan all buses thru bus #N. Can be
3362 useful if the kernel is unable to find your
3363 secondary buses and you want to tell it
3364 explicitly which ones they are.
3365 assign-busses [X86] Always assign all PCI bus
3366 numbers ourselves, overriding
3367 whatever the firmware may have done.
3368 usepirqmask [X86] Honor the possible IRQ mask stored
3369 in the BIOS $PIR table. This is needed on
3370 some systems with broken BIOSes, notably
3371 some HP Pavilion N5400 and Omnibook XE3
3372 notebooks. This will have no effect if ACPI
3373 IRQ routing is enabled.
3374 noacpi [X86] Do not use ACPI for IRQ routing
3375 or for PCI scanning.
3376 use_crs [X86] Use PCI host bridge window information
3377 from ACPI. On BIOSes from 2008 or later, this
3378 is enabled by default. If you need to use this,
3379 please report a bug.
3380 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3381 If you need to use this, please report a bug.
3382 routeirq Do IRQ routing for all PCI devices.
3383 This is normally done in pci_enable_device(),
3384 so this option is a temporary workaround
3385 for broken drivers that don't call it.
3386 skip_isa_align [X86] do not align io start addr, so can
3387 handle more pci cards
3388 noearly [X86] Don't do any early type 1 scanning.
3389 This might help on some broken boards which
3390 machine check when some devices' config space
3391 is read. But various workarounds are disabled
3392 and some IOMMU drivers will not work.
3393 bfsort Sort PCI devices into breadth-first order.
3394 This sorting is done to get a device
3395 order compatible with older (<= 2.4) kernels.
3396 nobfsort Don't sort PCI devices into breadth-first order.
3397 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3398 tuning and use the BIOS-configured MPS defaults.
3399 pcie_bus_safe Set every device's MPS to the largest value
3400 supported by all devices below the root complex.
3401 pcie_bus_perf Set device MPS to the largest allowable MPS
3402 based on its parent bus. Also set MRRS (Max
3403 Read Request Size) to the largest supported
3404 value (no larger than the MPS that the device
3405 or bus can support) for best performance.
3406 pcie_bus_peer2peer Set every device's MPS to 128B, which
3407 every device is guaranteed to support. This
3408 configuration allows peer-to-peer DMA between
3409 any pair of devices, possibly at the cost of
3410 reduced performance. This also guarantees
3411 that hot-added devices will work.
3412 cbiosize=nn[KMG] The fixed amount of bus space which is
3413 reserved for the CardBus bridge's IO window.
3414 The default value is 256 bytes.
3415 cbmemsize=nn[KMG] The fixed amount of bus space which is
3416 reserved for the CardBus bridge's memory
3417 window. The default value is 64 megabytes.
3420 [<order of align>@]<pci_dev>[; ...]
3421 Specifies alignment and device to reassign
3422 aligned memory resources. How to
3423 specify the device is described above.
3424 If <order of align> is not specified,
3425 PAGE_SIZE is used as alignment.
3426 PCI-PCI bridge can be specified, if resource
3427 windows need to be expanded.
3428 To specify the alignment for several
3429 instances of a device, the PCI vendor,
3430 device, subvendor, and subdevice may be
3431 specified, e.g., 4096@pci:8086:9c22:103c:198f
3432 ecrc= Enable/disable PCIe ECRC (transaction layer
3433 end-to-end CRC checking).
3434 bios: Use BIOS/firmware settings. This is the
3438 hpiosize=nn[KMG] The fixed amount of bus space which is
3439 reserved for hotplug bridge's IO window.
3440 Default size is 256 bytes.
3441 hpmemsize=nn[KMG] The fixed amount of bus space which is
3442 reserved for hotplug bridge's memory window.
3443 Default size is 2 megabytes.
3444 hpbussize=nn The minimum amount of additional bus numbers
3445 reserved for buses below a hotplug bridge.
3447 realloc= Enable/disable reallocating PCI bridge resources
3448 if allocations done by BIOS are too small to
3449 accommodate resources required by all child
3451 off: Turn realloc off
3453 realloc same as realloc=on
3454 noari do not use PCIe ARI.
3455 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
3456 do not use PCIe ATS (and IOMMU device IOTLB).
3457 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3458 only look for one device below a PCIe downstream
3460 big_root_window Try to add a big 64bit memory window to the PCIe
3461 root complex on AMD CPUs. Some GFX hardware
3462 can resize a BAR to allow access to all VRAM.
3463 Adding the window is slightly risky (it may
3464 conflict with unreported devices), so this
3466 disable_acs_redir=<pci_dev>[; ...]
3467 Specify one or more PCI devices (in the format
3468 specified above) separated by semicolons.
3469 Each device specified will have the PCI ACS
3470 redirect capabilities forced off which will
3471 allow P2P traffic between devices through
3472 bridges without forcing it upstream. Note:
3473 this removes isolation between devices and
3474 may put more devices in an IOMMU group.
3475 force_floating [S390] Force usage of floating interrupts.
3476 nomio [S390] Do not use MIO instructions.
3478 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3481 force Enable ASPM even on devices that claim not to support it.
3482 WARNING: Forcing ASPM on may cause system lockups.
3484 pcie_ports= [PCIE] PCIe port services handling:
3485 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
3486 even if the platform doesn't give the OS permission to
3487 use them. This may cause conflicts if the platform
3488 also tries to use these services.
3489 compat Disable native PCIe services (PME, AER, DPC, PCIe
3492 pcie_port_pm= [PCIE] PCIe port power management handling:
3493 off Disable power management of all PCIe ports
3494 force Forcibly enable power management of all PCIe ports
3496 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3497 nomsi Do not use MSI for native PCIe PME signaling (this makes
3498 all PCIe root ports use INTx for all services).
3500 pcmv= [HW,PCMCIA] BadgePAD 4
3504 Keep all power-domains already enabled by bootloader on,
3505 even if no driver has claimed them. This is useful
3506 for debug and development, but should not be
3507 needed on a platform with proper driver support.
3510 See Documentation/blockdev/paride.txt.
3512 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3515 See arch/parisc/kernel/pdc_chassis.c
3517 percpu_alloc= Select which percpu first chunk allocator to use.
3518 Currently supported values are "embed" and "page".
3519 Archs may support subset or none of the selections.
3520 See comments in mm/percpu.c for details on each
3521 allocator. This parameter is primarily for debugging
3522 and performance comparison.
3525 See Documentation/blockdev/paride.txt.
3528 See Documentation/blockdev/paride.txt.
3530 pirq= [SMP,APIC] Manual mp-table setup
3531 See Documentation/x86/i386/IO-APIC.txt.
3533 plip= [PPT,NET] Parallel port network link
3534 Format: { parport<nr> | timid | 0 }
3535 See also Documentation/admin-guide/parport.rst.
3537 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3538 Override pmtimer IOPort with a hex value.
3542 Enable PNP debug messages (depends on the
3543 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3544 via /sys/module/pnp/parameters/debug. We always show
3545 current resource usage; turning this on also shows
3546 possible settings and some assignment information.
3552 { on | off | curr | res | no-curr | no-res }
3555 [ISAPNP] Exclude IRQs for the autoconfiguration
3558 [ISAPNP] Exclude DMAs for the autoconfiguration
3560 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3561 Ranges are in pairs (I/O port base and size).
3564 [ISAPNP] Exclude memory regions for the
3566 Ranges are in pairs (memory base and size).
3568 ports= [IP_VS_FTP] IPVS ftp helper module
3570 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3572 Format: <port>,<port>....
3574 powersave=off [PPC] This option disables power saving features.
3575 It specifically disables cpuidle and sets the
3576 platform machine description specific power_save
3577 function to NULL. On Idle the CPU just reduces
3580 ppc_strict_facility_enable
3581 [PPC] This option catches any kernel floating point,
3582 Altivec, VSX and SPE outside of regions specifically
3583 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3584 There is some performance impact when enabling this.
3588 Disable Hardware Transactional Memory
3590 print-fatal-signals=
3591 [KNL] debug: print fatal signals
3593 If enabled, warn about various signal handling
3594 related application anomalies: too many signals,
3595 too many POSIX.1 timers, fatal signals causing a
3598 If you hit the warning due to signal overflow,
3599 you might want to try "ulimit -i unlimited".
3603 printk.always_kmsg_dump=
3604 Trigger kmsg_dump for cases other than kernel oops or
3606 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3609 printk.devkmsg={on,off,ratelimit}
3610 Control writing to /dev/kmsg.
3611 on - unlimited logging to /dev/kmsg from userspace
3612 off - logging to /dev/kmsg disabled
3613 ratelimit - ratelimit the logging
3616 printk.time= Show timing data prefixed to each printk message line
3617 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3619 processor.max_cstate= [HW,ACPI]
3620 Limit processor to maximum C-state
3621 max_cstate=9 overrides any DMI blacklist limit.
3623 processor.nocst [HW,ACPI]
3624 Ignore the _CST method to determine C-states,
3625 instead using the legacy FADT method
3627 profile= [KNL] Enable kernel profiling via /proc/profile
3628 Format: [<profiletype>,]<number>
3629 Param: <profiletype>: "schedule", "sleep", or "kvm"
3630 [defaults to kernel profiling]
3631 Param: "schedule" - profile schedule points.
3632 Param: "sleep" - profile D-state sleeping (millisecs).
3633 Requires CONFIG_SCHEDSTATS
3634 Param: "kvm" - profile VM exits.
3635 Param: <number> - step/bucket size as a power of 2 for
3636 statistical time based profiling.
3638 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3640 See Documentation/blockdev/ramdisk.txt.
3642 psi= [KNL] Enable or disable pressure stall information
3646 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3647 probe for; one of (bare|imps|exps|lifebook|any).
3648 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3650 psmouse.resetafter= [HW,MOUSE]
3651 Try to reset the device after so many bad packets
3654 [HW,MOUSE] Set desired mouse resolution, in dpi.
3655 psmouse.smartscroll=
3656 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3657 0 = disabled, 1 = enabled (default).
3659 pstore.backend= Specify the name of the pstore backend to use
3662 See Documentation/blockdev/paride.txt.
3664 pti= [X86_64] Control Page Table Isolation of user and
3665 kernel address spaces. Disabling this feature
3666 removes hardening, but improves performance of
3667 system calls and interrupts.
3669 on - unconditionally enable
3670 off - unconditionally disable
3671 auto - kernel detects whether your CPU model is
3672 vulnerable to issues that PTI mitigates
3674 Not specifying this option is equivalent to pti=auto.
3677 Equivalent to pti=off
3680 [KNL] Number of legacy pty's. Overwrites compiled-in
3683 quiet [KNL] Disable most log messages
3688 See Documentation/admin-guide/md.rst.
3690 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3691 See Documentation/blockdev/ramdisk.txt.
3693 random.trust_cpu={on,off}
3694 [KNL] Enable or disable trusting the use of the
3695 CPU's random number generator (if available) to
3696 fully seed the kernel's CRNG. Default is controlled
3697 by CONFIG_RANDOM_TRUST_CPU.
3699 ras=option[,option,...] [KNL] RAS-specific options
3702 Disable the Correctable Errors Collector,
3703 see CONFIG_RAS_CEC help text.
3706 The argument is a cpu list, as described above,
3707 except that the string "all" can be used to
3708 specify every CPU on the system.
3710 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3711 the specified list of CPUs to be no-callback CPUs.
3712 Invocation of these CPUs' RCU callbacks will be
3713 offloaded to "rcuox/N" kthreads created for that
3714 purpose, where "x" is "p" for RCU-preempt, and
3715 "s" for RCU-sched, and "N" is the CPU number.
3716 This reduces OS jitter on the offloaded CPUs,
3717 which can be useful for HPC and real-time
3718 workloads. It can also improve energy efficiency
3719 for asymmetric multiprocessors.
3722 Rather than requiring that offloaded CPUs
3723 (specified by rcu_nocbs= above) explicitly
3724 awaken the corresponding "rcuoN" kthreads,
3725 make these kthreads poll for callbacks.
3726 This improves the real-time response for the
3727 offloaded CPUs by relieving them of the need to
3728 wake up the corresponding kthread, but degrades
3729 energy efficiency by requiring that the kthreads
3730 periodically wake up to do the polling.
3732 rcutree.blimit= [KNL]
3733 Set maximum number of finished RCU callbacks to
3734 process in one batch.
3736 rcutree.dump_tree= [KNL]
3737 Dump the structure of the rcu_node combining tree
3738 out at early boot. This is used for diagnostic
3739 purposes, to verify correct tree setup.
3741 rcutree.gp_cleanup_delay= [KNL]
3742 Set the number of jiffies to delay each step of
3743 RCU grace-period cleanup.
3745 rcutree.gp_init_delay= [KNL]
3746 Set the number of jiffies to delay each step of
3747 RCU grace-period initialization.
3749 rcutree.gp_preinit_delay= [KNL]
3750 Set the number of jiffies to delay each step of
3751 RCU grace-period pre-initialization, that is,
3752 the propagation of recent CPU-hotplug changes up
3753 the rcu_node combining tree.
3755 rcutree.rcu_fanout_exact= [KNL]
3756 Disable autobalancing of the rcu_node combining
3757 tree. This is used by rcutorture, and might
3758 possibly be useful for architectures having high
3759 cache-to-cache transfer latencies.
3761 rcutree.rcu_fanout_leaf= [KNL]
3762 Change the number of CPUs assigned to each
3763 leaf rcu_node structure. Useful for very
3764 large systems, which will choose the value 64,
3765 and for NUMA systems with large remote-access
3766 latencies, which will choose a value aligned
3767 with the appropriate hardware boundaries.
3769 rcutree.jiffies_till_first_fqs= [KNL]
3770 Set delay from grace-period initialization to
3771 first attempt to force quiescent states.
3772 Units are jiffies, minimum value is zero,
3773 and maximum value is HZ.
3775 rcutree.jiffies_till_next_fqs= [KNL]
3776 Set delay between subsequent attempts to force
3777 quiescent states. Units are jiffies, minimum
3778 value is one, and maximum value is HZ.
3780 rcutree.jiffies_till_sched_qs= [KNL]
3781 Set required age in jiffies for a
3782 given grace period before RCU starts
3783 soliciting quiescent-state help from
3784 rcu_note_context_switch() and cond_resched().
3785 If not specified, the kernel will calculate
3786 a value based on the most recent settings
3787 of rcutree.jiffies_till_first_fqs
3788 and rcutree.jiffies_till_next_fqs.
3789 This calculated value may be viewed in
3790 rcutree.jiffies_to_sched_qs. Any attempt to set
3791 rcutree.jiffies_to_sched_qs will be cheerfully
3794 rcutree.kthread_prio= [KNL,BOOT]
3795 Set the SCHED_FIFO priority of the RCU per-CPU
3796 kthreads (rcuc/N). This value is also used for
3797 the priority of the RCU boost threads (rcub/N)
3798 and for the RCU grace-period kthreads (rcu_bh,
3799 rcu_preempt, and rcu_sched). If RCU_BOOST is
3800 set, valid values are 1-99 and the default is 1
3801 (the least-favored priority). Otherwise, when
3802 RCU_BOOST is not set, valid values are 0-99 and
3803 the default is zero (non-realtime operation).
3805 rcutree.rcu_nocb_leader_stride= [KNL]
3806 Set the number of NOCB kthread groups, which
3807 defaults to the square root of the number of
3808 CPUs. Larger numbers reduces the wakeup overhead
3809 on the per-CPU grace-period kthreads, but increases
3810 that same overhead on each group's leader.
3812 rcutree.qhimark= [KNL]
3813 Set threshold of queued RCU callbacks beyond which
3814 batch limiting is disabled.
3816 rcutree.qlowmark= [KNL]
3817 Set threshold of queued RCU callbacks below which
3818 batch limiting is re-enabled.
3820 rcutree.rcu_idle_gp_delay= [KNL]
3821 Set wakeup interval for idle CPUs that have
3822 RCU callbacks (RCU_FAST_NO_HZ=y).
3824 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3825 Set wakeup interval for idle CPUs that have
3826 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3827 Lazy RCU callbacks are those which RCU can
3828 prove do nothing more than free memory.
3830 rcutree.rcu_kick_kthreads= [KNL]
3831 Cause the grace-period kthread to get an extra
3832 wake_up() if it sleeps three times longer than
3833 it should at force-quiescent-state time.
3834 This wake_up() will be accompanied by a
3835 WARN_ONCE() splat and an ftrace_dump().
3837 rcutree.sysrq_rcu= [KNL]
3838 Commandeer a sysrq key to dump out Tree RCU's
3839 rcu_node tree with an eye towards determining
3840 why a new grace period has not yet started.
3842 rcuperf.gp_async= [KNL]
3843 Measure performance of asynchronous
3844 grace-period primitives such as call_rcu().
3846 rcuperf.gp_async_max= [KNL]
3847 Specify the maximum number of outstanding
3848 callbacks per writer thread. When a writer
3849 thread exceeds this limit, it invokes the
3850 corresponding flavor of rcu_barrier() to allow
3851 previously posted callbacks to drain.
3853 rcuperf.gp_exp= [KNL]
3854 Measure performance of expedited synchronous
3855 grace-period primitives.
3857 rcuperf.holdoff= [KNL]
3858 Set test-start holdoff period. The purpose of
3859 this parameter is to delay the start of the
3860 test until boot completes in order to avoid
3863 rcuperf.nreaders= [KNL]
3864 Set number of RCU readers. The value -1 selects
3865 N, where N is the number of CPUs. A value
3866 "n" less than -1 selects N-n+1, where N is again
3867 the number of CPUs. For example, -2 selects N
3868 (the number of CPUs), -3 selects N+1, and so on.
3869 A value of "n" less than or equal to -N selects
3872 rcuperf.nwriters= [KNL]
3873 Set number of RCU writers. The values operate
3874 the same as for rcuperf.nreaders.
3875 N, where N is the number of CPUs
3877 rcuperf.perf_type= [KNL]
3878 Specify the RCU implementation to test.
3880 rcuperf.shutdown= [KNL]
3881 Shut the system down after performance tests
3882 complete. This is useful for hands-off automated
3885 rcuperf.verbose= [KNL]
3886 Enable additional printk() statements.
3888 rcuperf.writer_holdoff= [KNL]
3889 Write-side holdoff between grace periods,
3890 in microseconds. The default of zero says
3893 rcutorture.fqs_duration= [KNL]
3894 Set duration of force_quiescent_state bursts
3897 rcutorture.fqs_holdoff= [KNL]
3898 Set holdoff time within force_quiescent_state bursts
3901 rcutorture.fqs_stutter= [KNL]
3902 Set wait time between force_quiescent_state bursts
3905 rcutorture.fwd_progress= [KNL]
3906 Enable RCU grace-period forward-progress testing
3907 for the types of RCU supporting this notion.
3909 rcutorture.fwd_progress_div= [KNL]
3910 Specify the fraction of a CPU-stall-warning
3911 period to do tight-loop forward-progress testing.
3913 rcutorture.fwd_progress_holdoff= [KNL]
3914 Number of seconds to wait between successive
3915 forward-progress tests.
3917 rcutorture.fwd_progress_need_resched= [KNL]
3918 Enclose cond_resched() calls within checks for
3919 need_resched() during tight-loop forward-progress
3922 rcutorture.gp_cond= [KNL]
3923 Use conditional/asynchronous update-side
3924 primitives, if available.
3926 rcutorture.gp_exp= [KNL]
3927 Use expedited update-side primitives, if available.
3929 rcutorture.gp_normal= [KNL]
3930 Use normal (non-expedited) asynchronous
3931 update-side primitives, if available.
3933 rcutorture.gp_sync= [KNL]
3934 Use normal (non-expedited) synchronous
3935 update-side primitives, if available. If all
3936 of rcutorture.gp_cond=, rcutorture.gp_exp=,
3937 rcutorture.gp_normal=, and rcutorture.gp_sync=
3938 are zero, rcutorture acts as if is interpreted
3939 they are all non-zero.
3941 rcutorture.n_barrier_cbs= [KNL]
3942 Set callbacks/threads for rcu_barrier() testing.
3944 rcutorture.nfakewriters= [KNL]
3945 Set number of concurrent RCU writers. These just
3946 stress RCU, they don't participate in the actual
3947 test, hence the "fake".
3949 rcutorture.nreaders= [KNL]
3950 Set number of RCU readers. The value -1 selects
3951 N-1, where N is the number of CPUs. A value
3952 "n" less than -1 selects N-n-2, where N is again
3953 the number of CPUs. For example, -2 selects N
3954 (the number of CPUs), -3 selects N+1, and so on.
3956 rcutorture.object_debug= [KNL]
3957 Enable debug-object double-call_rcu() testing.
3959 rcutorture.onoff_holdoff= [KNL]
3960 Set time (s) after boot for CPU-hotplug testing.
3962 rcutorture.onoff_interval= [KNL]
3963 Set time (jiffies) between CPU-hotplug operations,
3964 or zero to disable CPU-hotplug testing.
3966 rcutorture.shuffle_interval= [KNL]
3967 Set task-shuffle interval (s). Shuffling tasks
3968 allows some CPUs to go into dyntick-idle mode
3969 during the rcutorture test.
3971 rcutorture.shutdown_secs= [KNL]
3972 Set time (s) after boot system shutdown. This
3973 is useful for hands-off automated testing.
3975 rcutorture.stall_cpu= [KNL]
3976 Duration of CPU stall (s) to test RCU CPU stall
3977 warnings, zero to disable.
3979 rcutorture.stall_cpu_holdoff= [KNL]
3980 Time to wait (s) after boot before inducing stall.
3982 rcutorture.stall_cpu_irqsoff= [KNL]
3983 Disable interrupts while stalling if set.
3985 rcutorture.stat_interval= [KNL]
3986 Time (s) between statistics printk()s.
3988 rcutorture.stutter= [KNL]
3989 Time (s) to stutter testing, for example, specifying
3990 five seconds causes the test to run for five seconds,
3991 wait for five seconds, and so on. This tests RCU's
3992 ability to transition abruptly to and from idle.
3994 rcutorture.test_boost= [KNL]
3995 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
3996 "Maybe" means test if the RCU implementation
3997 under test support RCU priority boosting.
3999 rcutorture.test_boost_duration= [KNL]
4000 Duration (s) of each individual boost test.
4002 rcutorture.test_boost_interval= [KNL]
4003 Interval (s) between each boost test.
4005 rcutorture.test_no_idle_hz= [KNL]
4006 Test RCU's dyntick-idle handling. See also the
4007 rcutorture.shuffle_interval parameter.
4009 rcutorture.torture_type= [KNL]
4010 Specify the RCU implementation to test.
4012 rcutorture.verbose= [KNL]
4013 Enable additional printk() statements.
4015 rcupdate.rcu_cpu_stall_suppress= [KNL]
4016 Suppress RCU CPU stall warning messages.
4018 rcupdate.rcu_cpu_stall_timeout= [KNL]
4019 Set timeout for RCU CPU stall warning messages.
4021 rcupdate.rcu_expedited= [KNL]
4022 Use expedited grace-period primitives, for
4023 example, synchronize_rcu_expedited() instead
4024 of synchronize_rcu(). This reduces latency,
4025 but can increase CPU utilization, degrade
4026 real-time latency, and degrade energy efficiency.
4027 No effect on CONFIG_TINY_RCU kernels.
4029 rcupdate.rcu_normal= [KNL]
4030 Use only normal grace-period primitives,
4031 for example, synchronize_rcu() instead of
4032 synchronize_rcu_expedited(). This improves
4033 real-time latency, CPU utilization, and
4034 energy efficiency, but can expose users to
4035 increased grace-period latency. This parameter
4036 overrides rcupdate.rcu_expedited. No effect on
4037 CONFIG_TINY_RCU kernels.
4039 rcupdate.rcu_normal_after_boot= [KNL]
4040 Once boot has completed (that is, after
4041 rcu_end_inkernel_boot() has been invoked), use
4042 only normal grace-period primitives. No effect
4043 on CONFIG_TINY_RCU kernels.
4045 rcupdate.rcu_task_stall_timeout= [KNL]
4046 Set timeout in jiffies for RCU task stall warning
4047 messages. Disable with a value less than or equal
4050 rcupdate.rcu_self_test= [KNL]
4051 Run the RCU early boot self tests
4055 Run specified binary instead of /init from the ramdisk,
4056 used for early userspace startup. See initrd.
4059 Turn on/off individual RDT features. List is:
4060 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
4062 E.g. to turn on cmt and turn off mba use:
4066 Format (x86 or x86_64):
4067 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
4069 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
4071 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
4072 (prefix with 'panic_' to set mode for panic
4074 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
4075 reboot_force is either force or not specified,
4076 reboot_cpu is s[mp]#### with #### being the processor
4077 to be used for rebooting.
4080 [KNL, SMP] Set scheduler's default relax_domain_level.
4081 See Documentation/cgroup-v1/cpusets.txt.
4083 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
4084 Format: <base1>,<size1>[,<base2>,<size2>,...]
4085 Reserve I/O ports or memory so the kernel won't use
4086 them. If <base> is less than 0x10000, the region
4087 is assumed to be I/O ports; otherwise it is memory.
4089 reservetop= [X86-32]
4091 Reserves a hole at the top of the kernel virtual
4096 Set the amount of memory to reserve for BIOS at
4097 the bottom of the address space.
4099 reset_devices [KNL] Force drivers to reset the underlying device
4100 during initialization.
4103 Specify the partition device for software suspend
4105 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
4107 resume_offset= [SWSUSP]
4108 Specify the offset from the beginning of the partition
4109 given by "resume=" at which the swap header is located,
4110 in <PAGE_SIZE> units (needed only for swap files).
4111 See Documentation/power/swsusp-and-swap-files.txt
4113 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
4114 read the resume files
4116 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
4117 Useful for devices that are detected asynchronously
4118 (e.g. USB and MMC devices).
4120 hibernate= [HIBERNATION]
4121 noresume Don't check if there's a hibernation image
4122 present during boot.
4123 nocompress Don't compress/decompress hibernation images.
4124 no Disable hibernation and resume.
4125 protect_image Turn on image protection during restoration
4126 (that will set all pages holding image data
4127 during restoration read-only).
4129 retain_initrd [RAM] Keep initrd memory after extraction
4131 rfkill.default_state=
4132 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
4133 etc. communication is blocked by default.
4136 rfkill.master_switch_mode=
4137 0 The "airplane mode" button does nothing.
4138 1 The "airplane mode" button toggles between everything
4139 blocked and the previous configuration.
4140 2 The "airplane mode" button toggles between everything
4141 blocked and everything unblocked.
4143 rhash_entries= [KNL,NET]
4144 Set number of hash buckets for route cache
4147 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4150 ro [KNL] Mount root device read-only on boot
4153 on Mark read-only kernel memory as read-only (default).
4154 off Leave read-only kernel memory writable for debugging.
4157 Enable the uart passthrough on the designated usb port
4158 on Rockchip SoCs. When active, the signals of the
4159 debug-uart get routed to the D+ and D- pins of the usb
4160 port and the regular usb controller gets disabled.
4162 root= [KNL] Root filesystem
4163 See name_to_dev_t comment in init/do_mounts.c.
4165 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4166 mount the root filesystem
4168 rootflags= [KNL] Set root filesystem mount option string
4170 rootfstype= [KNL] Set root filesystem type
4172 rootwait [KNL] Wait (indefinitely) for root device to show up.
4173 Useful for devices that are detected asynchronously
4174 (e.g. USB and MMC devices).
4176 rproc_mem=nn[KMG][@address]
4177 [KNL,ARM,CMA] Remoteproc physical memory block.
4178 Memory area to be used by remote processor image,
4181 rw [KNL] Mount root device read-write on boot
4183 S [KNL] Run init in single mode
4185 s390_iommu= [HW,S390]
4186 Set s390 IOTLB flushing mode
4188 With strict flushing every unmap operation will result in
4189 an IOTLB flush. Default is lazy flushing before reuse,
4193 See drivers/net/irda/sa1100_ir.c.
4195 sbni= [NET] Granch SBNI12 leased line adapter
4197 sched_debug [KNL] Enables verbose scheduler debug messages.
4199 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4200 Allowed values are enable and disable. This feature
4201 incurs a small amount of overhead in the scheduler
4202 but is useful for debugging and performance tuning.
4204 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
4205 xtime_lock contention on larger systems, and/or RCU lock
4206 contention on all systems with CONFIG_MAXSMP set.
4207 Format: { "0" | "1" }
4208 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
4210 Note: increases power consumption, thus should only be
4211 enabled if running jitter sensitive (HPC/RT) workloads.
4213 security= [SECURITY] Choose a legacy "major" security module to
4214 enable at boot. This has been deprecated by the
4217 selinux= [SELINUX] Disable or enable SELinux at boot time.
4218 Format: { "0" | "1" }
4219 See security/selinux/Kconfig help text.
4222 Default value is set via kernel config option.
4223 If enabled at boot time, /selinux/disable can be used
4224 later to disable prior to initial policy load.
4226 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
4227 Format: { "0" | "1" }
4228 See security/apparmor/Kconfig help text
4231 Default value is set via kernel config option.
4233 serialnumber [BUGS=X86-32]
4236 Maximal number of shapers.
4244 Disable merging of slabs with similar size. May be
4245 necessary if there is some reason to distinguish
4246 allocs to different slabs, especially in hardened
4247 environments where the risk of heap overflows and
4248 layout control by attackers can usually be
4249 frustrated by disabling merging. This will reduce
4250 most of the exposure of a heap attack to a single
4251 cache (risks via metadata attacks are mostly
4252 unchanged). Debug options disable merging on their
4254 For more information see Documentation/vm/slub.rst.
4256 slab_max_order= [MM, SLAB]
4257 Determines the maximum allowed order for slabs.
4258 A high setting may cause OOMs due to memory
4259 fragmentation. Defaults to 1 for systems with
4260 more than 32MB of RAM, 0 otherwise.
4262 slub_debug[=options[,slabs]] [MM, SLUB]
4263 Enabling slub_debug allows one to determine the
4264 culprit if slab objects become corrupted. Enabling
4265 slub_debug can create guard zones around objects and
4266 may poison objects when not in use. Also tracks the
4267 last alloc / free. For more information see
4268 Documentation/vm/slub.rst.
4270 slub_memcg_sysfs= [MM, SLUB]
4271 Determines whether to enable sysfs directories for
4272 memory cgroup sub-caches. 1 to enable, 0 to disable.
4273 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
4274 Enabling this can lead to a very high number of debug
4275 directories and files being created under
4278 slub_max_order= [MM, SLUB]
4279 Determines the maximum allowed order for slabs.
4280 A high setting may cause OOMs due to memory
4281 fragmentation. For more information see
4282 Documentation/vm/slub.rst.
4284 slub_min_objects= [MM, SLUB]
4285 The minimum number of objects per slab. SLUB will
4286 increase the slab order up to slub_max_order to
4287 generate a sufficiently large slab able to contain
4288 the number of objects indicated. The higher the number
4289 of objects the smaller the overhead of tracking slabs
4290 and the less frequently locks need to be acquired.
4291 For more information see Documentation/vm/slub.rst.
4293 slub_min_order= [MM, SLUB]
4294 Determines the minimum page order for slabs. Must be
4295 lower than slub_max_order.
4296 For more information see Documentation/vm/slub.rst.
4298 slub_nomerge [MM, SLUB]
4299 Same with slab_nomerge. This is supported for legacy.
4300 See slab_nomerge for more information.
4303 Format: <io1>[,<io2>[,...,<io8>]]
4305 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
4306 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4307 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4308 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4309 smsc-ircc2.ircc_irq= [HW] IRQ line
4310 smsc-ircc2.ircc_dma= [HW] DMA channel
4311 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4312 0: Toshiba Satellite 1800 (GP data pin select)
4313 1: Fast pin select (default)
4316 smt [KNL,S390] Set the maximum number of threads (logical
4317 CPUs) to use per physical CPU on systems capable of
4318 symmetric multithreading (SMT). Will be capped to the
4319 actual hardware limit.
4321 Default: -1 (no limit)
4324 [KNL] Should the soft-lockup detector generate panics.
4327 A nonzero value instructs the soft-lockup detector
4328 to panic the machine when a soft-lockup occurs. This
4329 is also controlled by CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC
4330 which is the respective build-time switch to that
4333 softlockup_all_cpu_backtrace=
4334 [KNL] Should the soft-lockup detector generate
4335 backtraces on all cpus.
4338 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4339 See Documentation/laptops/sonypi.txt
4341 spectre_v2= [X86] Control mitigation of Spectre variant 2
4342 (indirect branch speculation) vulnerability.
4343 The default operation protects the kernel from
4346 on - unconditionally enable, implies
4348 off - unconditionally disable, implies
4350 auto - kernel detects whether your CPU model is
4353 Selecting 'on' will, and 'auto' may, choose a
4354 mitigation method at run time according to the
4355 CPU, the available microcode, the setting of the
4356 CONFIG_RETPOLINE configuration option, and the
4357 compiler with which the kernel was built.
4359 Selecting 'on' will also enable the mitigation
4360 against user space to user space task attacks.
4362 Selecting 'off' will disable both the kernel and
4363 the user space protections.
4365 Specific mitigations can also be selected manually:
4367 retpoline - replace indirect branches
4368 retpoline,generic - google's original retpoline
4369 retpoline,amd - AMD-specific minimal thunk
4371 Not specifying this option is equivalent to
4375 [X86] Control mitigation of Spectre variant 2
4376 (indirect branch speculation) vulnerability between
4379 on - Unconditionally enable mitigations. Is
4380 enforced by spectre_v2=on
4382 off - Unconditionally disable mitigations. Is
4383 enforced by spectre_v2=off
4385 prctl - Indirect branch speculation is enabled,
4386 but mitigation can be enabled via prctl
4387 per thread. The mitigation control state
4388 is inherited on fork.
4391 - Like "prctl" above, but only STIBP is
4392 controlled per thread. IBPB is issued
4393 always when switching between different user
4397 - Same as "prctl" above, but all seccomp
4398 threads will enable the mitigation unless
4399 they explicitly opt out.
4402 - Like "seccomp" above, but only STIBP is
4403 controlled per thread. IBPB is issued
4404 always when switching between different
4405 user space processes.
4407 auto - Kernel selects the mitigation depending on
4408 the available CPU features and vulnerability.
4411 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4413 Not specifying this option is equivalent to
4414 spectre_v2_user=auto.
4416 spec_store_bypass_disable=
4417 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4418 (Speculative Store Bypass vulnerability)
4420 Certain CPUs are vulnerable to an exploit against a
4421 a common industry wide performance optimization known
4422 as "Speculative Store Bypass" in which recent stores
4423 to the same memory location may not be observed by
4424 later loads during speculative execution. The idea
4425 is that such stores are unlikely and that they can
4426 be detected prior to instruction retirement at the
4427 end of a particular speculation execution window.
4429 In vulnerable processors, the speculatively forwarded
4430 store can be used in a cache side channel attack, for
4431 example to read memory to which the attacker does not
4432 directly have access (e.g. inside sandboxed code).
4434 This parameter controls whether the Speculative Store
4435 Bypass optimization is used.
4437 On x86 the options are:
4439 on - Unconditionally disable Speculative Store Bypass
4440 off - Unconditionally enable Speculative Store Bypass
4441 auto - Kernel detects whether the CPU model contains an
4442 implementation of Speculative Store Bypass and
4443 picks the most appropriate mitigation. If the
4444 CPU is not vulnerable, "off" is selected. If the
4445 CPU is vulnerable the default mitigation is
4446 architecture and Kconfig dependent. See below.
4447 prctl - Control Speculative Store Bypass per thread
4448 via prctl. Speculative Store Bypass is enabled
4449 for a process by default. The state of the control
4450 is inherited on fork.
4451 seccomp - Same as "prctl" above, but all seccomp threads
4452 will disable SSB unless they explicitly opt out.
4454 Default mitigations:
4455 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4457 On powerpc the options are:
4459 on,auto - On Power8 and Power9 insert a store-forwarding
4460 barrier on kernel entry and exit. On Power7
4461 perform a software flush on kernel entry and
4465 Not specifying this option is equivalent to
4466 spec_store_bypass_disable=auto.
4468 spia_io_base= [HW,MTD]
4473 srcutree.counter_wrap_check [KNL]
4474 Specifies how frequently to check for
4475 grace-period sequence counter wrap for the
4476 srcu_data structure's ->srcu_gp_seq_needed field.
4477 The greater the number of bits set in this kernel
4478 parameter, the less frequently counter wrap will
4479 be checked for. Note that the bottom two bits
4482 srcutree.exp_holdoff [KNL]
4483 Specifies how many nanoseconds must elapse
4484 since the end of the last SRCU grace period for
4485 a given srcu_struct until the next normal SRCU
4486 grace period will be considered for automatic
4487 expediting. Set to zero to disable automatic
4491 Speculative Store Bypass Disable control
4493 On CPUs that are vulnerable to the Speculative
4494 Store Bypass vulnerability and offer a
4495 firmware based mitigation, this parameter
4496 indicates how the mitigation should be used:
4498 force-on: Unconditionally enable mitigation for
4499 for both kernel and userspace
4500 force-off: Unconditionally disable mitigation for
4501 for both kernel and userspace
4502 kernel: Always enable mitigation in the
4503 kernel, and offer a prctl interface
4504 to allow userspace to register its
4505 interest in being mitigated too.
4507 stack_guard_gap= [MM]
4508 override the default stack gap protection. The value
4509 is in page units and it defines how many pages prior
4510 to (for stacks growing down) resp. after (for stacks
4511 growing up) the main stack are reserved for no other
4512 mapping. Default value is 256 pages.
4515 Enabled the stack tracer on boot up.
4517 stacktrace_filter=[function-list]
4518 [FTRACE] Limit the functions that the stack tracer
4519 will trace at boot up. function-list is a comma separated
4520 list of functions. This list can be changed at run
4521 time by the stack_trace_filter file in the debugfs
4522 tracing directory. Note, this enables stack tracing
4523 and the stacktrace above is not needed.
4527 Set the STI (builtin display/keyboard on the HP-PARISC
4528 machines) console (graphic card) which should be used
4529 as the initial boot-console.
4530 See also comment in drivers/video/console/sticore.c.
4533 See comment in drivers/video/console/sticore.c.
4536 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4538 sunrpc.min_resvport=
4539 sunrpc.max_resvport=
4541 SunRPC servers often require that client requests
4542 originate from a privileged port (i.e. a port in the
4543 range 0 < portnr < 1024).
4544 An administrator who wishes to reserve some of these
4545 ports for other uses may adjust the range that the
4546 kernel's sunrpc client considers to be privileged
4547 using these two parameters to set the minimum and
4548 maximum port values.
4550 sunrpc.svc_rpc_per_connection_limit=
4552 Limit the number of requests that the server will
4553 process in parallel from a single connection.
4554 The default value is 0 (no limit).
4558 Control how the NFS server code allocates CPUs to
4559 service thread pools. Depending on how many NICs
4560 you have and where their interrupts are bound, this
4561 option will affect which CPUs will do NFS serving.
4562 Note: this parameter cannot be changed while the
4563 NFS server is running.
4565 auto the server chooses an appropriate mode
4566 automatically using heuristics
4567 global a single global pool contains all CPUs
4568 percpu one pool for each CPU
4569 pernode one pool for each NUMA node (equivalent
4570 to global on non-NUMA machines)
4572 sunrpc.tcp_slot_table_entries=
4573 sunrpc.udp_slot_table_entries=
4575 Sets the upper limit on the number of simultaneous
4576 RPC calls that can be sent from the client to a
4577 server. Increasing these values may allow you to
4578 improve throughput, but will also increase the
4579 amount of memory reserved for use by the client.
4581 suspend.pm_test_delay=
4583 Sets the number of seconds to remain in a suspend test
4584 mode before resuming the system (see
4585 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4586 is set. Default value is 5.
4589 [KNL] Enable accounting of swap in memory resource
4590 controller if no parameter or 1 is given or disable
4591 it if 0 is given (See Documentation/cgroup-v1/memory.txt)
4593 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4594 Format: { <int> | force | noforce }
4595 <int> -- Number of I/O TLB slabs
4596 force -- force using of bounce buffers even if they
4597 wouldn't be automatically used by the kernel
4598 noforce -- Never use bounce buffers (for debugging)
4602 sysfs.deprecated=0|1 [KNL]
4603 Enable/disable old style sysfs layout for old udev
4604 on older distributions. When this option is enabled
4605 very new udev will not work anymore. When this option
4606 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4607 in older udev will not work anymore.
4608 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4609 the kernel configuration.
4611 sysrq_always_enabled
4613 Ignore sysrq setting - this boot parameter will
4614 neutralize any effect of /proc/sys/kernel/sysrq.
4615 Useful for debugging.
4617 tcpmhash_entries= [KNL,NET]
4618 Set the number of tcp_metrics_hash slots.
4619 Default value is 8192 or 16384 depending on total
4620 ram pages. This is used to specify the TCP metrics
4621 cache size. See Documentation/networking/ip-sysctl.txt
4622 "tcp_no_metrics_save" section for more details.
4626 test_suspend= [SUSPEND][,N]
4627 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4628 standby suspend) or "freeze" (for suspend type freeze)
4629 as the system sleep state during system startup with
4630 the optional capability to repeat N number of times.
4631 The system is woken from this state using a
4632 wakeup-capable RTC alarm.
4634 thash_entries= [KNL,NET]
4635 Set number of hash buckets for TCP connection
4637 thermal.act= [HW,ACPI]
4638 -1: disable all active trip points in all thermal zones
4639 <degrees C>: override all lowest active trip points
4641 thermal.crt= [HW,ACPI]
4642 -1: disable all critical trip points in all thermal zones
4643 <degrees C>: override all critical trip points
4645 thermal.nocrt= [HW,ACPI]
4646 Set to disable actions on ACPI thermal zone
4647 critical and hot trip points.
4649 thermal.off= [HW,ACPI]
4650 1: disable ACPI thermal control
4652 thermal.psv= [HW,ACPI]
4653 -1: disable all passive trip points
4654 <degrees C>: override all passive trip points to this
4657 thermal.tzp= [HW,ACPI]
4658 Specify global default ACPI thermal zone polling rate
4659 <deci-seconds>: poll all this frequency
4660 0: no polling (default)
4663 Force threading of all interrupt handlers except those
4664 marked explicitly IRQF_NO_THREAD.
4667 Enable the Transcendent memory driver if built-in.
4669 tmem.cleancache=0|1 [KNL, XEN]
4670 Default is on (1). Disable the usage of the cleancache
4671 API to send anonymous pages to the hypervisor.
4673 tmem.frontswap=0|1 [KNL, XEN]
4674 Default is on (1). Disable the usage of the frontswap
4675 API to send swap pages to the hypervisor. If disabled
4676 the selfballooning and selfshrinking are force disabled.
4678 tmem.selfballooning=0|1 [KNL, XEN]
4679 Default is on (1). Disable the driving of swap pages
4682 tmem.selfshrinking=0|1 [KNL, XEN]
4683 Default is on (1). Partial swapoff that immediately
4684 transfers pages from Xen hypervisor back to the
4685 kernel based on different criteria.
4689 Specify if the kernel should make use of the cpu
4690 topology information if the hardware supports this.
4691 The scheduler will make use of this information and
4692 e.g. base its process migration decisions on it.
4695 topology_updates= [KNL, PPC, NUMA]
4697 Specify if the kernel should ignore (off)
4698 topology updates sent by the hypervisor to this
4703 tpm_suspend_pcr=[HW,TPM]
4704 Format: integer pcr id
4705 Specify that at suspend time, the tpm driver
4706 should extend the specified pcr with zeros,
4707 as a workaround for some chips which fail to
4708 flush the last written pcr on TPM_SaveState.
4709 This will guarantee that all the other pcrs
4712 trace_buf_size=nn[KMG]
4713 [FTRACE] will set tracing buffer size on each cpu.
4715 trace_event=[event-list]
4716 [FTRACE] Set and start specified trace events in order
4717 to facilitate early boot debugging. The event-list is a
4718 comma separated list of trace events to enable. See
4719 also Documentation/trace/events.rst
4721 trace_options=[option-list]
4722 [FTRACE] Enable or disable tracer options at boot.
4723 The option-list is a comma delimited list of options
4724 that can be enabled or disabled just as if you were
4725 to echo the option name into
4727 /sys/kernel/debug/tracing/trace_options
4729 For example, to enable stacktrace option (to dump the
4730 stack trace of each event), add to the command line:
4732 trace_options=stacktrace
4734 See also Documentation/trace/ftrace.rst "trace options"
4738 Have the tracepoints sent to printk as well as the
4739 tracing ring buffer. This is useful for early boot up
4740 where the system hangs or reboots and does not give the
4741 option for reading the tracing buffer or performing a
4742 ftrace_dump_on_oops.
4744 To turn off having tracepoints sent to printk,
4745 echo 0 > /proc/sys/kernel/tracepoint_printk
4746 Note, echoing 1 into this file without the
4747 tracepoint_printk kernel cmdline option has no effect.
4751 Having tracepoints sent to printk() and activating high
4752 frequency tracepoints such as irq or sched, can cause
4753 the system to live lock.
4756 [FTRACE] enable this option to disable tracing when a
4757 warning is hit. This turns off "tracing_on". Tracing can
4758 be enabled again by echoing '1' into the "tracing_on"
4759 file located in /sys/kernel/debug/tracing/
4761 This option is useful, as it disables the trace before
4762 the WARNING dump is called, which prevents the trace to
4763 be filled with content caused by the warning output.
4765 This option can also be set at run time via the sysctl
4766 option: kernel/traceoff_on_warning
4768 transparent_hugepage=
4770 Format: [always|madvise|never]
4771 Can be used to control the default behavior of the system
4772 with respect to transparent hugepages.
4773 See Documentation/admin-guide/mm/transhuge.rst
4776 tsc= Disable clocksource stability checks for TSC.
4778 [x86] reliable: mark tsc clocksource as reliable, this
4779 disables clocksource verification at runtime, as well
4780 as the stability checks done at bootup. Used to enable
4781 high-resolution timer mode on older hardware, and in
4782 virtualized environment.
4783 [x86] noirqtime: Do not use TSC to do irq accounting.
4784 Used to run time disable IRQ_TIME_ACCOUNTING on any
4785 platforms where RDTSC is slow and this accounting
4787 [x86] unstable: mark the TSC clocksource as unstable, this
4788 marks the TSC unconditionally unstable at bootup and
4789 avoids any further wobbles once the TSC watchdog notices.
4790 [x86] nowatchdog: disable clocksource watchdog. Used
4791 in situations with strict latency requirements (where
4792 interruptions from clocksource watchdog are not
4795 turbografx.map[2|3]= [HW,JOY]
4796 TurboGraFX parallel port interface
4798 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4799 See also Documentation/input/devices/joystick-parport.rst
4801 udbg-immortal [PPC] When debugging early kernel crashes that
4802 happen after console_init() and before a proper
4803 console driver takes over, this boot options might
4804 help "seeing" what's going on.
4806 uhash_entries= [KNL,NET]
4807 Set number of hash buckets for UDP/UDP-Lite connections
4810 [USB] Ignore overcurrent events (default N).
4811 Some badly-designed motherboards generate lots of
4812 bogus events, for ports that aren't wired to
4813 anything. Set this parameter to avoid log spamming.
4814 Note that genuine overcurrent events won't be
4818 [X86] Cause panic on unknown NMI.
4820 usbcore.authorized_default=
4821 [USB] Default USB device authorization:
4822 (default -1 = authorized except for wireless USB,
4823 0 = not authorized, 1 = authorized, 2 = authorized
4824 if device connected to internal port)
4826 usbcore.autosuspend=
4827 [USB] The autosuspend time delay (in seconds) used
4828 for newly-detected USB devices (default 2). This
4829 is the time required before an idle device will be
4830 autosuspended. Devices for which the delay is set
4831 to a negative value won't be autosuspended at all.
4833 usbcore.usbfs_snoop=
4834 [USB] Set to log all usbfs traffic (default 0 = off).
4836 usbcore.usbfs_snoop_max=
4837 [USB] Maximum number of bytes to snoop in each URB
4840 usbcore.blinkenlights=
4841 [USB] Set to cycle leds on hubs (default 0 = off).
4843 usbcore.old_scheme_first=
4844 [USB] Start with the old device initialization
4845 scheme, applies only to low and full-speed devices
4848 usbcore.usbfs_memory_mb=
4849 [USB] Memory limit (in MB) for buffers allocated by
4850 usbfs (default = 16, 0 = max = 2047).
4852 usbcore.use_both_schemes=
4853 [USB] Try the other device initialization scheme
4854 if the first one fails (default 1 = enabled).
4856 usbcore.initial_descriptor_timeout=
4857 [USB] Specifies timeout for the initial 64-byte
4858 USB_REQ_GET_DESCRIPTOR request in milliseconds
4859 (default 5000 = 5.0 seconds).
4861 usbcore.nousb [USB] Disable the USB subsystem
4864 [USB] A list of quirk entries to augment the built-in
4865 usb core quirk list. List entries are separated by
4866 commas. Each entry has the form
4867 VendorID:ProductID:Flags. The IDs are 4-digit hex
4868 numbers and Flags is a set of letters. Each letter
4869 will change the built-in quirk; setting it if it is
4870 clear and clearing it if it is set. The letters have
4871 the following meanings:
4872 a = USB_QUIRK_STRING_FETCH_255 (string
4873 descriptors must not be fetched using
4875 b = USB_QUIRK_RESET_RESUME (device can't resume
4876 correctly so reset it instead);
4877 c = USB_QUIRK_NO_SET_INTF (device can't handle
4878 Set-Interface requests);
4879 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
4880 handle its Configuration or Interface
4882 e = USB_QUIRK_RESET (device can't be reset
4883 (e.g morph devices), don't use reset);
4884 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
4885 more interface descriptions than the
4886 bNumInterfaces count, and can't handle
4887 talking to these interfaces);
4888 g = USB_QUIRK_DELAY_INIT (device needs a pause
4889 during initialization, after we read
4890 the device descriptor);
4891 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
4892 high speed and super speed interrupt
4893 endpoints, the USB 2.0 and USB 3.0 spec
4894 require the interval in microframes (1
4895 microframe = 125 microseconds) to be
4896 calculated as interval = 2 ^
4898 Devices with this quirk report their
4899 bInterval as the result of this
4900 calculation instead of the exponent
4901 variable used in the calculation);
4902 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
4903 handle device_qualifier descriptor
4905 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
4906 generates spurious wakeup, ignore
4907 remote wakeup capability);
4908 k = USB_QUIRK_NO_LPM (device can't handle Link
4910 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
4911 (Device reports its bInterval as linear
4912 frames instead of the USB 2.0
4914 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
4915 to be disconnected before suspend to
4916 prevent spurious wakeup);
4917 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
4918 pause after every control message);
4919 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
4920 delay after resetting its port);
4921 Example: quirks=0781:5580:bk,0a5c:5834:gij
4924 [USBHID] The interval which mice are to be polled at.
4927 [USBHID] The interval which joysticks are to be polled at.
4930 [USBHID] The interval which keyboards are to be polled at.
4932 usb-storage.delay_use=
4933 [UMS] The delay in seconds before a new device is
4934 scanned for Logical Units (default 1).
4937 [UMS] A list of quirks entries to supplement or
4938 override the built-in unusual_devs list. List
4939 entries are separated by commas. Each entry has
4940 the form VID:PID:Flags where VID and PID are Vendor
4941 and Product ID values (4-digit hex numbers) and
4942 Flags is a set of characters, each corresponding
4943 to a common usb-storage quirk flag as follows:
4944 a = SANE_SENSE (collect more than 18 bytes
4946 b = BAD_SENSE (don't collect more than 18
4947 bytes of sense data);
4948 c = FIX_CAPACITY (decrease the reported
4949 device capacity by one sector);
4950 d = NO_READ_DISC_INFO (don't use
4951 READ_DISC_INFO command);
4952 e = NO_READ_CAPACITY_16 (don't use
4953 READ_CAPACITY_16 command);
4954 f = NO_REPORT_OPCODES (don't use report opcodes
4956 g = MAX_SECTORS_240 (don't transfer more than
4957 240 sectors at a time, uas only);
4958 h = CAPACITY_HEURISTICS (decrease the
4959 reported device capacity by one
4960 sector if the number is odd);
4961 i = IGNORE_DEVICE (don't bind to this
4963 j = NO_REPORT_LUNS (don't use report luns
4965 l = NOT_LOCKABLE (don't try to lock and
4966 unlock ejectable media);
4967 m = MAX_SECTORS_64 (don't transfer more
4968 than 64 sectors = 32 KB at a time);
4969 n = INITIAL_READ10 (force a retry of the
4970 initial READ(10) command);
4971 o = CAPACITY_OK (accept the capacity
4972 reported by the device);
4973 p = WRITE_CACHE (the device cache is ON
4975 r = IGNORE_RESIDUE (the device reports
4976 bogus residue values);
4977 s = SINGLE_LUN (the device has only one
4979 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
4980 commands, uas only);
4981 u = IGNORE_UAS (don't bind to the uas driver);
4982 w = NO_WP_DETECT (don't test whether the
4983 medium is write-protected).
4984 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
4985 even if the device claims no cache)
4986 Example: quirks=0419:aaf5:rl,0421:0433:rc
4988 user_debug= [KNL,ARM]
4990 See arch/arm/Kconfig.debug help text.
4991 1 - undefined instruction events
4993 4 - invalid data aborts
4996 Example: user_debug=31
4999 [X86] Flags controlling user PTE allocations.
5001 nohigh = do not allocate PTE pages in
5002 HIGHMEM regardless of setting
5006 On X86_32, this is an alias for vdso32=. Otherwise:
5008 vdso=1: enable VDSO (the default)
5009 vdso=0: disable VDSO mapping
5011 vdso32= [X86] Control the 32-bit vDSO
5012 vdso32=1: enable 32-bit VDSO
5013 vdso32=0 or vdso32=2: disable 32-bit VDSO
5015 See the help text for CONFIG_COMPAT_VDSO for more
5016 details. If CONFIG_COMPAT_VDSO is set, the default is
5017 vdso32=0; otherwise, the default is vdso32=1.
5019 For compatibility with older kernels, vdso32=2 is an
5022 Try vdso32=0 if you encounter an error that says:
5023 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
5026 vector=percpu: enable percpu vector domain
5028 video= [FB] Frame buffer configuration
5029 See Documentation/fb/modedb.txt.
5031 video.brightness_switch_enabled= [0,1]
5032 If set to 1, on receiving an ACPI notify event
5033 generated by hotkey, video driver will adjust brightness
5034 level and then send out the event to user space through
5035 the allocated input device; If set to 0, video driver
5036 will only send out the event without touching backlight
5041 [VMMIO] Memory mapped virtio (platform) device.
5043 <size>@<baseaddr>:<irq>[:<id>]
5045 <size> := size (can use standard suffixes
5047 <baseaddr> := physical base address
5048 <irq> := interrupt number (as passed to
5050 <id> := (optional) platform device id
5052 virtio_mmio.device=1K@0x100b0000:48:7
5054 Can be used multiple times for multiple devices.
5056 vga= [BOOT,X86-32] Select a particular video mode
5057 See Documentation/x86/boot.txt and
5058 Documentation/svga.txt.
5059 Use vga=ask for menu.
5060 This is actually a boot loader parameter; the value is
5061 passed to the kernel using a special protocol.
5063 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
5064 May slow down system boot speed, especially when
5065 enabled on systems with a large amount of memory.
5066 All options are enabled by default, and this
5067 interface is meant to allow for selectively
5068 enabling or disabling specific virtual memory
5071 Available options are:
5072 P Enable page structure init time poisoning
5073 - Disable all of the above options
5075 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
5076 size of <nn>. This can be used to increase the
5077 minimum size (128MB on x86). It can also be used to
5078 decrease the size and leave more room for directly
5081 vmcp_cma=nn[MG] [KNL,S390]
5082 Sets the memory size reserved for contiguous memory
5083 allocations for the vmcp device driver.
5085 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
5088 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
5091 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
5095 Controls the behavior of vsyscalls (i.e. calls to
5096 fixed addresses of 0xffffffffff600x00 from legacy
5097 code). Most statically-linked binaries and older
5098 versions of glibc use these calls. Because these
5099 functions are at fixed addresses, they make nice
5100 targets for exploits that can control RIP.
5102 emulate [default] Vsyscalls turn into traps and are
5103 emulated reasonably safely.
5105 native Vsyscalls are native syscall instructions.
5106 This is a little bit faster than trapping
5107 and makes a few dynamic recompilers work
5108 better than they would in emulation mode.
5109 It also makes exploits much easier to write.
5111 none Vsyscalls don't work at all. This makes
5112 them quite hard to use for exploits but
5113 might break your system.
5115 vt.color= [VT] Default text color.
5116 Format: 0xYX, X = foreground, Y = background.
5117 Default: 0x07 = light gray on black.
5119 vt.cur_default= [VT] Default cursor shape.
5120 Format: 0xCCBBAA, where AA, BB, and CC are the same as
5121 the parameters of the <Esc>[?A;B;Cc escape sequence;
5122 see VGA-softcursor.txt. Default: 2 = underline.
5124 vt.default_blu= [VT]
5125 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
5126 Change the default blue palette of the console.
5127 This is a 16-member array composed of values
5130 vt.default_grn= [VT]
5131 Format: <green0>,<green1>,<green2>,...,<green15>
5132 Change the default green palette of the console.
5133 This is a 16-member array composed of values
5136 vt.default_red= [VT]
5137 Format: <red0>,<red1>,<red2>,...,<red15>
5138 Change the default red palette of the console.
5139 This is a 16-member array composed of values
5145 Set system-wide default UTF-8 mode for all tty's.
5146 Default is 1, i.e. UTF-8 mode is enabled for all
5147 newly opened terminals.
5149 vt.global_cursor_default=
5152 Set system-wide default for whether a cursor
5153 is shown on new VTs. Default is -1,
5154 i.e. cursors will be created by default unless
5155 overridden by individual drivers. 0 will hide
5156 cursors, 1 will display them.
5158 vt.italic= [VT] Default color for italic text; 0-15.
5161 vt.underline= [VT] Default color for underlined text; 0-15.
5164 watchdog timers [HW,WDT] For information on watchdog timers,
5165 see Documentation/watchdog/watchdog-parameters.txt
5166 or other driver-specific files in the
5167 Documentation/watchdog/ directory.
5171 Set the hard lockup detector stall duration
5172 threshold in seconds. The soft lockup detector
5173 threshold is set to twice the value. A value of 0
5174 disables both lockup detectors. Default is 10
5177 workqueue.watchdog_thresh=
5178 If CONFIG_WQ_WATCHDOG is configured, workqueue can
5179 warn stall conditions and dump internal state to
5180 help debugging. 0 disables workqueue stall
5181 detection; otherwise, it's the stall threshold
5182 duration in seconds. The default value is 30 and
5183 it can be updated at runtime by writing to the
5184 corresponding sysfs file.
5186 workqueue.disable_numa
5187 By default, all work items queued to unbound
5188 workqueues are affine to the NUMA nodes they're
5189 issued on, which results in better behavior in
5190 general. If NUMA affinity needs to be disabled for
5191 whatever reason, this option can be used. Note
5192 that this also can be controlled per-workqueue for
5193 workqueues visible under /sys/bus/workqueue/.
5195 workqueue.power_efficient
5196 Per-cpu workqueues are generally preferred because
5197 they show better performance thanks to cache
5198 locality; unfortunately, per-cpu workqueues tend to
5199 be more power hungry than unbound workqueues.
5201 Enabling this makes the per-cpu workqueues which
5202 were observed to contribute significantly to power
5203 consumption unbound, leading to measurably lower
5204 power usage at the cost of small performance
5207 The default value of this parameter is determined by
5208 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
5210 workqueue.debug_force_rr_cpu
5211 Workqueue used to implicitly guarantee that work
5212 items queued without explicit CPU specified are put
5213 on the local CPU. This guarantee is no longer true
5214 and while local CPU is still preferred work items
5215 may be put on foreign CPUs. This debug option
5216 forces round-robin CPU selection to flush out
5217 usages which depend on the now broken guarantee.
5218 When enabled, memory and cache locality will be
5221 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
5222 default x2apic cluster mode on platforms
5225 x86_intel_mid_timer= [X86-32,APBT]
5226 Choose timer option for x86 Intel MID platform.
5227 Two valid options are apbt timer only and lapic timer
5228 plus one apbt timer for broadcast timer.
5229 x86_intel_mid_timer=apbt_only | lapic_and_apbt
5231 xen_512gb_limit [KNL,X86-64,XEN]
5232 Restricts the kernel running paravirtualized under Xen
5233 to use only up to 512 GB of RAM. The reason to do so is
5234 crash analysis tools and Xen tools for doing domain
5235 save/restore/migration must be enabled to handle larger
5238 xen_emul_unplug= [HW,X86,XEN]
5239 Unplug Xen emulated devices
5240 Format: [unplug0,][unplug1]
5241 ide-disks -- unplug primary master IDE devices
5242 aux-ide-disks -- unplug non-primary-master IDE devices
5243 nics -- unplug network devices
5244 all -- unplug all emulated devices (NICs and IDE disks)
5245 unnecessary -- unplugging emulated devices is
5246 unnecessary even if the host did not respond to
5248 never -- do not unplug even if version check succeeds
5250 xen_nopvspin [X86,XEN]
5251 Disables the ticketlock slowpath using Xen PV
5255 Disables the PV optimizations forcing the HVM guest to
5256 run as generic HVM guest with no PV drivers.
5258 xen_scrub_pages= [XEN]
5259 Boolean option to control scrubbing pages before giving them back
5260 to Xen, for use by other domains. Can be also changed at runtime
5261 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
5262 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
5264 xen_timer_slop= [X86-64,XEN]
5265 Set the timer slop (in nanoseconds) for the virtual Xen
5266 timers (default is 100000). This adjusts the minimum
5267 delta of virtualized Xen timers, where lower values
5268 improve timer resolution at the expense of processing
5269 more timer interrupts.
5271 xirc2ps_cs= [NET,PCMCIA]
5273 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
5275 xhci-hcd.quirks [USB,KNL]
5276 A hex value specifying bitmask with supplemental xhci
5277 host controller quirks. Meaning of each bit can be
5278 consulted in header drivers/usb/host/xhci.h.