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.rst, 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]
25 { vendor | video | native | none }
26 If set to vendor, prefer vendor-specific driver
27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
28 of the ACPI video.ko driver.
29 If set to video, use the ACPI video.ko driver.
30 If set to native, use the device's native backlight mode.
31 If set to none, disable the ACPI backlight interface.
33 acpi_force_32bit_fadt_addr
34 force FADT to use 32 bit addresses rather than the
35 64 bit X_* addresses. Some firmware have broken 64
36 bit addresses for force ACPI ignore these and use
37 the older legacy 32 bit addresses.
39 acpica_no_return_repair [HW, ACPI]
40 Disable AML predefined validation mechanism
41 This mechanism can repair the evaluation result to make
42 the return objects more ACPI specification compliant.
43 This option is useful for developers to identify the
44 root cause of an AML interpreter issue when the issue
45 has something to do with the repair mechanism.
47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
51 debug output. Bits in debug_layer correspond to a
52 _COMPONENT in an ACPI source file, e.g.,
53 #define _COMPONENT ACPI_PCI_COMPONENT
54 Bits in debug_level correspond to a level in
55 ACPI_DEBUG_PRINT statements, e.g.,
56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
57 The debug_level mask defaults to "info". See
58 Documentation/firmware-guide/acpi/debug.rst for more information about
59 debug layers and levels.
61 Enable processor driver info messages:
62 acpi.debug_layer=0x20000000
63 Enable PCI/PCI interrupt routing info messages:
64 acpi.debug_layer=0x400000
65 Enable AML "Debug" output, i.e., stores to the Debug
66 object while interpreting AML:
67 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
68 Enable all messages related to ACPI hardware:
69 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
71 Some values produce so much output that the system is
72 unusable. The "log_buf_len" parameter may be useful
73 if you need to capture more output.
75 acpi_enforce_resources= [ACPI]
77 Check for resource conflicts between native drivers
78 and ACPI OperationRegions (SystemIO and SystemMemory
79 only). IO ports and memory declared in ACPI might be
80 used by the ACPI subsystem in arbitrary AML code and
81 can interfere with legacy drivers.
82 strict (default): access to resources claimed by ACPI
83 is denied; legacy drivers trying to access reserved
84 resources will fail to bind to device using them.
85 lax: access to resources claimed by ACPI is allowed;
86 legacy drivers trying to access reserved resources
87 will bind successfully but a warning message is logged.
88 no: ACPI OperationRegions are not marked as reserved,
89 no further checks are performed.
91 acpi_force_table_verification [HW,ACPI]
92 Enable table checksum verification during early stage.
93 By default, this is disabled due to x86 early mapping
96 acpi_irq_balance [HW,ACPI]
97 ACPI will balance active IRQs
100 acpi_irq_nobalance [HW,ACPI]
101 ACPI will not move active IRQs (default)
104 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
105 Format: <irq>,<irq>...
107 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
109 Format: <irq>,<irq>...
111 acpi_mask_gpe= [HW,ACPI]
112 Due to the existence of _Lxx/_Exx, some GPEs triggered
113 by unsupported hardware/firmware features can result in
114 GPE floodings that cannot be automatically disabled by
116 This facility can be used to prevent such uncontrolled
120 acpi_no_auto_serialize [HW,ACPI]
121 Disable auto-serialization of AML methods
122 AML control methods that contain the opcodes to create
123 named objects will be marked as "Serialized" by the
124 auto-serialization feature.
125 This feature is enabled by default.
126 This option allows to turn off the feature.
128 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
131 acpi_no_static_ssdt [HW,ACPI]
132 Disable installation of static SSDTs at early boot time
133 By default, SSDTs contained in the RSDT/XSDT will be
134 installed automatically and they will appear under
135 /sys/firmware/acpi/tables.
136 This option turns off this feature.
137 Note that specifying this option does not affect
138 dynamic table installation which will install SSDT
139 tables to /sys/firmware/acpi/tables/dynamic.
141 acpi_no_watchdog [HW,ACPI,WDT]
142 Ignore the ACPI-based watchdog interface (WDAT) and let
143 a native driver control the watchdog device instead.
145 acpi_rsdp= [ACPI,EFI,KEXEC]
146 Pass the RSDP address to the kernel, mostly used
147 on machines running EFI runtime service to boot the
148 second kernel for kdump.
150 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
151 Format: To spoof as Windows 98: ="Microsoft Windows"
153 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
154 of 2 which is mandated by ACPI 6) as the supported ACPI
155 specification revision (when using this switch, it may
156 be necessary to carry out a cold reboot _twice_ in a
157 row to make it take effect on the platform firmware).
159 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
160 acpi_osi="string1" # add string1
161 acpi_osi="!string2" # remove string2
162 acpi_osi=!* # remove all strings
163 acpi_osi=! # disable all built-in OS vendor
165 acpi_osi=!! # enable all built-in OS vendor
167 acpi_osi= # disable all strings
169 'acpi_osi=!' can be used in combination with single or
170 multiple 'acpi_osi="string1"' to support specific OS
171 vendor string(s). Note that such command can only
172 affect the default state of the OS vendor strings, thus
173 it cannot affect the default state of the feature group
174 strings and the current state of the OS vendor strings,
175 specifying it multiple times through kernel command line
176 is meaningless. This command is useful when one do not
177 care about the state of the feature group strings which
178 should be controlled by the OSPM.
180 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
181 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
182 can make '_OSI("Windows 2000")' TRUE.
184 'acpi_osi=' cannot be used in combination with other
185 'acpi_osi=' command lines, the _OSI method will not
186 exist in the ACPI namespace. NOTE that such command can
187 only affect the _OSI support state, thus specifying it
188 multiple times through kernel command line is also
191 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
194 'acpi_osi=!*' can be used in combination with single or
195 multiple 'acpi_osi="string1"' to support specific
196 string(s). Note that such command can affect the
197 current state of both the OS vendor strings and the
198 feature group strings, thus specifying it multiple times
199 through kernel command line is meaningful. But it may
200 still not able to affect the final state of a string if
201 there are quirks related to this string. This command
202 is useful when one want to control the state of the
203 feature group strings to debug BIOS issues related to
206 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
207 '_OSI("Module Device")' FALSE.
208 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
209 '_OSI("Module Device")' TRUE.
210 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
212 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
214 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
215 they all will make '_OSI("Windows 2000")' TRUE.
218 Override the pmtimer bug detection: force the kernel
219 to assume that this machine's pmtimer latches its value
220 and always returns good values.
222 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
223 Format: { level | edge | high | low }
225 acpi_skip_timer_override [HW,ACPI]
226 Recognize and ignore IRQ0/pin2 Interrupt Override.
227 For broken nForce2 BIOS resulting in XT-PIC timer.
229 acpi_sleep= [HW,ACPI] Sleep options
230 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
231 old_ordering, nonvs, sci_force_enable, nobl }
232 See Documentation/power/video.rst for information on
234 s3_beep is for debugging; it makes the PC's speaker beep
235 as soon as the kernel's real-mode entry point is called.
236 s4_nohwsig prevents ACPI hardware signature from being
237 used during resume from hibernation.
238 old_ordering causes the ACPI 1.0 ordering of the _PTS
239 control method, with respect to putting devices into
240 low power states, to be enforced (the ACPI 2.0 ordering
241 of _PTS is used by default).
242 nonvs prevents the kernel from saving/restoring the
243 ACPI NVS memory during suspend/hibernation and resume.
244 sci_force_enable causes the kernel to set SCI_EN directly
245 on resume from S1/S3 (which is against the ACPI spec,
246 but some broken systems don't work without it).
247 nobl causes the internal blacklist of systems known to
248 behave incorrectly in some ways with respect to system
249 suspend and resume to be ignored (use wisely).
251 acpi_use_timer_override [HW,ACPI]
252 Use timer override. For some broken Nvidia NF5 boards
253 that require a timer override, but don't have HPET
255 add_efi_memmap [EFI; X86] Include EFI memory map in
256 kernel's map of available physical RAM.
259 { off | try_unsupported }
260 off: disable AGP support
261 try_unsupported: try to drive unsupported chipsets
262 (may crash computer or cause data corruption)
265 See Documentation/sound/alsa-configuration.rst
268 Allow the default userspace alignment fault handler
269 behaviour to be specified. Bit 0 enables warnings,
270 bit 1 enables fixups, and bit 2 sends a segfault.
272 align_va_addr= [X86-64]
273 Align virtual addresses by clearing slice [14:12] when
274 allocating a VMA at process creation time. This option
275 gives you up to 3% performance improvement on AMD F15h
276 machines (where it is enabled by default) for a
277 CPU-intensive style benchmark, and it can vary highly in
278 a microbenchmark depending on workload and compiler.
280 32: only for 32-bit processes
281 64: only for 64-bit processes
282 on: enable for both 32- and 64-bit processes
283 off: disable for both 32- and 64-bit processes
285 alloc_snapshot [FTRACE]
286 Allocate the ftrace snapshot buffer on boot up when the
287 main buffer is allocated. This is handy if debugging
288 and you need to use tracing_snapshot() on boot up, and
289 do not want to use tracing_snapshot_alloc() as it needs
290 to be done where GFP_KERNEL allocations are allowed.
292 amd_iommu= [HW,X86-64]
293 Pass parameters to the AMD IOMMU driver in the system.
295 fullflush - enable flushing of IO/TLB entries when
296 they are unmapped. Otherwise they are
297 flushed before they will be reused, which
299 off - do not initialize any AMD IOMMU found in
301 force_isolation - Force device isolation for all
302 devices. The IOMMU driver is not
303 allowed anymore to lift isolation
304 requirements as needed. This option
305 does not override iommu=pt
307 amd_iommu_dump= [HW,X86-64]
308 Enable AMD IOMMU driver option to dump the ACPI table
309 for AMD IOMMU. With this option enabled, AMD IOMMU
310 driver will print ACPI tables for AMD IOMMU during
311 IOMMU initialization.
313 amd_iommu_intr= [HW,X86-64]
314 Specifies one of the following AMD IOMMU interrupt
316 legacy - Use legacy interrupt remapping mode.
317 vapic - Use virtual APIC mode, which allows IOMMU
318 to inject interrupts directly into guest.
319 This mode requires kvm-amd.avic=1.
320 (Default when IOMMU HW support is present.)
322 amijoy.map= [HW,JOY] Amiga joystick support
323 Map of devices attached to JOY0DAT and JOY1DAT
325 See also Documentation/input/joydev/joystick.rst
327 analog.map= [HW,JOY] Analog joystick and gamepad support
328 Specifies type or capabilities of an analog joystick
329 connected to one of 16 gameports
330 Format: <type1>,<type2>,..<type16>
333 Power management functions (SPARCstation-4/5 + deriv.)
335 Disable APC CPU standby support. SPARCstation-Fox does
336 not play well with APC CPU idle - disable it if you have
337 APC and your system crashes randomly.
339 apic= [APIC,X86] Advanced Programmable Interrupt Controller
340 Change the output verbosity while booting
341 Format: { quiet (default) | verbose | debug }
342 Change the amount of debugging information output
343 when initialising the APIC and IO-APIC components.
344 For X86-32, this can also be used to specify an APIC
346 Format: apic=driver_name
347 Examples: apic=bigsmp
349 apic_extnmi= [APIC,X86] External NMI delivery setting
350 Format: { bsp (default) | all | none }
351 bsp: External NMI is delivered only to CPU 0
352 all: External NMIs are broadcast to all CPUs as a
354 none: External NMI is masked for all CPUs. This is
355 useful so that a dump capture kernel won't be
359 See Documentation/networking/ipv6.rst.
361 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
362 Limit apic dumping. The parameter defines the maximal
363 number of local apics being dumped. Also it is possible
364 to set it to "all" by meaning -- no limit here.
365 Format: { 1 (default) | 2 | ... | all }.
366 The parameter valid if only apic=debug or
367 apic=verbose is specified.
368 Example: apic=debug show_lapic=all
370 apm= [APM] Advanced Power Management
371 See header of arch/x86/kernel/apm_32.c.
373 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
374 Format: <io>,<irq>,<nodeID>
378 atarimouse= [HW,MOUSE] Atari Mouse
380 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
381 EzKey and similar keyboards
383 atkbd.reset= [HW] Reset keyboard during initialization
385 atkbd.set= [HW] Select keyboard code set
386 Format: <int> (2 = AT (default), 3 = PS/2)
388 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
391 atkbd.softraw= [HW] Choose between synthetic and real raw mode
392 Format: <bool> (0 = real, 1 = synthetic (default))
394 atkbd.softrepeat= [HW]
395 Use software keyboard repeat
397 audit= [KNL] Enable the audit sub-system
398 Format: { "0" | "1" | "off" | "on" }
399 0 | off - kernel audit is disabled and can not be
400 enabled until the next reboot
401 unset - kernel audit is initialized but disabled and
402 will be fully enabled by the userspace auditd.
403 1 | on - kernel audit is initialized and partially
404 enabled, storing at most audit_backlog_limit
405 messages in RAM until it is fully enabled by the
409 audit_backlog_limit= [KNL] Set the audit queue size limit.
410 Format: <int> (must be >=0)
413 bau= [X86_UV] Enable the BAU on SGI UV. The default
414 behavior is to disable the BAU (i.e. bau=0).
415 Format: { "0" | "1" }
418 unset - Disable the BAU.
420 baycom_epp= [HW,AX25]
423 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
425 See header of drivers/net/hamradio/baycom_par.c.
427 baycom_ser_fdx= [HW,AX25]
428 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
429 Format: <io>,<irq>,<mode>[,<baud>]
430 See header of drivers/net/hamradio/baycom_ser_fdx.c.
432 baycom_ser_hdx= [HW,AX25]
433 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
434 Format: <io>,<irq>,<mode>
435 See header of drivers/net/hamradio/baycom_ser_hdx.c.
437 blkdevparts= Manual partition parsing of block device(s) for
438 embedded devices based on command line input.
439 See Documentation/block/cmdline-partition.rst
441 boot_delay= Milliseconds to delay each printk during boot.
442 Values larger than 10 seconds (10000) are changed to
447 Extended command line options can be added to an initrd
448 and this will cause the kernel to look for it.
450 See Documentation/admin-guide/bootconfig.rst
453 Disable BERT OS support on buggy BIOSes.
455 bgrt_disable [ACPI][X86]
456 Disable BGRT to avoid flickering OEM logo.
458 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
459 bttv.radio= Most important insmod options are available as
461 bttv.pll= See Documentation/admin-guide/media/bttv.rst
464 bulk_remove=off [PPC] This parameter disables the use of the pSeries
465 firmware feature for flushing multiple hpte entries
468 c101= [NET] Moxa C101 synchronous serial card
470 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
471 Sometimes CPU hardware bugs make them report the cache
472 size incorrectly. The kernel will attempt work arounds
473 to fix known problems, but for some CPUs it is not
474 possible to determine what the correct size should be.
475 This option provides an override for these situations.
478 [NET] Specifies amount of time (in seconds) that
479 the kernel should wait for a network carrier. By default
480 it waits 120 seconds.
482 ca_keys= [KEYS] This parameter identifies a specific key(s) on
483 the system trusted keyring to be used for certificate
485 format: { id:<keyid> | builtin }
487 cca= [MIPS] Override the kernel pages' cache coherency
488 algorithm. Accepted values range from 0 to 7
489 inclusive. See arch/mips/include/asm/pgtable-bits.h
490 for platform specific values (SB1, Loongson3 and
493 ccw_timeout_log [S390]
494 See Documentation/s390/common_io.rst for details.
496 cgroup_disable= [KNL] Disable a particular controller
497 Format: {name of the controller(s) to disable}
498 The effects of cgroup_disable=foo are:
499 - foo isn't auto-mounted if you mount all cgroups in
501 - foo isn't visible as an individually mountable
503 {Currently only "memory" controller deal with this and
504 cut the overhead, others just disable the usage. So
505 only cgroup_disable=memory is actually worthy}
507 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
508 Format: { { controller | "all" | "named" }
509 [,{ controller | "all" | "named" }...] }
510 Like cgroup_disable, but only applies to cgroup v1;
511 the blacklisted controllers remain available in cgroup2.
512 "all" blacklists all controllers and "named" disables
513 named mounts. Specifying both "all" and "named" disables
516 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
518 nosocket -- Disable socket memory accounting.
519 nokmem -- Disable kernel memory accounting.
521 checkreqprot [SELINUX] Set initial checkreqprot flag value.
522 Format: { "0" | "1" }
523 See security/selinux/Kconfig help text.
524 0 -- check protection applied by kernel (includes
525 any implied execute protection).
526 1 -- check protection requested by application.
527 Default value is set via a kernel config option.
528 Value can be changed at runtime via
529 /sys/fs/selinux/checkreqprot.
530 Setting checkreqprot to 1 is deprecated.
533 See Documentation/s390/common_io.rst for details.
536 Prevents the clock framework from automatically gating
537 clocks that have not been explicitly enabled by a Linux
538 device driver but are enabled in hardware at reset or
539 by the bootloader/firmware. Note that this does not
540 force such clocks to be always-on nor does it reserve
541 those clocks in any way. This parameter is useful for
542 debug and development, but should not be needed on a
543 platform with proper driver support. For more
544 information, see Documentation/driver-api/clk.rst.
546 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
548 Forces specified clocksource (if available) to be used
549 when calculating gettimeofday(). If specified
550 clocksource is not available, it defaults to PIT.
551 Format: { pit | tsc | cyclone | pmtmr }
553 clocksource= Override the default clocksource
555 Override the default clocksource and use the clocksource
556 with the name specified.
557 Some clocksource names to choose from, depending on
559 [all] jiffies (this is the base, fallback clocksource)
561 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
562 pxa_timer,timer3,32k_counter,timer0_1
563 [X86-32] pit,hpet,tsc;
564 scx200_hrt on Geode; cyclone on IBM x440
572 clocksource.arm_arch_timer.evtstrm=
575 Enable/disable the eventstream feature of the ARM
576 architected timer so that code using WFE-based polling
577 loops can be debugged more effectively on production
580 clearcpuid=BITNUM [X86]
581 Disable CPUID feature X for the kernel. See
582 arch/x86/include/asm/cpufeatures.h for the valid bit
583 numbers. Note the Linux specific bits are not necessarily
584 stable over kernel options, but the vendor specific
586 Also note that user programs calling CPUID directly
587 or using the feature without checking anything
588 will still see it. This just prevents it from
589 being used by the kernel or shown in /proc/cpuinfo.
590 Also note the kernel might malfunction if you disable
593 cma=nn[MG]@[start[MG][-end[MG]]]
595 Sets the size of kernel global memory area for
596 contiguous memory allocations and optionally the
597 placement constraint by the physical address range of
598 memory allocations. A value of 0 disables CMA
599 altogether. For more information, see
600 include/linux/dma-contiguous.h
602 cmo_free_hint= [PPC] Format: { yes | no }
603 Specify whether pages are marked as being inactive
604 when they are freed. This is used in CMO environments
605 to determine OS memory pressure for page stealing by
609 coherent_pool=nn[KMG] [ARM,KNL]
610 Sets the size of memory pool for coherent, atomic dma
611 allocations, by default set to 256K.
613 com20020= [HW,NET] ARCnet - COM20020 chipset
615 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
617 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
621 ARCnet - COM90xx chipset (memory-mapped buffers)
622 Format: <io>[,<irq>[,<memstart>]]
624 condev= [HW,S390] console device
627 console= [KNL] Output console device and options.
629 tty<n> Use the virtual console device <n>.
633 Use the specified serial port. The options are of
634 the form "bbbbpnf", where "bbbb" is the baud rate,
635 "p" is parity ("n", "o", or "e"), "n" is number of
636 bits, and "f" is flow control ("r" for RTS or
637 omit it). Default is "9600n8".
639 See Documentation/admin-guide/serial-console.rst for more
641 Documentation/networking/netconsole.rst for an
644 uart[8250],io,<addr>[,options]
645 uart[8250],mmio,<addr>[,options]
646 uart[8250],mmio16,<addr>[,options]
647 uart[8250],mmio32,<addr>[,options]
648 uart[8250],0x<addr>[,options]
649 Start an early, polled-mode console on the 8250/16550
650 UART at the specified I/O port or MMIO address,
651 switching to the matching ttyS device later.
652 MMIO inter-register address stride is either 8-bit
653 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
654 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
655 to be equivalent to 'mmio'. 'options' are specified in
656 the same format described for ttyS above; if unspecified,
657 the h/w is not re-initialized.
659 hvc<n> Use the hypervisor console device <n>. This is for
660 both Xen and PowerPC hypervisors.
662 If the device connected to the port is not a TTY but a braille
663 device, prepend "brl," before the device type, for instance
665 For now, only VisioBraille is supported.
668 [KNL] Change console messages format
670 By default we print messages on consoles in
671 "[time stamp] text\n" format (time stamp may not be
672 printed, depending on CONFIG_PRINTK_TIME or
673 `printk_time' param).
675 Switch to syslog format: "<%u>[time stamp] text\n"
676 IOW, each message will have a facility and loglevel
677 prefix. The format is similar to one used by syslog()
678 syscall, or to executing "dmesg -S --raw" or to reading
681 consoleblank= [KNL] The console blank (screen saver) timeout in
682 seconds. A value of 0 disables the blank timer.
686 [KNL] Change the default value for
687 /proc/<pid>/coredump_filter.
688 See also Documentation/filesystems/proc.rst.
690 coresight_cpu_debug.enable
693 Enable/disable the CPU sampling based debugging.
694 0: default value, disable debugging
695 1: enable debugging at boot time
697 cpuidle.off=1 [CPU_IDLE]
698 disable the cpuidle sub-system
701 [CPU_IDLE] Name of the cpuidle governor to use.
703 cpufreq.off=1 [CPU_FREQ]
704 disable the cpufreq sub-system
706 cpufreq.default_governor=
707 [CPU_FREQ] Name of the default cpufreq governor or
708 policy to use. This governor must be registered in the
709 kernel before the cpufreq driver probes.
712 [X86] Delay for N microsec between assert and de-assert
713 of APIC INIT to start processors. This delay occurs
714 on every CPU online, such as boot, and resume from suspend.
717 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
719 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
721 crashkernel=size[KMG][@offset[KMG]]
722 [KNL] Using kexec, Linux can switch to a 'crash kernel'
723 upon panic. This parameter reserves the physical
724 memory region [offset, offset + size] for that kernel
725 image. If '@offset' is omitted, then a suitable offset
726 is selected automatically.
727 [KNL, x86_64] select a region under 4G first, and
728 fall back to reserve region above 4G when '@offset'
729 hasn't been specified.
730 See Documentation/admin-guide/kdump/kdump.rst for further details.
732 crashkernel=range1:size1[,range2:size2,...][@offset]
733 [KNL] Same as above, but depends on the memory
734 in the running system. The syntax of range is
735 start-[end] where start and end are both
736 a memory unit (amount[KMG]). See also
737 Documentation/admin-guide/kdump/kdump.rst for an example.
739 crashkernel=size[KMG],high
740 [KNL, x86_64] range could be above 4G. Allow kernel
741 to allocate physical memory region from top, so could
742 be above 4G if system have more than 4G ram installed.
743 Otherwise memory region will be allocated below 4G, if
745 It will be ignored if crashkernel=X is specified.
746 crashkernel=size[KMG],low
747 [KNL, x86_64] range under 4G. When crashkernel=X,high
748 is passed, kernel could allocate physical memory region
749 above 4G, that cause second kernel crash on system
750 that require some amount of low memory, e.g. swiotlb
751 requires at least 64M+32K low memory, also enough extra
752 low memory is needed to make sure DMA buffers for 32-bit
753 devices won't run out. Kernel would try to allocate at
754 at least 256M below 4G automatically.
755 This one let user to specify own low range under 4G
756 for second kernel instead.
757 0: to disable low allocation.
758 It will be ignored when crashkernel=X,high is not used
759 or memory reserved is below 4G.
762 [KNL] Disable crypto self-tests
767 cs89x0_media= [HW,NET]
768 Format: { rj45 | aui | bnc }
771 See header of drivers/s390/block/dasd_devmap.c.
773 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
774 (one device per port)
775 Format: <port#>,<type>
776 See also Documentation/input/devices/joystick-parport.rst
778 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
780 Documentation/admin-guide/dynamic-debug-howto.rst for
781 details. Deprecated, see dyndbg.
783 debug [KNL] Enable kernel debugging (events log level).
786 [KNL] Enable printing [hashed] pointers early in the
787 boot sequence. If enabled, we use a weak hash instead
788 of siphash to hash pointers. Use this option if you are
789 seeing instances of '(___ptrval___)') and need to see a
790 value (hashed pointer) instead. Cryptographically
791 insecure, please do not use on production kernels.
794 [KNL] verbose self-tests
796 Print debugging info while doing the locking API
798 We default to 0 (no extra messages), setting it to
799 1 will print _a lot_ more information - normally
800 only useful to kernel developers.
802 debug_objects [KNL] Enable object debugging
805 [KNL] Disable object debugging
807 debug_guardpage_minorder=
808 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
809 parameter allows control of the order of pages that will
810 be intentionally kept free (and hence protected) by the
811 buddy allocator. Bigger value increase the probability
812 of catching random memory corruption, but reduce the
813 amount of memory for normal system use. The maximum
814 possible value is MAX_ORDER/2. Setting this parameter
815 to 1 or 2 should be enough to identify most random
816 memory corruption problems caused by bugs in kernel or
817 driver code when a CPU writes to (or reads from) a
818 random memory location. Note that there exists a class
819 of memory corruptions problems caused by buggy H/W or
820 F/W or by drivers badly programing DMA (basically when
821 memory is written at bus level and the CPU MMU is
822 bypassed) which are not detectable by
823 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
824 tracking down these problems.
827 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
828 enables the feature at boot time. By default, it is
829 disabled and the system will work mostly the same as a
830 kernel built without CONFIG_DEBUG_PAGEALLOC.
831 Note: to get most of debug_pagealloc error reports, it's
832 useful to also enable the page_owner functionality.
833 on: enable the feature
835 debugfs= [KNL] This parameter enables what is exposed to userspace
836 and debugfs internal clients.
837 Format: { on, no-mount, off }
838 on: All functions are enabled.
840 Filesystem is not registered but kernel clients can
841 access APIs and a crashkernel can be used to read
842 its content. There is nothing to mount.
843 off: Filesystem is not registered and clients
844 get a -EPERM as result when trying to register files
845 or directories within debugfs.
846 This is equivalent of the runtime functionality if
847 debugfs was not enabled in the kernel at all.
848 Default value is set in build-time with a kernel configuration.
850 debugpat [X86] Enable PAT debugging
852 decnet.addr= [HW,NET]
853 Format: <area>[,<node>]
854 See also Documentation/networking/decnet.rst.
857 [HW] The size of the default HugeTLB page. This is
858 the size represented by the legacy /proc/ hugepages
859 APIs. In addition, this is the default hugetlb size
860 used for shmget(), mmap() and mounting hugetlbfs
861 filesystems. If not specified, defaults to the
862 architecture's default huge page size. Huge page
863 sizes are architecture dependent. See also
864 Documentation/admin-guide/mm/hugetlbpage.rst.
867 deferred_probe_timeout=
868 [KNL] Debugging option to set a timeout in seconds for
869 deferred probe to give up waiting on dependencies to
870 probe. Only specific dependencies (subsystems or
871 drivers) that have opted in will be ignored. A timeout of 0
872 will timeout at the end of initcalls. This option will also
873 dump out devices still on the deferred probe list after
877 Format: { on | off | def_only | inf_only | always }
878 on: s390 zlib hardware support for compression on
879 level 1 and decompression (default)
880 off: No s390 zlib hardware support
881 def_only: s390 zlib hardware support for deflate
882 only (compression on level 1)
883 inf_only: s390 zlib hardware support for inflate
885 always: Same as 'on' but ignores the selected compression
886 level always using hardware support (used for debugging)
889 Set number of hash buckets for dentry cache.
891 disable_1tb_segments [PPC]
892 Disables the use of 1TB hash page table segments. This
893 causes the kernel to fall back to 256MB segments which
894 can be useful when debugging issues that require an SLB
898 Limits the number of kernel SLB entries, and flushes
899 them frequently to increase the rate of SLB faults
903 See Documentation/networking/ipv6.rst.
906 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
907 hardening is enabled for this boot. Hardened
908 usercopy checking is used to protect the kernel
909 from reading or writing beyond known memory
910 allocation boundaries as a proactive defense
911 against bounds-checking flaws in the kernel's
912 copy_to_user()/copy_from_user() interface.
913 on Perform hardened usercopy checks (default).
914 off Disable hardened usercopy checks.
917 Disable RADIX MMU mode on POWER9
920 Disable TLBIE instruction. Currently does not work
921 with KVM, with HASH MMU, or with coherent accelerators.
923 disable_cpu_apicid= [X86,APIC,SMP]
925 The number of initial APIC ID for the
926 corresponding CPU to be disabled at boot,
927 mostly used for the kdump 2nd kernel to
928 disable BSP to wake up multiple CPUs without
929 causing system reset or hang due to sending
932 perf_v4_pmi= [X86,INTEL]
934 Disable Intel PMU counter freezing feature.
935 The feature only exists starting from
936 Arch Perfmon v4 (Skylake and newer).
938 disable_ddw [PPC/PSERIES]
939 Disable Dynamic DMA Window support. Use this if
940 to workaround buggy firmware.
943 See Documentation/networking/ipv6.rst.
945 disable_mtrr_cleanup [X86]
946 The kernel tries to adjust MTRR layout from continuous
947 to discrete, to make X server driver able to add WB
948 entry later. This parameter disables that.
950 disable_mtrr_trim [X86, Intel and AMD only]
951 By default the kernel will trim any uncacheable
952 memory out of your available memory pool based on
953 MTRR settings. This parameter disables that behavior,
954 possibly causing your machine to run very slowly.
956 disable_timer_pin_1 [X86]
957 Disable PIN 1 of APIC timer
958 Can be useful to work around chipset bugs.
960 dis_ucode_ldr [X86] Disable the microcode loader.
962 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
963 this option disables the debugging code at boot.
965 dma_debug_entries=<number>
966 This option allows to tune the number of preallocated
967 entries for DMA-API debugging code. One entry is
968 required per DMA-API allocation. Use this if the
969 DMA-API debugging code disables itself because the
970 architectural default is too low.
972 dma_debug_driver=<driver_name>
973 With this option the DMA-API debugging driver
974 filter feature can be enabled at boot time. Just
975 pass the driver to filter for as the parameter.
976 The filter can be disabled or changed to another
977 driver later using sysfs.
979 driver_async_probe= [KNL]
980 List of driver names to be probed asynchronously.
981 Format: <driver_name1>,<driver_name2>...
983 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
984 Broken monitors, graphic adapters, KVMs and EDIDless
985 panels may send no or incorrect EDID data sets.
986 This parameter allows to specify an EDID data sets
987 in the /lib/firmware directory that are used instead.
988 Generic built-in EDID data sets are used, if one of
989 edid/1024x768.bin, edid/1280x1024.bin,
990 edid/1680x1050.bin, or edid/1920x1080.bin is given
991 and no file with the same name exists. Details and
992 instructions how to build your own EDID data are
993 available in Documentation/admin-guide/edid.rst. An EDID
994 data set will only be used for a particular connector,
995 if its name and a colon are prepended to the EDID
996 name. Each connector may use a unique EDID data
997 set by separating the files with a comma. An EDID
998 data set with no connector name will be used for
999 any connectors not explicitly specified.
1004 Format: {"off" | "known"}
1005 Control how the dt_cpu_ftrs device-tree binding is
1006 used for CPU feature discovery and setup (if it
1008 off: Do not use it, fall back to legacy cpu table.
1009 known: Do not pass through unknown features to guests
1010 or userspace, only those that the kernel is aware of.
1012 dump_apple_properties [X86]
1013 Dump name and content of EFI device properties on
1014 x86 Macs. Useful for driver authors to determine
1015 what data is available or for reverse-engineering.
1017 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
1018 module.dyndbg[="val"]
1019 Enable debug messages at boot time. See
1020 Documentation/admin-guide/dynamic-debug-howto.rst
1023 nopku [X86] Disable Memory Protection Keys CPU feature found
1026 module.async_probe [KNL]
1027 Enable asynchronous probe on this module.
1029 early_ioremap_debug [KNL]
1030 Enable debug messages in early_ioremap support. This
1031 is useful for tracking down temporary early mappings
1032 which are not unmapped.
1034 earlycon= [KNL] Output early console device and options.
1036 When used with no options, the early console is
1037 determined by stdout-path property in device tree's
1038 chosen node or the ACPI SPCR table if supported by
1041 cdns,<addr>[,options]
1042 Start an early, polled-mode console on a Cadence
1043 (xuartps) serial port at the specified address. Only
1044 supported option is baud rate. If baud rate is not
1045 specified, the serial port must already be setup and
1048 uart[8250],io,<addr>[,options]
1049 uart[8250],mmio,<addr>[,options]
1050 uart[8250],mmio32,<addr>[,options]
1051 uart[8250],mmio32be,<addr>[,options]
1052 uart[8250],0x<addr>[,options]
1053 Start an early, polled-mode console on the 8250/16550
1054 UART at the specified I/O port or MMIO address.
1055 MMIO inter-register address stride is either 8-bit
1056 (mmio) or 32-bit (mmio32 or mmio32be).
1057 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1058 to be equivalent to 'mmio'. 'options' are specified
1059 in the same format described for "console=ttyS<n>"; if
1060 unspecified, the h/w is not initialized.
1064 Start an early, polled-mode console on a pl011 serial
1065 port at the specified address. The pl011 serial port
1066 must already be setup and configured. Options are not
1067 yet supported. If 'mmio32' is specified, then only
1068 the driver will use only 32-bit accessors to read/write
1069 the device registers.
1072 Start an early, polled-mode console on a meson serial
1073 port at the specified address. The serial port must
1074 already be setup and configured. Options are not yet
1078 Start an early, polled-mode console on an msm serial
1079 port at the specified address. The serial port
1080 must already be setup and configured. Options are not
1083 msm_serial_dm,<addr>
1084 Start an early, polled-mode console on an msm serial
1085 dm port at the specified address. The serial port
1086 must already be setup and configured. Options are not
1090 Start an early, polled-mode console on a serial port
1091 of an Actions Semi SoC, such as S500 or S900, at the
1092 specified address. The serial port must already be
1093 setup and configured. Options are not yet supported.
1096 Start an early, polled-mode console on a serial port
1097 of an RDA Micro SoC, such as RDA8810PL, at the
1098 specified address. The serial port must already be
1099 setup and configured. Options are not yet supported.
1102 Use RISC-V SBI (Supervisor Binary Interface) for early
1105 smh Use ARM semihosting calls for early console.
1113 Use early console provided by serial driver available
1114 on Samsung SoCs, requires selecting proper type and
1115 a correct base address of the selected UART port. The
1116 serial port must already be setup and configured.
1117 Options are not yet supported.
1120 Start an early, polled-mode console on a lantiq serial
1121 (lqasc) port at the specified address. The serial port
1122 must already be setup and configured. Options are not
1127 Use early console provided by Freescale LP UART driver
1128 found on Freescale Vybrid and QorIQ LS1021A processors.
1129 A valid base address must be provided, and the serial
1130 port must already be setup and configured.
1134 Start an early, polled-mode, output-only console on the
1135 Freescale i.MX UART at the specified address. The UART
1136 must already be setup and configured.
1139 Start an early, polled-mode console on the
1140 Armada 3700 serial port at the specified
1141 address. The serial port must already be setup
1142 and configured. Options are not yet supported.
1145 Start an early, polled-mode console on a Qualcomm
1146 Generic Interface (GENI) based serial port at the
1147 specified address. The serial port must already be
1148 setup and configured. Options are not yet supported.
1151 Start an early, unaccelerated console on the EFI
1152 memory mapped framebuffer (if available). On cache
1153 coherent non-x86 systems that use system memory for
1154 the framebuffer, pass the 'ram' option so that it is
1155 mapped with the correct attributes.
1158 Use early console provided by Freescale LINFlexD UART
1159 serial driver for NXP S32V234 SoCs. A valid base
1160 address must be provided, and the serial port must
1161 already be setup and configured.
1163 earlyprintk= [X86,SH,ARM,M68k,S390]
1167 earlyprintk=serial[,ttySn[,baudrate]]
1168 earlyprintk=serial[,0x...[,baudrate]]
1169 earlyprintk=ttySn[,baudrate]
1170 earlyprintk=dbgp[debugController#]
1171 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1172 earlyprintk=xdbc[xhciController#]
1174 earlyprintk is useful when the kernel crashes before
1175 the normal console is initialized. It is not enabled by
1176 default because it has some cosmetic problems.
1178 Append ",keep" to not disable it when the real console
1181 Only one of vga, efi, serial, or usb debug port can
1184 Currently only ttyS0 and ttyS1 may be specified by
1185 name. Other I/O ports may be explicitly specified
1186 on some architectures (x86 and arm at least) by
1187 replacing ttySn with an I/O port address, like this:
1188 earlyprintk=serial,0x1008,115200
1189 You can find the port for a given device in
1190 /proc/tty/driver/serial:
1191 2: uart:ST16650V2 port:00001008 irq:18 ...
1193 Interaction with the standard serial driver is not
1196 The VGA and EFI output is eventually overwritten by
1199 The xen output can only be used by Xen PV guests.
1201 The sclp output can only be used on s390.
1203 The optional "force" to "pciserial" enables use of a
1204 PCI device even when its classcode is not of the
1207 edac_report= [HW,EDAC] Control how to report EDAC event
1208 Format: {"on" | "off" | "force"}
1209 on: enable EDAC to report H/W event. May be overridden
1210 by other higher priority error reporting module.
1211 off: disable H/W event reporting through EDAC.
1212 force: enforce the use of EDAC to report H/W event.
1215 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1218 This is designed to be used in conjunction with
1219 the boot argument: earlyprintk=vga
1221 This parameter works in place of the kgdboc parameter
1222 but can only be used if the backing tty is available
1223 very early in the boot process. For early debugging
1224 via a serial port see kgdboc_earlycon instead.
1227 Format: {"off" | "on" | "skip[mbr]"}
1230 Format: { "debug", "disable_early_pci_dma",
1231 "nochunk", "noruntime", "nosoftreserve",
1232 "novamap", "no_disable_early_pci_dma",
1234 debug: enable misc debug output.
1235 disable_early_pci_dma: disable the busmaster bit on all
1236 PCI bridges while in the EFI boot stub.
1237 nochunk: disable reading files in "chunks" in the EFI
1238 boot stub, as chunking can cause problems with some
1239 firmware implementations.
1240 noruntime : disable EFI runtime services support
1241 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose)
1242 attribute may cause the kernel to reserve the
1243 memory range for a memory mapping driver to
1244 claim. Specify efi=nosoftreserve to disable this
1245 reservation and treat the memory by its base type
1246 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM").
1247 novamap: do not call SetVirtualAddressMap().
1248 no_disable_early_pci_dma: Leave the busmaster bit set
1249 on all PCI bridges while in the EFI boot stub
1250 old_map [X86-64]: switch to the old ioremap-based EFI
1251 runtime services mapping. [Needs CONFIG_X86_UV=y]
1253 efi_no_storage_paranoia [EFI; X86]
1254 Using this parameter you can use more than 50% of
1255 your efi variable storage. Use this parameter only if
1256 you are really sure that your UEFI does sane gc and
1257 fulfills the spec otherwise your board may brick.
1259 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1260 Add arbitrary attribute to specific memory range by
1261 updating original EFI memory map.
1262 Region of memory which aa attribute is added to is
1265 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1266 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1267 attribute is added to range 0x100000000-0x180000000 and
1268 0x10a0000000-0x1120000000.
1270 If efi_fake_mem=8G@9G:0x40000 is specified, the
1271 EFI_MEMORY_SP(0x40000) attribute is added to
1272 range 0x240000000-0x43fffffff.
1274 Using this parameter you can do debugging of EFI memmap
1275 related features. For example, you can do debugging of
1276 Address Range Mirroring feature even if your box
1277 doesn't support it, or mark specific memory as
1280 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1281 that is to be dynamically loaded by Linux. If there are
1282 multiple variables with the same name but with different
1283 vendor GUIDs, all of them will be loaded. See
1284 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1287 eisa_irq_edge= [PARISC,HW]
1288 See header of drivers/parisc/eisa.c.
1291 See comment before function elanfreq_setup() in
1292 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1294 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1295 Specifies physical address of start of kernel core
1296 image elf header and optionally the size. Generally
1297 kexec loader will pass this option to capture kernel.
1298 See Documentation/admin-guide/kdump/kdump.rst for details.
1300 enable_mtrr_cleanup [X86]
1301 The kernel tries to adjust MTRR layout from continuous
1302 to discrete, to make X server driver able to add WB
1303 entry later. This parameter enables that.
1305 enable_timer_pin_1 [X86]
1306 Enable PIN 1 of APIC timer
1307 Can be useful to work around chipset bugs
1308 (in particular on some ATI chipsets).
1309 The kernel tries to set a reasonable default.
1311 enforcing [SELINUX] Set initial enforcing status.
1313 See security/selinux/Kconfig help text.
1314 0 -- permissive (log only, no denials).
1315 1 -- enforcing (deny and log).
1317 Value can be changed at runtime via
1318 /sys/fs/selinux/enforce.
1321 Disable Error Record Serialization Table (ERST)
1324 ether= [HW,NET] Ethernet cards parameters
1325 This option is obsoleted by the "netdev=" option, which
1326 has equivalent usage. See its documentation for details.
1330 Permit 'security.evm' to be updated regardless of
1331 current integrity status.
1335 fail_make_request=[KNL]
1336 General fault injection mechanism.
1337 Format: <interval>,<probability>,<space>,<times>
1338 See also Documentation/fault-injection/.
1341 See Documentation/admin-guide/blockdev/floppy.rst.
1343 force_pal_cache_flush
1344 [IA-64] Avoid check_sal_cache_flush which may hang on
1345 buggy SAL_CACHE_FLUSH implementations. Using this
1346 parameter will force ia64_sal_cache_flush to call
1347 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1350 Forcefully enable Physical Address Extension (PAE).
1351 Many Pentium M systems disable PAE but may have a
1352 functionally usable PAE implementation.
1353 Warning: use of this parameter will taint the kernel
1354 and may cause unknown problems.
1357 [FTRACE] will set and start the specified tracer
1358 as early as possible in order to facilitate early
1361 ftrace_dump_on_oops[=orig_cpu]
1362 [FTRACE] will dump the trace buffers on oops.
1363 If no parameter is passed, ftrace will dump
1364 buffers of all CPUs, but if you pass orig_cpu, it will
1365 dump only the buffer of the CPU that triggered the
1368 ftrace_filter=[function-list]
1369 [FTRACE] Limit the functions traced by the function
1370 tracer at boot up. function-list is a comma separated
1371 list of functions. This list can be changed at run
1372 time by the set_ftrace_filter file in the debugfs
1375 ftrace_notrace=[function-list]
1376 [FTRACE] Do not trace the functions specified in
1377 function-list. This list can be changed at run time
1378 by the set_ftrace_notrace file in the debugfs
1381 ftrace_graph_filter=[function-list]
1382 [FTRACE] Limit the top level callers functions traced
1383 by the function graph tracer at boot up.
1384 function-list is a comma separated list of functions
1385 that can be changed at run time by the
1386 set_graph_function file in the debugfs tracing directory.
1388 ftrace_graph_notrace=[function-list]
1389 [FTRACE] Do not trace from the functions specified in
1390 function-list. This list is a comma separated list of
1391 functions that can be changed at run time by the
1392 set_graph_notrace file in the debugfs tracing directory.
1394 ftrace_graph_max_depth=<uint>
1395 [FTRACE] Used with the function graph tracer. This is
1396 the max depth it will trace into a function. This value
1397 can be changed at run time by the max_graph_depth file
1398 in the tracefs tracing directory. default: 0 (no limit)
1400 fw_devlink= [KNL] Create device links between consumer and supplier
1401 devices by scanning the firmware to infer the
1402 consumer/supplier relationships. This feature is
1403 especially useful when drivers are loaded as modules as
1404 it ensures proper ordering of tasks like device probing
1405 (suppliers first, then consumers), supplier boot state
1406 clean up (only after all consumers have probed),
1407 suspend/resume & runtime PM (consumers first, then
1409 Format: { off | permissive | on | rpm }
1410 off -- Don't create device links from firmware info.
1411 permissive -- Create device links from firmware info
1412 but use it only for ordering boot state clean
1413 up (sync_state() calls).
1414 on -- Create device links from firmware info and use it
1415 to enforce probe and suspend/resume ordering.
1416 rpm -- Like "on", but also use to order runtime PM.
1419 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1420 support via parallel port (up to 5 devices per port)
1421 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1422 See also Documentation/input/devices/joystick-parport.rst
1426 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1430 gcov_persist= [GCOV] When non-zero (default), profiling data for
1431 kernel modules is saved and remains accessible via
1432 debugfs, even when the module is unloaded/reloaded.
1433 When zero, profiling data is discarded and associated
1434 debugfs files are removed at module unload time.
1436 goldfish [X86] Enable the goldfish android emulator platform.
1437 Don't use this when you are not running on the
1440 gpt [EFI] Forces disk with valid GPT signature but
1441 invalid Protective MBR to be treated as GPT. If the
1442 primary GPT is corrupted, it enables the backup/alternate
1443 GPT to be used instead.
1445 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1446 the "Enable 0" bit of the configuration register.
1449 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1450 the "Enable 0" bit of the configuration register.
1453 grcan.select= [HW] Select which physical interface to use.
1456 grcan.txsize= [HW] Sets the size of the tx buffer.
1457 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1459 grcan.rxsize= [HW] Sets the size of the rx buffer.
1460 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1463 gpio-mockup.gpio_mockup_ranges
1464 [HW] Sets the ranges of gpiochip of for this device.
1465 Format: <start1>,<end1>,<start2>,<end2>...
1467 hardlockup_all_cpu_backtrace=
1468 [KNL] Should the hard-lockup detector generate
1469 backtraces on all cpus.
1472 hashdist= [KNL,NUMA] Large hashes allocated during boot
1473 are distributed across NUMA nodes. Defaults on
1474 for 64-bit NUMA, off otherwise.
1475 Format: 0 | 1 (for off | on)
1477 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1479 hd= [EIDE] (E)IDE hard drive subsystem geometry
1480 Format: <cyl>,<head>,<sect>
1483 Disable Hardware Error Source Table (HEST) support;
1484 corresponding firmware-first mode error processing
1485 logic will be disabled.
1487 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1488 size of <nn>. This works even on boxes that have no
1489 highmem otherwise. This also works to reduce highmem
1490 size on bigger boxes.
1492 highres= [KNL] Enable/disable high resolution timer mode.
1493 Valid parameters: "on", "off"
1498 hpet= [X86-32,HPET] option to control HPET usage
1499 Format: { enable (default) | disable | force |
1501 disable: disable HPET and use PIT instead
1502 force: allow force enabled of undocumented chips (ICH4,
1504 verbose: show contents of HPET registers during setup
1506 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1507 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1509 hugetlb_cma= [HW] The size of a cma area used for allocation
1510 of gigantic hugepages.
1513 Reserve a cma area of given size and allocate gigantic
1514 hugepages using the cma allocator. If enabled, the
1515 boot-time allocation of gigantic hugepages is skipped.
1517 hugepages= [HW] Number of HugeTLB pages to allocate at boot.
1518 If this follows hugepagesz (below), it specifies
1519 the number of pages of hugepagesz to be allocated.
1520 If this is the first HugeTLB parameter on the command
1521 line, it specifies the number of pages to allocate for
1522 the default huge page size. See also
1523 Documentation/admin-guide/mm/hugetlbpage.rst.
1527 [HW] The size of the HugeTLB pages. This is used in
1528 conjunction with hugepages (above) to allocate huge
1529 pages of a specific size at boot. The pair
1530 hugepagesz=X hugepages=Y can be specified once for
1531 each supported huge page size. Huge page sizes are
1532 architecture dependent. See also
1533 Documentation/admin-guide/mm/hugetlbpage.rst.
1537 [KNL] Should the hung task detector generate panics.
1540 A value of 1 instructs the kernel to panic when a
1541 hung task is detected. The default value is controlled
1542 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1543 option. The value selected by this boot parameter can
1544 be changed later by the kernel.hung_task_panic sysctl.
1546 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1547 terminal devices. Valid values: 0..8
1548 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1549 If specified, z/VM IUCV HVC accepts connections
1550 from listed z/VM user IDs only.
1552 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1553 which allow the hypervisor to 'idle' the
1554 guest on lock contention.
1557 Do not unregister boot console at start. This is only
1558 useful for debugging when something happens in the window
1559 between unregistering the boot console and initializing
1562 i2c_bus= [HW] Override the default board specific I2C bus speed
1563 or register an additional I2C bus that is not
1564 registered from board initialization code.
1568 i8042.debug [HW] Toggle i8042 debug mode
1569 i8042.unmask_kbd_data
1570 [HW] Enable printing of interrupt data from the KBD port
1571 (disabled by default, and as a pre-condition
1572 requires that i8042.debug=1 be enabled)
1573 i8042.direct [HW] Put keyboard port into non-translated mode
1574 i8042.dumbkbd [HW] Pretend that controller can only read data from
1575 keyboard and cannot control its state
1576 (Don't attempt to blink the leds)
1577 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1578 i8042.nokbd [HW] Don't check/create keyboard port
1579 i8042.noloop [HW] Disable the AUX Loopback command while probing
1581 i8042.nomux [HW] Don't check presence of an active multiplexing
1583 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1585 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1586 i8042.reset [HW] Reset the controller during init, cleanup and
1587 suspend-to-ram transitions, only during s2r
1588 transitions, or never reset
1589 Format: { 1 | Y | y | 0 | N | n }
1590 1, Y, y: always reset controller
1591 0, N, n: don't ever reset controller
1592 Default: only on s2r transitions on x86; most other
1593 architectures force reset to be always executed
1594 i8042.unlock [HW] Unlock (ignore) the keylock
1595 i8042.kbdreset [HW] Reset device connected to KBD port
1599 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1600 indicates that the driver is running on unsupported
1602 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1603 does not match list of supported models.
1605 [HW] Report power status in /proc/i8k
1606 (disabled by default)
1607 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1610 i915.invert_brightness=
1611 [DRM] Invert the sense of the variable that is used to
1612 set the brightness of the panel backlight. Normally a
1613 brightness value of 0 indicates backlight switched off,
1614 and the maximum of the brightness value sets the backlight
1615 to maximum brightness. If this parameter is set to 0
1616 (default) and the machine requires it, or this parameter
1617 is set to 1, a brightness value of 0 sets the backlight
1618 to maximum brightness, and the maximum of the brightness
1619 value switches the backlight off.
1620 -1 -- never invert brightness
1621 0 -- machine default
1622 1 -- force brightness inversion
1625 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1627 ide-core.nodma= [HW] (E)IDE subsystem
1628 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1629 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1630 .cdrom .chs .ignore_cable are additional options
1631 See Documentation/ide/ide.rst.
1633 ide-generic.probe-mask= [HW] (E)IDE subsystem
1635 Probe mask for legacy ISA IDE ports. Depending on
1636 platform up to 6 ports are supported, enabled by
1637 setting corresponding bits in the mask to 1. The
1638 default value is 0x0, which has a special meaning.
1639 On systems that have PCI, it triggers scanning the
1640 PCI bus for the first and the second port, which
1641 are then probed. On systems without PCI the value
1642 of 0x0 enables probing the two first ports as if it
1645 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1646 Claim all unknown PCI IDE storage controllers.
1649 Format: idle=poll, idle=halt, idle=nomwait
1650 Poll forces a polling idle loop that can slightly
1651 improve the performance of waking up a idle CPU, but
1652 will use a lot of power and make the system run hot.
1654 idle=halt: Halt is forced to be used for CPU idle.
1655 In such case C2/C3 won't be used again.
1656 idle=nomwait: Disable mwait for CPU C-states
1658 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1659 Format: { strict | legacy | 2008 | relaxed }
1662 Choose which programs will be accepted for execution
1663 based on the IEEE 754 NaN encoding(s) supported by
1664 the FPU and the NaN encoding requested with the value
1665 of an ELF file header flag individually set by each
1666 binary. Hardware implementations are permitted to
1667 support either or both of the legacy and the 2008 NaN
1670 Available settings are as follows:
1671 strict accept binaries that request a NaN encoding
1672 supported by the FPU
1673 legacy only accept legacy-NaN binaries, if supported
1675 2008 only accept 2008-NaN binaries, if supported
1677 relaxed accept any binaries regardless of whether
1678 supported by the FPU
1680 The FPU emulator is always able to support both NaN
1681 encodings, so if no FPU hardware is present or it has
1682 been disabled with 'nofpu', then the settings of
1683 'legacy' and '2008' strap the emulator accordingly,
1684 'relaxed' straps the emulator for both legacy-NaN and
1685 2008-NaN, whereas 'strict' enables legacy-NaN only on
1686 legacy processors and both NaN encodings on MIPS32 or
1689 The setting for ABS.fmt/NEG.fmt instruction execution
1690 mode generally follows that for the NaN encoding,
1691 except where unsupported by hardware.
1693 ignore_loglevel [KNL]
1694 Ignore loglevel setting - this will print /all/
1695 kernel messages to the console. Useful for debugging.
1696 We also add it as printk module parameter, so users
1697 could change it dynamically, usually by
1698 /sys/module/printk/parameters/ignore_loglevel.
1701 Ignore RLIMIT_DATA setting for data mappings,
1702 print warning at first misuse. Can be changed via
1703 /sys/module/kernel/parameters/ignore_rlimit_data.
1705 ihash_entries= [KNL]
1706 Set number of hash buckets for inode cache.
1708 ima_appraise= [IMA] appraise integrity measurements
1709 Format: { "off" | "enforce" | "fix" | "log" }
1712 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1713 The builtin appraise policy appraises all files
1716 ima_canonical_fmt [IMA]
1717 Use the canonical format for the binary runtime
1718 measurements, instead of host native format.
1721 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1725 The list of supported hash algorithms is defined
1726 in crypto/hash_info.h.
1729 The builtin policies to load during IMA setup.
1730 Format: "tcb | appraise_tcb | secure_boot |
1733 The "tcb" policy measures all programs exec'd, files
1734 mmap'd for exec, and all files opened with the read
1735 mode bit set by either the effective uid (euid=0) or
1738 The "appraise_tcb" policy appraises the integrity of
1739 all files owned by root.
1741 The "secure_boot" policy appraises the integrity
1742 of files (eg. kexec kernel image, kernel modules,
1743 firmware, policy, etc) based on file signatures.
1745 The "fail_securely" policy forces file signature
1746 verification failure also on privileged mounted
1747 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1750 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1751 Load a policy which meets the needs of the Trusted
1752 Computing Base. This means IMA will measure all
1753 programs exec'd, files mmap'd for exec, and all files
1754 opened for read by uid=0.
1757 Select one of defined IMA measurements template formats.
1758 Formats: { "ima" | "ima-ng" | "ima-sig" }
1762 [IMA] Define a custom template format.
1763 Format: { "field1|...|fieldN" }
1765 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1766 Format: <min_file_size>
1767 Set the minimal file size for using asynchronous hash.
1768 If left unspecified, ahash usage is disabled.
1770 ahash performance varies for different data sizes on
1771 different crypto accelerators. This option can be used
1772 to achieve the best performance for a particular HW.
1774 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1776 Set hashing buffer size. Default: 4k.
1778 ahash performance varies for different chunk sizes on
1779 different crypto accelerators. This option can be used
1780 to achieve best performance for particular HW.
1784 Run specified binary instead of /sbin/init as init
1787 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1788 for working out where the kernel is dying during
1791 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1792 initcall functions. Useful for debugging built-in
1793 modules and initcalls.
1795 initrd= [BOOT] Specify the location of the initial ramdisk
1797 initrdmem= [KNL] Specify a physical address and size from which to
1798 load the initrd. If an initrd is compiled in or
1799 specified in the bootparams, it takes priority over this
1801 Format: ss[KMG],nn[KMG]
1804 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
1807 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
1809 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
1811 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
1813 init_pkru= [x86] Specify the default memory protection keys rights
1814 register contents for all processes. 0x55555554 by
1815 default (disallow access to all but pkey 0). Can
1816 override in debugfs after boot.
1818 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1821 int_pln_enable [x86] Enable power limit notification interrupt
1823 integrity_audit=[IMA]
1824 Format: { "0" | "1" }
1825 0 -- basic integrity auditing messages. (Default)
1826 1 -- additional integrity auditing messages.
1828 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1830 Enable intel iommu driver.
1832 Disable intel iommu driver.
1833 igfx_off [Default Off]
1834 By default, gfx is mapped as normal device. If a gfx
1835 device has a dedicated DMAR unit, the DMAR unit is
1836 bypassed by not enabling DMAR with this option. In
1837 this case, gfx device will use physical address for
1840 With this option iommu will not optimize to look
1841 for io virtual address below 32-bit forcing dual
1842 address cycle on pci bus for cards supporting greater
1843 than 32-bit addressing. The default is to look
1844 for translation below 32-bit and if not available
1845 then look in the higher range.
1846 strict [Default Off]
1847 With this option on every unmap_single operation will
1848 result in a hardware IOTLB flush operation as opposed
1849 to batching them for performance.
1850 sp_off [Default Off]
1851 By default, super page will be supported if Intel IOMMU
1852 has the capability. With this option, super page will
1855 By default, scalable mode will be disabled even if the
1856 hardware advertises that it has support for the scalable
1857 mode translation. With this option set, scalable mode
1858 will be used on hardware which claims to support it.
1859 tboot_noforce [Default Off]
1860 Do not force the Intel IOMMU enabled under tboot.
1861 By default, tboot will force Intel IOMMU on, which
1862 could harm performance of some high-throughput
1863 devices like 40GBit network cards, even if identity
1865 Note that using this option lowers the security
1866 provided by tboot because it makes the system
1867 vulnerable to DMA attacks.
1868 nobounce [Default off]
1869 Disable bounce buffer for untrusted devices such as
1870 the Thunderbolt devices. This will treat the untrusted
1871 devices as the trusted ones, hence might expose security
1872 risks of DMA attacks.
1874 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1875 0 disables intel_idle and fall back on acpi_idle.
1876 1 to 9 specify maximum depth of C-state.
1880 Do not enable intel_pstate as the default
1881 scaling driver for the supported processors
1883 Use intel_pstate as a scaling driver, but configure it
1884 to work with generic cpufreq governors (instead of
1885 enabling its internal governor). This mode cannot be
1886 used along with the hardware-managed P-states (HWP)
1889 Enable intel_pstate on systems that prohibit it by default
1890 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1891 instead of acpi-cpufreq may disable platform features, such
1892 as thermal controls and power capping, that rely on ACPI
1893 P-States information being indicated to OSPM and therefore
1894 should be used with caution. This option does not work with
1895 processors that aren't supported by the intel_pstate driver
1896 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1898 Do not enable hardware P state control (HWP)
1901 Only load intel_pstate on systems which support
1902 hardware P state control (HWP) if available.
1904 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1905 Description Table, specifies preferred power management
1906 profile as "Enterprise Server" or "Performance Server",
1907 then this feature is turned on by default.
1909 Allow per-logical-CPU P-State performance control limits using
1910 cpufreq sysfs interface
1912 intremap= [X86-64, Intel-IOMMU]
1913 on enable Interrupt Remapping (default)
1914 off disable Interrupt Remapping
1915 nosid disable Source ID checking
1917 BIOS x2APIC opt-out request will be ignored
1918 nopost disable Interrupt Posting
1920 iomem= Disable strict checking of access to MMIO memory
1921 strict regions from userspace.
1936 nobypass [PPC/POWERNV]
1937 Disable IOMMU bypass, using IOMMU for PCI devices.
1939 iommu.strict= [ARM64] Configure TLB invalidation behaviour
1940 Format: { "0" | "1" }
1942 Request that DMA unmap operations use deferred
1943 invalidation of hardware TLBs, for increased
1944 throughput at the cost of reduced device isolation.
1945 Will fall back to strict mode if not supported by
1946 the relevant IOMMU driver.
1947 1 - Strict mode (default).
1948 DMA unmap operations invalidate IOMMU hardware TLBs
1952 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
1953 Format: { "0" | "1" }
1954 0 - Use IOMMU translation for DMA.
1955 1 - Bypass the IOMMU for DMA.
1956 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
1958 io7= [HW] IO7 for Marvel based alpha systems
1959 See comment before marvel_specify_io7 in
1960 arch/alpha/kernel/core_marvel.c.
1962 io_delay= [X86] I/O delay method
1964 Standard port 0x80 based delay
1966 Alternate port 0xed based delay (needed on some systems)
1968 Simple two microseconds delay
1973 See Documentation/admin-guide/nfs/nfsroot.rst.
1975 ipcmni_extend [KNL] Extend the maximum number of unique System V
1976 IPC identifiers from 32,768 to 16,777,216.
1978 irqaffinity= [SMP] Set the default irq affinity mask
1979 The argument is a cpu list, as described above.
1981 irqchip.gicv2_force_probe=
1984 Force the kernel to look for the second 4kB page
1985 of a GICv2 controller even if the memory range
1986 exposed by the device tree is too small.
1988 irqchip.gicv3_nolpi=
1990 Force the kernel to ignore the availability of
1991 LPIs (and by consequence ITSs). Intended for system
1992 that use the kernel as a bootloader, and thus want
1993 to let secondary kernels in charge of setting up
1996 irqchip.gicv3_pseudo_nmi= [ARM64]
1997 Enables support for pseudo-NMIs in the kernel. This
1998 requires the kernel to be built with
1999 CONFIG_ARM64_PSEUDO_NMI.
2002 When an interrupt is not handled search all handlers
2003 for it. Intended to get systems with badly broken
2007 When an interrupt is not handled search all handlers
2008 for it. Also check all handlers each timer
2009 interrupt. Intended to get systems with badly broken
2013 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
2015 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
2016 [Deprecated - use cpusets instead]
2017 Format: [flag-list,]<cpu-list>
2019 Specify one or more CPUs to isolate from disturbances
2020 specified in the flag list (default: domain):
2023 Disable the tick when a single task runs.
2025 A residual 1Hz tick is offloaded to workqueues, which you
2026 need to affine to housekeeping through the global
2027 workqueue's affinity configured via the
2028 /sys/devices/virtual/workqueue/cpumask sysfs file, or
2029 by using the 'domain' flag described below.
2031 NOTE: by default the global workqueue runs on all CPUs,
2032 so to protect individual CPUs the 'cpumask' file has to
2033 be configured manually after bootup.
2036 Isolate from the general SMP balancing and scheduling
2037 algorithms. Note that performing domain isolation this way
2038 is irreversible: it's not possible to bring back a CPU to
2039 the domains once isolated through isolcpus. It's strongly
2040 advised to use cpusets instead to disable scheduler load
2041 balancing through the "cpuset.sched_load_balance" file.
2042 It offers a much more flexible interface where CPUs can
2043 move in and out of an isolated set anytime.
2045 You can move a process onto or off an "isolated" CPU via
2046 the CPU affinity syscalls or cpuset.
2047 <cpu number> begins at 0 and the maximum value is
2048 "number of CPUs in system - 1".
2052 Isolate from being targeted by managed interrupts
2053 which have an interrupt mask containing isolated
2054 CPUs. The affinity of managed interrupts is
2055 handled by the kernel and cannot be changed via
2056 the /proc/irq/* interfaces.
2058 This isolation is best effort and only effective
2059 if the automatically assigned interrupt mask of a
2060 device queue contains isolated and housekeeping
2061 CPUs. If housekeeping CPUs are online then such
2062 interrupts are directed to the housekeeping CPU
2063 so that IO submitted on the housekeeping CPU
2064 cannot disturb the isolated CPU.
2066 If a queue's affinity mask contains only isolated
2067 CPUs then this parameter has no effect on the
2068 interrupt routing decision, though interrupts are
2069 only delivered when tasks running on those
2070 isolated CPUs submit IO. IO submitted on
2071 housekeeping CPUs has no influence on those
2074 The format of <cpu-list> is described above.
2078 ivrs_ioapic [HW,X86_64]
2079 Provide an override to the IOAPIC-ID<->DEVICE-ID
2080 mapping provided in the IVRS ACPI table. For
2081 example, to map IOAPIC-ID decimal 10 to
2082 PCI device 00:14.0 write the parameter as:
2083 ivrs_ioapic[10]=00:14.0
2085 ivrs_hpet [HW,X86_64]
2086 Provide an override to the HPET-ID<->DEVICE-ID
2087 mapping provided in the IVRS ACPI table. For
2088 example, to map HPET-ID decimal 0 to
2089 PCI device 00:14.0 write the parameter as:
2090 ivrs_hpet[0]=00:14.0
2092 ivrs_acpihid [HW,X86_64]
2093 Provide an override to the ACPI-HID:UID<->DEVICE-ID
2094 mapping provided in the IVRS ACPI table. For
2095 example, to map UART-HID:UID AMD0020:0 to
2096 PCI device 00:14.5 write the parameter as:
2097 ivrs_acpihid[00:14.5]=AMD0020:0
2099 js= [HW,JOY] Analog joystick
2100 See Documentation/input/joydev/joystick.rst.
2103 When CONFIG_RANDOMIZE_BASE is set, this disables
2104 kernel and module base offset ASLR (Address Space
2105 Layout Randomization).
2108 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
2109 report on every invalid memory access. Without this
2110 parameter KASAN will print report only for the first
2115 kernelcore= [KNL,X86,IA-64,PPC]
2116 Format: nn[KMGTPE] | nn% | "mirror"
2117 This parameter specifies the amount of memory usable by
2118 the kernel for non-movable allocations. The requested
2119 amount is spread evenly throughout all nodes in the
2120 system as ZONE_NORMAL. The remaining memory is used for
2121 movable memory in its own zone, ZONE_MOVABLE. In the
2122 event, a node is too small to have both ZONE_NORMAL and
2123 ZONE_MOVABLE, kernelcore memory will take priority and
2124 other nodes will have a larger ZONE_MOVABLE.
2126 ZONE_MOVABLE is used for the allocation of pages that
2127 may be reclaimed or moved by the page migration
2128 subsystem. Note that allocations like PTEs-from-HighMem
2129 still use the HighMem zone if it exists, and the Normal
2130 zone if it does not.
2132 It is possible to specify the exact amount of memory in
2133 the form of "nn[KMGTPE]", a percentage of total system
2134 memory in the form of "nn%", or "mirror". If "mirror"
2135 option is specified, mirrored (reliable) memory is used
2136 for non-movable allocations and remaining memory is used
2137 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
2138 are exclusive, so you cannot specify multiple forms.
2140 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
2141 Format: <Controller#>[,poll interval]
2142 The controller # is the number of the ehci usb debug
2143 port as it is probed via PCI. The poll interval is
2144 optional and is the number seconds in between
2145 each poll cycle to the debug port in case you need
2146 the functionality for interrupting the kernel with
2147 gdb or control-c on the dbgp connection. When
2148 not using this parameter you use sysrq-g to break into
2149 the kernel debugger.
2151 kgdboc= [KGDB,HW] kgdb over consoles.
2152 Requires a tty driver that supports console polling,
2153 or a supported polling keyboard driver (non-usb).
2154 Serial only format: <serial_device>[,baud]
2155 keyboard only format: kbd
2156 keyboard and serial format: kbd,<serial_device>[,baud]
2157 Optional Kernel mode setting:
2158 kms, kbd format: kms,kbd
2159 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2161 kgdboc_earlycon= [KGDB,HW]
2162 If the boot console provides the ability to read
2163 characters and can work in polling mode, you can use
2164 this parameter to tell kgdb to use it as a backend
2165 until the normal console is registered. Intended to
2166 be used together with the kgdboc parameter which
2167 specifies the normal console to transition to.
2169 The name of the early console should be specified
2170 as the value of this parameter. Note that the name of
2171 the early console might be different than the tty
2172 name passed to kgdboc. It's OK to leave the value
2173 blank and the first boot console that implements
2174 read() will be picked.
2176 kgdbwait [KGDB] Stop kernel execution and enter the
2177 kernel debugger at the earliest opportunity.
2179 kmac= [MIPS] korina ethernet MAC address.
2180 Configure the RouterBoard 532 series on-chip
2181 Ethernet adapter MAC address.
2183 kmemleak= [KNL] Boot-time kmemleak enable/disable
2184 Valid arguments: on, off
2186 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2189 kprobe_event=[probe-list]
2190 [FTRACE] Add kprobe events and enable at boot time.
2191 The probe-list is a semicolon delimited list of probe
2192 definitions. Each definition is same as kprobe_events
2193 interface, but the parameters are comma delimited.
2194 For example, to add a kprobe event on vfs_read with
2195 arg1 and arg2, add to the command line;
2197 kprobe_event=p,vfs_read,$arg1,$arg2
2199 See also Documentation/trace/kprobetrace.rst "Kernel
2200 Boot Parameter" section.
2202 kpti= [ARM64] Control page table isolation of user
2203 and kernel address spaces.
2204 Default: enabled on cores which need mitigation.
2208 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2209 Default is 0 (don't ignore, but inject #GP)
2211 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2212 Default is false (don't support).
2214 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
2219 [KVM] Controls the software workaround for the
2220 X86_BUG_ITLB_MULTIHIT bug.
2221 force : Always deploy workaround.
2222 off : Never deploy workaround.
2223 auto : Deploy workaround based on the presence of
2224 X86_BUG_ITLB_MULTIHIT.
2228 If the software workaround is enabled for the host,
2229 guests do need not to enable it for nested guests.
2231 kvm.nx_huge_pages_recovery_ratio=
2232 [KVM] Controls how many 4KiB pages are periodically zapped
2233 back to huge pages. 0 disables the recovery, otherwise if
2234 the value is N KVM will zap 1/Nth of the 4KiB pages every
2235 minute. The default is 60.
2237 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2238 Default is 1 (enabled)
2240 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2242 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2244 kvm-arm.vgic_v3_group0_trap=
2245 [KVM,ARM] Trap guest accesses to GICv3 group-0
2248 kvm-arm.vgic_v3_group1_trap=
2249 [KVM,ARM] Trap guest accesses to GICv3 group-1
2252 kvm-arm.vgic_v3_common_trap=
2253 [KVM,ARM] Trap guest accesses to GICv3 common
2256 kvm-arm.vgic_v4_enable=
2257 [KVM,ARM] Allow use of GICv4 for direct injection of
2260 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2261 (virtualized MMU) support on capable Intel chips.
2262 Default is 1 (enabled)
2264 kvm-intel.emulate_invalid_guest_state=
2265 [KVM,Intel] Enable emulation of invalid guest states
2266 Default is 0 (disabled)
2268 kvm-intel.flexpriority=
2269 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2270 Default is 1 (enabled)
2273 [KVM,Intel] Enable VMX nesting (nVMX).
2274 Default is 0 (disabled)
2276 kvm-intel.unrestricted_guest=
2277 [KVM,Intel] Disable unrestricted guest feature
2278 (virtualized real and unpaged mode) on capable
2279 Intel chips. Default is 1 (enabled)
2281 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2284 Valid arguments: never, cond, always
2286 always: L1D cache flush on every VMENTER.
2287 cond: Flush L1D on VMENTER only when the code between
2288 VMEXIT and VMENTER can leak host memory.
2289 never: Disables the mitigation
2291 Default is cond (do L1 cache flush in specific instances)
2293 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2294 feature (tagged TLBs) on capable Intel chips.
2295 Default is 1 (enabled)
2297 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2300 The kernel PTE inversion protection is unconditionally
2301 enabled and cannot be disabled.
2304 Provides all available mitigations for the
2305 L1TF vulnerability. Disables SMT and
2306 enables all mitigations in the
2307 hypervisors, i.e. unconditional L1D flush.
2309 SMT control and L1D flush control via the
2310 sysfs interface is still possible after
2311 boot. Hypervisors will issue a warning
2312 when the first VM is started in a
2313 potentially insecure configuration,
2314 i.e. SMT enabled or L1D flush disabled.
2317 Same as 'full', but disables SMT and L1D
2318 flush runtime control. Implies the
2319 'nosmt=force' command line option.
2320 (i.e. sysfs control of SMT is disabled.)
2323 Leaves SMT enabled and enables the default
2324 hypervisor mitigation, i.e. conditional
2327 SMT control and L1D flush control via the
2328 sysfs interface is still possible after
2329 boot. Hypervisors will issue a warning
2330 when the first VM is started in a
2331 potentially insecure configuration,
2332 i.e. SMT enabled or L1D flush disabled.
2336 Disables SMT and enables the default
2337 hypervisor mitigation.
2339 SMT control and L1D flush control via the
2340 sysfs interface is still possible after
2341 boot. Hypervisors will issue a warning
2342 when the first VM is started in a
2343 potentially insecure configuration,
2344 i.e. SMT enabled or L1D flush disabled.
2347 Same as 'flush', but hypervisors will not
2348 warn when a VM is started in a potentially
2349 insecure configuration.
2352 Disables hypervisor mitigations and doesn't
2354 It also drops the swap size and available
2355 RAM limit restriction on both hypervisor and
2360 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2366 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2369 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2370 value for LAPIC timer one-shot implementation. Default
2371 back to the programmable timer unit in the LAPIC.
2373 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2376 libata.dma= [LIBATA] DMA control
2377 libata.dma=0 Disable all PATA and SATA DMA
2378 libata.dma=1 PATA and SATA Disk DMA only
2379 libata.dma=2 ATAPI (CDROM) DMA only
2380 libata.dma=4 Compact Flash DMA only
2381 Combinations also work, so libata.dma=3 enables DMA
2382 for disks and CDROMs, but not CFs.
2384 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2385 libata.ignore_hpa=0 keep BIOS limits (default)
2386 libata.ignore_hpa=1 ignore limits, using full disk
2388 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2392 libata.force= [LIBATA] Force configurations. The format is comma
2393 separated list of "[ID:]VAL" where ID is
2394 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2395 matching port, link or device. Basically, it matches
2396 the ATA ID string printed on console by libata. If
2397 the whole ID part is omitted, the last PORT and DEVICE
2398 values are used. If ID hasn't been specified yet, the
2399 configuration applies to all ports, links and devices.
2401 If only DEVICE is omitted, the parameter applies to
2402 the port and all links and devices behind it. DEVICE
2403 number of 0 either selects the first device or the
2404 first fan-out link behind PMP device. It does not
2405 select the host link. DEVICE number of 15 selects the
2406 host link and device attached to it.
2408 The VAL specifies the configuration to force. As long
2409 as there's no ambiguity shortcut notation is allowed.
2410 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2411 The following configurations can be forced.
2413 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2414 Any ID with matching PORT is used.
2416 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2418 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2419 udma[/][16,25,33,44,66,100,133] notation is also
2422 * [no]ncq: Turn on or off NCQ.
2424 * [no]ncqtrim: Turn off queued DSM TRIM.
2426 * nohrst, nosrst, norst: suppress hard, soft
2429 * rstonce: only attempt one reset during
2430 hot-unplug link recovery
2432 * dump_id: dump IDENTIFY data.
2434 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2436 * disable: Disable this device.
2438 If there are multiple matching configurations changing
2439 the same attribute, the last one is used.
2441 memblock=debug [KNL] Enable memblock debug messages.
2443 load_ramdisk= [RAM] List of ramdisks to load from floppy
2444 See Documentation/admin-guide/blockdev/ramdisk.rst.
2446 lockd.nlm_grace_period=P [NFS] Assign grace period.
2449 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2452 lockd.nlm_timeout=T [NFS] Assign timeout value.
2455 lockd.nlm_udpport=M [NFS] Assign UDP port.
2458 lockdown= [SECURITY]
2459 { integrity | confidentiality }
2460 Enable the kernel lockdown feature. If set to
2461 integrity, kernel features that allow userland to
2462 modify the running kernel are disabled. If set to
2463 confidentiality, kernel features that allow userland
2464 to extract confidential information from the kernel
2467 locktorture.nreaders_stress= [KNL]
2468 Set the number of locking read-acquisition kthreads.
2469 Defaults to being automatically set based on the
2470 number of online CPUs.
2472 locktorture.nwriters_stress= [KNL]
2473 Set the number of locking write-acquisition kthreads.
2475 locktorture.onoff_holdoff= [KNL]
2476 Set time (s) after boot for CPU-hotplug testing.
2478 locktorture.onoff_interval= [KNL]
2479 Set time (s) between CPU-hotplug operations, or
2480 zero to disable CPU-hotplug testing.
2482 locktorture.shuffle_interval= [KNL]
2483 Set task-shuffle interval (jiffies). Shuffling
2484 tasks allows some CPUs to go into dyntick-idle
2485 mode during the locktorture test.
2487 locktorture.shutdown_secs= [KNL]
2488 Set time (s) after boot system shutdown. This
2489 is useful for hands-off automated testing.
2491 locktorture.stat_interval= [KNL]
2492 Time (s) between statistics printk()s.
2494 locktorture.stutter= [KNL]
2495 Time (s) to stutter testing, for example,
2496 specifying five seconds causes the test to run for
2497 five seconds, wait for five seconds, and so on.
2498 This tests the locking primitive's ability to
2499 transition abruptly to and from idle.
2501 locktorture.torture_type= [KNL]
2502 Specify the locking implementation to test.
2504 locktorture.verbose= [KNL]
2505 Enable additional printk() statements.
2507 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2510 loglevel= All Kernel Messages with a loglevel smaller than the
2511 console loglevel will be printed to the console. It can
2512 also be changed with klogd or other programs. The
2513 loglevels are defined as follows:
2515 0 (KERN_EMERG) system is unusable
2516 1 (KERN_ALERT) action must be taken immediately
2517 2 (KERN_CRIT) critical conditions
2518 3 (KERN_ERR) error conditions
2519 4 (KERN_WARNING) warning conditions
2520 5 (KERN_NOTICE) normal but significant condition
2521 6 (KERN_INFO) informational
2522 7 (KERN_DEBUG) debug-level messages
2524 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2525 in bytes. n must be a power of two and greater
2526 than the minimal size. The minimal size is defined
2527 by LOG_BUF_SHIFT kernel config parameter. There is
2528 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2529 that allows to increase the default size depending on
2530 the number of CPUs. See init/Kconfig for more details.
2532 logo.nologo [FB] Disables display of the built-in Linux logo.
2533 This may be used to provide more screen space for
2534 kernel log messages and is useful when debugging
2535 kernel boot problems.
2537 lp=0 [LP] Specify parallel ports to use, e.g,
2538 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2539 lp=reset first parallel port). 'lp=0' disables the
2540 lp=auto printer driver. 'lp=reset' (which can be
2541 specified in addition to the ports) causes
2542 attached printers to be reset. Using
2543 lp=port1,port2,... specifies the parallel ports
2544 to associate lp devices with, starting with
2545 lp0. A port specification may be 'none' to skip
2546 that lp device, or a parport name such as
2547 'parport0'. Specifying 'lp=auto' instead of a
2548 port specification list means that device IDs
2549 from each port should be examined, to see if
2550 an IEEE 1284-compliant printer is attached; if
2551 so, the driver will manage that printer.
2552 See also header of drivers/char/lp.c.
2555 Sets loops_per_jiffy to given constant, thus avoiding
2556 time-consuming boot-time autodetection (up to 250 ms per
2557 CPU). 0 enables autodetection (default). To determine
2558 the correct value for your kernel, boot with normal
2559 autodetection and see what value is printed. Note that
2560 on SMP systems the preset will be applied to all CPUs,
2561 which is likely to cause problems if your CPUs need
2562 significantly divergent settings. An incorrect value
2563 will cause delays in the kernel to be wrong, leading to
2564 unpredictable I/O errors and other breakage. Although
2565 unlikely, in the extreme case this might damage your
2569 Format: <io>,<irq>,<dma>
2571 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2574 [SECURITY] Choose order of LSM initialization. This
2575 overrides CONFIG_LSM, and the "security=" parameter.
2577 machvec= [IA-64] Force the use of a particular machine-vector
2578 (machvec) in a generic kernel.
2579 Example: machvec=hpzx1
2581 machtype= [Loongson] Share the same kernel image file between different
2583 Example: machtype=lemote-yeeloong-2f-7inch
2585 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2586 than or equal to this physical address is ignored.
2588 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2589 will bring up during bootup. maxcpus=n : n >= 0 limits
2590 the kernel to bring up 'n' processors. Surely after
2591 bootup you can bring up the other plugged cpu by executing
2592 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2593 only takes effect during system bootup.
2594 While n=0 is a special case, it is equivalent to "nosmp",
2595 which also disables the IO APIC.
2597 max_loop= [LOOP] The number of loop block devices that get
2598 (loop.max_loop) unconditionally pre-created at init time. The default
2599 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2600 of statically allocating a predefined number, loop
2601 devices can be requested on-demand with the
2602 /dev/loop-control interface.
2604 mce [X86-32] Machine Check Exception
2606 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst
2608 md= [HW] RAID subsystems devices and level
2609 See Documentation/admin-guide/md.rst.
2612 Format: <first>,<last>
2613 Specifies range of consoles to be captured by the MDA.
2616 Control mitigation for the Micro-architectural Data
2617 Sampling (MDS) vulnerability.
2619 Certain CPUs are vulnerable to an exploit against CPU
2620 internal buffers which can forward information to a
2621 disclosure gadget under certain conditions.
2623 In vulnerable processors, the speculatively
2624 forwarded data can be used in a cache side channel
2625 attack, to access data to which the attacker does
2626 not have direct access.
2628 This parameter controls the MDS mitigation. The
2631 full - Enable MDS mitigation on vulnerable CPUs
2632 full,nosmt - Enable MDS mitigation and disable
2633 SMT on vulnerable CPUs
2634 off - Unconditionally disable MDS mitigation
2636 On TAA-affected machines, mds=off can be prevented by
2637 an active TAA mitigation as both vulnerabilities are
2638 mitigated with the same mechanism so in order to disable
2639 this mitigation, you need to specify tsx_async_abort=off
2642 Not specifying this option is equivalent to
2645 For details see: Documentation/admin-guide/hw-vuln/mds.rst
2647 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2648 Amount of memory to be used in cases as follows:
2651 2 when the kernel is not able to see the whole system memory;
2652 3 memory that lies after 'mem=' boundary is excluded from
2653 the hypervisor, then assigned to KVM guests.
2655 [X86] Work as limiting max address. Use together
2656 with memmap= to avoid physical address space collisions.
2657 Without memmap= PCI devices could be placed at addresses
2658 belonging to unused RAM.
2660 Note that this only takes effects during boot time since
2661 in above case 3, memory may need be hot added after boot
2662 if system memory of hypervisor is not sufficient.
2664 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2668 [KNL,SH] Allow user to override the default size for
2669 per-device physically contiguous DMA buffers.
2671 memhp_default_state=online/offline
2672 [KNL] Set the initial state for the memory hotplug
2673 onlining policy. If not specified, the default value is
2674 set according to the
2675 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2677 See Documentation/admin-guide/mm/memory-hotplug.rst.
2679 memmap=exactmap [KNL,X86] Enable setting of an exact
2680 E820 memory map, as specified by the user.
2681 Such memmap=exactmap lines can be constructed based on
2682 BIOS output or other requirements. See the memmap=nn@ss
2685 memmap=nn[KMG]@ss[KMG]
2686 [KNL] Force usage of a specific region of memory.
2687 Region of memory to be used is from ss to ss+nn.
2688 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2689 which limits max address to nn[KMG].
2690 Multiple different regions can be specified,
2693 memmap=100M@2G,100M#3G,1G!1024G
2695 memmap=nn[KMG]#ss[KMG]
2696 [KNL,ACPI] Mark specific memory as ACPI data.
2697 Region of memory to be marked is from ss to ss+nn.
2699 memmap=nn[KMG]$ss[KMG]
2700 [KNL,ACPI] Mark specific memory as reserved.
2701 Region of memory to be reserved is from ss to ss+nn.
2702 Example: Exclude memory from 0x18690000-0x1869ffff
2703 memmap=64K$0x18690000
2705 memmap=0x10000$0x18690000
2706 Some bootloaders may need an escape character before '$',
2707 like Grub2, otherwise '$' and the following number
2710 memmap=nn[KMG]!ss[KMG]
2711 [KNL,X86] Mark specific memory as protected.
2712 Region of memory to be used, from ss to ss+nn.
2713 The memory region may be marked as e820 type 12 (0xc)
2714 and is NVDIMM or ADR memory.
2716 memmap=<size>%<offset>-<oldtype>+<newtype>
2717 [KNL,ACPI] Convert memory within the specified region
2718 from <oldtype> to <newtype>. If "-<oldtype>" is left
2719 out, the whole region will be marked as <newtype>,
2720 even if previously unavailable. If "+<newtype>" is left
2721 out, matching memory will be removed. Types are
2722 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2723 3 = ACPI, 12 = PRAM.
2725 memory_corruption_check=0/1 [X86]
2726 Some BIOSes seem to corrupt the first 64k of
2727 memory when doing things like suspend/resume.
2728 Setting this option will scan the memory
2729 looking for corruption. Enabling this will
2730 both detect corruption and prevent the kernel
2731 from using the memory being corrupted.
2732 However, its intended as a diagnostic tool; if
2733 repeatable BIOS-originated corruption always
2734 affects the same memory, you can use memmap=
2735 to prevent the kernel from using that memory.
2737 memory_corruption_check_size=size [X86]
2738 By default it checks for corruption in the low
2739 64k, making this memory unavailable for normal
2740 use. Use this parameter to scan for
2741 corruption in more or less memory.
2743 memory_corruption_check_period=seconds [X86]
2744 By default it checks for corruption every 60
2745 seconds. Use this parameter to check at some
2746 other rate. 0 disables periodic checking.
2748 memtest= [KNL,X86,ARM,PPC] Enable memtest
2750 default : 0 <disable>
2751 Specifies the number of memtest passes to be
2752 performed. Each pass selects another test
2753 pattern from a given set of patterns. Memtest
2754 fills the memory with this pattern, validates
2755 memory contents and reserves bad memory
2756 regions that are detected.
2758 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2759 Valid arguments: on, off
2760 Default (depends on kernel configuration option):
2761 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2762 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2763 mem_encrypt=on: Activate SME
2764 mem_encrypt=off: Do not activate SME
2766 Refer to Documentation/virt/kvm/amd-memory-encryption.rst
2767 for details on when memory encryption can be activated.
2769 mem_sleep_default= [SUSPEND] Default system suspend mode:
2770 s2idle - Suspend-To-Idle
2771 shallow - Power-On Suspend or equivalent (if supported)
2772 deep - Suspend-To-RAM or equivalent (if supported)
2773 See Documentation/admin-guide/pm/sleep-states.rst.
2775 meye.*= [HW] Set MotionEye Camera parameters
2776 See Documentation/admin-guide/media/meye.rst.
2778 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2779 Multi-Function General Purpose Timers on AMD Geode
2782 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2783 the BIOS has incorrectly applied a workaround. TinyBIOS
2784 version 0.98 is known to be affected, 0.99 fixes the
2785 problem by letting the user disable the workaround.
2789 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2790 physical address is ignored.
2792 mini2440= [ARM,HW,KNL]
2793 Format:[0..2][b][c][t]
2795 MINI2440 configuration specification:
2796 0 - The attached screen is the 3.5" TFT
2797 1 - The attached screen is the 7" TFT
2798 2 - The VGA Shield is attached (1024x768)
2799 Leaving out the screen size parameter will not load
2800 the TFT driver, and the framebuffer will be left
2802 b - Enable backlight. The TFT backlight pin will be
2803 linked to the kernel VESA blanking code and a GPIO
2804 LED. This parameter is not necessary when using the
2806 c - Enable the s3c camera interface.
2807 t - Reserved for enabling touchscreen support. The
2808 touchscreen support is not enabled in the mainstream
2809 kernel as of 2.6.30, a preliminary port can be found
2810 in the "bleeding edge" mini2440 support kernel at
2811 https://repo.or.cz/w/linux-2.6/mini2440.git
2814 [X86,PPC,S390,ARM64] Control optional mitigations for
2815 CPU vulnerabilities. This is a set of curated,
2816 arch-independent options, each of which is an
2817 aggregation of existing arch-specific options.
2820 Disable all optional CPU mitigations. This
2821 improves system performance, but it may also
2822 expose users to several CPU vulnerabilities.
2823 Equivalent to: nopti [X86,PPC]
2825 nospectre_v1 [X86,PPC]
2827 nospectre_v2 [X86,PPC,S390,ARM64]
2828 spectre_v2_user=off [X86]
2829 spec_store_bypass_disable=off [X86,PPC]
2830 ssbd=force-off [ARM64]
2833 tsx_async_abort=off [X86]
2834 kvm.nx_huge_pages=off [X86]
2837 This does not have any effect on
2838 kvm.nx_huge_pages when
2839 kvm.nx_huge_pages=force.
2842 Mitigate all CPU vulnerabilities, but leave SMT
2843 enabled, even if it's vulnerable. This is for
2844 users who don't want to be surprised by SMT
2845 getting disabled across kernel upgrades, or who
2846 have other ways of avoiding SMT-based attacks.
2847 Equivalent to: (default behavior)
2850 Mitigate all CPU vulnerabilities, disabling SMT
2851 if needed. This is for users who always want to
2852 be fully mitigated, even if it means losing SMT.
2853 Equivalent to: l1tf=flush,nosmt [X86]
2854 mds=full,nosmt [X86]
2855 tsx_async_abort=full,nosmt [X86]
2858 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2859 parameter allows control of the logging verbosity for
2860 the additional memory initialisation checks. A value
2861 of 0 disables mminit logging and a level of 4 will
2862 log everything. Information is printed at KERN_DEBUG
2863 so loglevel=8 may also need to be specified.
2866 [KNL] When CONFIG_MODULE_SIG is set, this means that
2867 modules without (valid) signatures will fail to load.
2868 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2869 is always true, so this option does nothing.
2871 module_blacklist= [KNL] Do not load a comma-separated list of
2872 modules. Useful for debugging problem modules.
2875 [MOUSE] Maximum time between finger touching and
2876 leaving touchpad surface for touch to be considered
2877 a tap and be reported as a left button click (for
2878 touchpads working in absolute mode only).
2880 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2881 reporting absolute coordinates, such as tablets
2882 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2883 reporting absolute coordinates, such as tablets
2885 movablecore= [KNL,X86,IA-64,PPC]
2886 Format: nn[KMGTPE] | nn%
2887 This parameter is the complement to kernelcore=, it
2888 specifies the amount of memory used for migratable
2889 allocations. If both kernelcore and movablecore is
2890 specified, then kernelcore will be at *least* the
2891 specified value but may be more. If movablecore on its
2892 own is specified, the administrator must be careful
2893 that the amount of memory usable for all allocations
2896 movable_node [KNL] Boot-time switch to make hotplugable memory
2897 NUMA nodes to be movable. This means that the memory
2898 of such nodes will be usable only for movable
2899 allocations which rules out almost all kernel
2900 allocations. Use with caution!
2902 MTD_Partition= [MTD]
2903 Format: <name>,<region-number>,<size>,<offset>
2905 MTD_Region= [MTD] Format:
2906 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2909 See drivers/mtd/parsers/cmdlinepart.c
2911 multitce=off [PPC] This parameter disables the use of the pSeries
2912 firmware feature for updating multiple TCE entries
2915 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2917 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2919 boundary - index of last SLC block on Flex-OneNAND.
2920 The remaining blocks are configured as MLC blocks.
2921 lock - Configure if Flex-OneNAND boundary should be locked.
2922 Once locked, the boundary cannot be changed.
2923 1 indicates lock status, 0 indicates unlock status.
2926 ARM/S3C2412 JIVE boot control
2928 See arch/arm/mach-s3c2412/mach-jive.c
2930 mtouchusb.raw_coordinates=
2931 [HW] Make the MicroTouch USB driver use raw coordinates
2932 ('y', default) or cooked coordinates ('n')
2934 mtrr_chunk_size=nn[KMG] [X86]
2935 used for mtrr cleanup. It is largest continuous chunk
2936 that could hold holes aka. UC entries.
2938 mtrr_gran_size=nn[KMG] [X86]
2939 Used for mtrr cleanup. It is granularity of mtrr block.
2941 Large value could prevent small alignment from
2944 mtrr_spare_reg_nr=n [X86]
2946 Range: 0,7 : spare reg number
2948 Used for mtrr cleanup. It is spare mtrr entries number.
2949 Set to 2 or more if your graphical card needs more.
2951 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2953 netdev= [NET] Network devices parameters
2954 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2955 Note that mem_start is often overloaded to mean
2956 something different and driver-specific.
2957 This usage is only documented in each driver source
2961 [NETFILTER] Enable connection tracking flow accounting
2962 0 to disable accounting
2963 1 to enable accounting
2966 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2967 See Documentation/admin-guide/nfs/nfsroot.rst.
2969 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2970 See Documentation/admin-guide/nfs/nfsroot.rst.
2972 nfsrootdebug [NFS] enable nfsroot debugging messages.
2973 See Documentation/admin-guide/nfs/nfsroot.rst.
2975 nfs.callback_nr_threads=
2976 [NFSv4] set the total number of threads that the
2977 NFS client will assign to service NFSv4 callback
2980 nfs.callback_tcpport=
2981 [NFS] set the TCP port on which the NFSv4 callback
2982 channel should listen.
2985 [NFS] sets the pathname to the program which is used
2986 to update the NFS client cache entries.
2988 nfs.cache_getent_timeout=
2989 [NFS] sets the timeout after which an attempt to
2990 update a cache entry is deemed to have failed.
2992 nfs.idmap_cache_timeout=
2993 [NFS] set the maximum lifetime for idmapper cache
2997 [NFS] enable 64-bit inode numbers.
2998 If zero, the NFS client will fake up a 32-bit inode
2999 number for the readdir() and stat() syscalls instead
3000 of returning the full 64-bit number.
3001 The default is to return 64-bit inode numbers.
3003 nfs.max_session_cb_slots=
3004 [NFSv4.1] Sets the maximum number of session
3005 slots the client will assign to the callback
3006 channel. This determines the maximum number of
3007 callbacks the client will process in parallel for
3008 a particular server.
3010 nfs.max_session_slots=
3011 [NFSv4.1] Sets the maximum number of session slots
3012 the client will attempt to negotiate with the server.
3013 This limits the number of simultaneous RPC requests
3014 that the client can send to the NFSv4.1 server.
3015 Note that there is little point in setting this
3016 value higher than the max_tcp_slot_table_limit.
3018 nfs.nfs4_disable_idmapping=
3019 [NFSv4] When set to the default of '1', this option
3020 ensures that both the RPC level authentication
3021 scheme and the NFS level operations agree to use
3022 numeric uids/gids if the mount is using the
3023 'sec=sys' security flavour. In effect it is
3024 disabling idmapping, which can make migration from
3025 legacy NFSv2/v3 systems to NFSv4 easier.
3026 Servers that do not support this mode of operation
3027 will be autodetected by the client, and it will fall
3028 back to using the idmapper.
3029 To turn off this behaviour, set the value to '0'.
3031 [NFS4] Specify an additional fixed unique ident-
3032 ification string that NFSv4 clients can insert into
3033 their nfs_client_id4 string. This is typically a
3034 UUID that is generated at system install time.
3036 nfs.send_implementation_id =
3037 [NFSv4.1] Send client implementation identification
3038 information in exchange_id requests.
3039 If zero, no implementation identification information
3041 The default is to send the implementation identification
3044 nfs.recover_lost_locks =
3045 [NFSv4] Attempt to recover locks that were lost due
3046 to a lease timeout on the server. Please note that
3047 doing this risks data corruption, since there are
3048 no guarantees that the file will remain unchanged
3049 after the locks are lost.
3050 If you want to enable the kernel legacy behaviour of
3051 attempting to recover these locks, then set this
3053 The default parameter value of '0' causes the kernel
3054 not to attempt recovery of lost locks.
3056 nfs4.layoutstats_timer =
3057 [NFSv4.2] Change the rate at which the kernel sends
3058 layoutstats to the pNFS metadata server.
3060 Setting this to value to 0 causes the kernel to use
3061 whatever value is the default set by the layout
3062 driver. A non-zero value sets the minimum interval
3063 in seconds between layoutstats transmissions.
3065 nfsd.nfs4_disable_idmapping=
3066 [NFSv4] When set to the default of '1', the NFSv4
3067 server will return only numeric uids and gids to
3068 clients using auth_sys, and will accept numeric uids
3069 and gids from such clients. This is intended to ease
3070 migration from NFSv2/v3.
3072 nmi_debug= [KNL,SH] Specify one or more actions to take
3073 when a NMI is triggered.
3074 Format: [state][,regs][,debounce][,die]
3076 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
3077 Format: [panic,][nopanic,][num]
3079 0 - turn hardlockup detector in nmi_watchdog off
3080 1 - turn hardlockup detector in nmi_watchdog on
3081 When panic is specified, panic when an NMI watchdog
3082 timeout occurs (or 'nopanic' to not panic on an NMI
3083 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
3084 To disable both hard and soft lockup detectors,
3085 please see 'nowatchdog'.
3086 This is useful when you use a panic=... timeout and
3087 need the box quickly up again.
3089 These settings can be accessed at runtime via
3090 the nmi_watchdog and hardlockup_panic sysctls.
3092 netpoll.carrier_timeout=
3093 [NET] Specifies amount of time (in seconds) that
3094 netpoll should wait for a carrier. By default netpoll
3097 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
3098 emulation library even if a 387 maths coprocessor
3101 no5lvl [X86-64] Disable 5-level paging mode. Forces
3102 kernel to use 4-level paging instead.
3104 nofsgsbase [X86] Disables FSGSBASE instructions.
3107 [HW] Never suspend the console
3108 Disable suspending of consoles during suspend and
3109 hibernate operations. Once disabled, debugging
3110 messages can reach various consoles while the rest
3111 of the system is being put to sleep (ie, while
3112 debugging driver suspend/resume hooks). This may
3113 not work reliably with all consoles, but is known
3114 to work with serial and VGA consoles.
3115 To facilitate more flexible debugging, we also add
3116 console_suspend, a printk module parameter to control
3117 it. Users could use console_suspend (usually
3118 /sys/module/printk/parameters/console_suspend) to
3119 turn on/off it dynamically.
3121 novmcoredd [KNL,KDUMP]
3122 Disable device dump. Device dump allows drivers to
3123 append dump data to vmcore so you can collect driver
3124 specified debug info. Drivers can append the data
3125 without any limit and this data is stored in memory,
3126 so this may cause significant memory stress. Disabling
3127 device dump can help save memory but the driver debug
3128 data will be no longer available. This parameter
3129 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
3132 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
3133 caches in the slab allocator. Saves per-node memory,
3134 but will impact performance.
3138 noaltinstr [S390] Disables alternative instructions patching
3139 (CPU alternatives feature).
3141 noapic [SMP,APIC] Tells the kernel to not make use of any
3142 IOAPICs that may be present in the system.
3144 noautogroup Disable scheduler automatic task group creation.
3146 nobats [PPC] Do not use BATs for mapping kernel lowmem
3147 on "Classic" PPC cores.
3151 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
3153 nodelayacct [KNL] Disable per-task delay accounting
3155 nodsp [SH] Disable hardware DSP at boot time.
3157 noefi Disable EFI runtime services support.
3162 On X86-32 available only on PAE configured kernels.
3163 noexec=on: enable non-executable mappings (default)
3164 noexec=off: disable non-executable mappings
3167 Disable SMAP (Supervisor Mode Access Prevention)
3168 even if it is supported by processor.
3171 Disable SMEP (Supervisor Mode Execution Prevention)
3172 even if it is supported by processor.
3175 This affects only 32-bit executables.
3176 noexec32=on: enable non-executable mappings (default)
3177 read doesn't imply executable mappings
3178 noexec32=off: disable non-executable mappings
3179 read implies executable mappings
3181 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3183 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3184 register save and restore. The kernel will only save
3185 legacy floating-point registers on task switch.
3187 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings.
3189 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
3190 Equivalent to smt=1.
3192 [KNL,x86] Disable symmetric multithreading (SMT).
3193 nosmt=force: Force disable SMT, cannot be undone
3194 via the sysfs control file.
3196 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3197 (bounds check bypass). With this option data leaks are
3198 possible in the system.
3200 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for
3201 the Spectre variant 2 (indirect branch prediction)
3202 vulnerability. System may allow data leaks with this
3205 nospec_store_bypass_disable
3206 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3208 noxsave [BUGS=X86] Disables x86 extended register state save
3209 and restore using xsave. The kernel will fallback to
3210 enabling legacy floating-point and sse state.
3212 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3213 register states. The kernel will fall back to use
3214 xsave to save the states. By using this parameter,
3215 performance of saving the states is degraded because
3216 xsave doesn't support modified optimization while
3217 xsaveopt supports it on xsaveopt enabled systems.
3219 noxsaves [X86] Disables xsaves and xrstors used in saving and
3220 restoring x86 extended register state in compacted
3221 form of xsave area. The kernel will fall back to use
3222 xsaveopt and xrstor to save and restore the states
3223 in standard form of xsave area. By using this
3224 parameter, xsave area per process might occupy more
3225 memory on xsaves enabled systems.
3227 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
3228 wfi(ARM) instruction doesn't work correctly and not to
3229 use it. This is also useful when using JTAG debugger.
3231 no_file_caps Tells the kernel not to honor file capabilities. The
3232 only way then for a file to be executed with privilege
3233 is to be setuid root or executed by root.
3235 nohalt [IA-64] Tells the kernel not to use the power saving
3236 function PAL_HALT_LIGHT when idle. This increases
3237 power-consumption. On the positive side, it reduces
3238 interrupt wake-up latency, which may improve performance
3239 in certain environments such as networked servers or
3242 nohibernate [HIBERNATION] Disable hibernation and resume.
3244 nohz= [KNL] Boottime enable/disable dynamic ticks
3245 Valid arguments: on, off
3248 nohz_full= [KNL,BOOT,SMP,ISOL]
3249 The argument is a cpu list, as described above.
3250 In kernels built with CONFIG_NO_HZ_FULL=y, set
3251 the specified list of CPUs whose tick will be stopped
3252 whenever possible. The boot CPU will be forced outside
3253 the range to maintain the timekeeping. Any CPUs
3254 in this list will have their RCU callbacks offloaded,
3255 just as if they had also been called out in the
3256 rcu_nocbs= boot parameter.
3258 noiotrap [SH] Disables trapped I/O port accesses.
3260 noirqdebug [X86-32] Disables the code which attempts to detect and
3261 disable unhandled interrupt sources.
3263 no_timer_check [X86,APIC] Disables the code which tests for
3264 broken timer IRQ sources.
3266 noisapnp [ISAPNP] Disables ISA PnP code.
3268 noinitrd [RAM] Tells the kernel not to load any configured
3271 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3273 [Deprecated - use intremap=off]
3277 noinvpcid [X86] Disable the INVPCID cpu feature.
3279 nojitter [IA-64] Disables jitter checking for ITC timers.
3281 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3283 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3287 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3288 clock and use the default one.
3290 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time
3291 accounting. steal time is computed, but won't
3292 influence scheduler behaviour
3294 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3296 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3298 noltlbs [PPC] Do not use large page/tlb entries for kernel
3299 lowmem mapping on PPC40x and PPC8xx
3301 nomca [IA-64] Disable machine check abort handling
3303 nomce [X86-32] Disable Machine Check Exception
3305 nomfgpt [X86-32] Disable Multi-Function General Purpose
3306 Timer usage (for AMD Geode machines).
3308 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3309 shutdown the other cpus. Instead use the REBOOT_VECTOR
3312 nomodule Disable module load
3314 nopat [X86] Disable PAT (page attribute table extension of
3315 pagetables) support.
3317 nopcid [X86-64] Disable the PCID cpu feature.
3319 norandmaps Don't use address space randomization. Equivalent to
3320 echo 0 > /proc/sys/kernel/randomize_va_space
3322 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3323 with UP alternatives
3325 nordrand [X86] Disable kernel use of the RDRAND and
3326 RDSEED instructions even if they are supported
3327 by the processor. RDRAND and RDSEED are still
3328 available to user space applications.
3330 noresume [SWSUSP] Disables resume and restores original swap
3333 no-scroll [VGA] Disables scrollback.
3334 This is required for the Braillex ib80-piezo Braille
3335 reader made by F.H. Papenmeier (Germany).
3339 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
3341 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3342 and disable the IO APIC. legacy for "maxcpus=0".
3344 nosoftlockup [KNL] Disable the soft-lockup detector.
3346 nosync [HW,M68K] Disables sync negotiation for all devices.
3348 nowatchdog [KNL] Disable both lockup detectors, i.e.
3349 soft-lockup and NMI watchdog (hard-lockup).
3353 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3355 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
3356 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
3357 Some features depend on CPU0. Known dependencies are:
3358 1. Resume from suspend/hibernate depends on CPU0.
3359 Suspend/hibernate will fail if CPU0 is offline and you
3360 need to online CPU0 before suspend/hibernate.
3361 2. PIC interrupts also depend on CPU0. CPU0 can't be
3362 removed if a PIC interrupt is detected.
3363 It's said poweroff/reboot may depend on CPU0 on some
3364 machines although I haven't seen such issues so far
3365 after CPU0 is offline on a few tested machines.
3366 If the dependencies are under your control, you can
3367 turn on cpu0_hotplug.
3369 nps_mtm_hs_ctr= [KNL,ARC]
3370 This parameter sets the maximum duration, in
3371 cycles, each HW thread of the CTOP can run
3372 without interruptions, before HW switches it.
3373 The actual maximum duration is 16 times this
3375 Format: integer between 1 and 255
3378 nptcg= [IA-64] Override max number of concurrent global TLB
3379 purges which is reported from either PAL_VM_SUMMARY or
3382 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3383 could support. nr_cpus=n : n >= 1 limits the kernel to
3384 support 'n' processors. It could be larger than the
3385 number of already plugged CPU during bootup, later in
3386 runtime you can physically add extra cpu until it reaches
3387 n. So during boot up some boot time memory for per-cpu
3388 variables need be pre-allocated for later physical cpu
3391 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3393 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
3394 Allowed values are enable and disable
3396 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3397 'node', 'default' can be specified
3398 This can be set from sysctl after boot.
3399 See Documentation/admin-guide/sysctl/vm.rst for details.
3401 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3402 See Documentation/core-api/debugging-via-ohci1394.rst for more
3405 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3406 Rather than timing out after 20 ms if an EC
3407 command is not properly ACKed, override the length
3408 of the timeout. We have interrupts disabled while
3409 waiting for the ACK, so if this is set too high
3410 interrupts *may* be lost!
3412 omap_mux= [OMAP] Override bootloader pin multiplexing.
3413 Format: <mux_mode0.mode_name=value>...
3414 For example, to override I2C bus2:
3415 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3417 oprofile.timer= [HW]
3418 Use timer interrupt instead of performance counters
3420 oprofile.cpu_type= Force an oprofile cpu type
3421 This might be useful if you have an older oprofile
3422 userland or if you want common events.
3423 Format: { arch_perfmon }
3424 arch_perfmon: [X86] Force use of architectural
3425 perfmon on Intel CPUs instead of the
3426 CPU specific event set.
3427 timer: [X86] Force use of architectural NMI
3428 timer mode (see also oprofile.timer
3429 for generic hr timer mode)
3431 oops=panic Always panic on oopses. Default is to just kill the
3432 process, but there is a small probability of
3433 deadlocking the machine.
3434 This will also cause panics on machine check exceptions.
3435 Useful together with panic=30 to trigger a reboot.
3438 [KNL] Boolean flag to control whether the page allocator
3439 should randomize its free lists. The randomization may
3440 be automatically enabled if the kernel detects it is
3441 running on a platform with a direct-mapped memory-side
3442 cache, and this parameter can be used to
3443 override/disable that behavior. The state of the flag
3444 can be read from sysfs at:
3445 /sys/module/page_alloc/parameters/shuffle.
3447 page_owner= [KNL] Boot-time page_owner enabling option.
3448 Storage of the information about who allocated
3449 each page is disabled in default. With this switch,
3451 on: enable the feature
3453 page_poison= [KNL] Boot-time parameter changing the state of
3454 poisoning on the buddy allocator, available with
3455 CONFIG_PAGE_POISONING=y.
3456 off: turn off poisoning (default)
3457 on: turn on poisoning
3459 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3460 timeout > 0: seconds before rebooting
3461 timeout = 0: wait forever
3462 timeout < 0: reboot immediately
3465 panic_print= Bitmask for printing system info when panic happens.
3466 User can chose combination of the following bits:
3467 bit 0: print all tasks info
3468 bit 1: print system memory info
3469 bit 2: print timer info
3470 bit 3: print locks info if CONFIG_LOCKDEP is on
3471 bit 4: print ftrace buffer
3472 bit 5: print all printk messages in buffer
3474 panic_on_taint= Bitmask for conditionally calling panic() in add_taint()
3475 Format: <hex>[,nousertaint]
3476 Hexadecimal bitmask representing the set of TAINT flags
3477 that will cause the kernel to panic when add_taint() is
3478 called with any of the flags in this set.
3479 The optional switch "nousertaint" can be utilized to
3480 prevent userspace forced crashes by writing to sysctl
3481 /proc/sys/kernel/tainted any flagset matching with the
3482 bitmask set on panic_on_taint.
3483 See Documentation/admin-guide/tainted-kernels.rst for
3484 extra details on the taint flags that users can pick
3485 to compose the bitmask to assign to panic_on_taint.
3487 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3490 crash_kexec_post_notifiers
3491 Run kdump after running panic-notifiers and dumping
3492 kmsg. This only for the users who doubt kdump always
3493 succeeds in any situation.
3494 Note that this also increases risks of kdump failure,
3495 because some panic notifiers can make the crashed
3496 kernel more unstable.
3498 parkbd.port= [HW] Parallel port number the keyboard adapter is
3499 connected to, default is 0.
3501 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3502 0 for XT, 1 for AT (default is AT).
3505 parport= [HW,PPT] Specify parallel ports. 0 disables.
3506 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3507 Use 'auto' to force the driver to use any
3508 IRQ/DMA settings detected (the default is to
3509 ignore detected IRQ/DMA settings because of
3510 possible conflicts). You can specify the base
3511 address, IRQ, and DMA settings; IRQ and DMA
3512 should be numbers, or 'auto' (for using detected
3513 settings on that particular port), or 'nofifo'
3514 (to avoid using a FIFO even if it is detected).
3515 Parallel ports are assigned in the order they
3516 are specified on the command line, starting
3519 parport_init_mode= [HW,PPT]
3520 Configure VIA parallel port to operate in
3521 a specific mode. This is necessary on Pegasos
3522 computer where firmware has no options for setting
3523 up parallel port mode and sets it to spp.
3524 Currently this function knows 686a and 8231 chips.
3525 Format: [spp|ps2|epp|ecp|ecpepp]
3528 Halt all CPUs after the first oops has been printed for
3529 the specified number of seconds. This is to be used if
3530 your oopses keep scrolling off the screen.
3535 See header of drivers/block/paride/pcd.c.
3536 See also Documentation/admin-guide/blockdev/paride.rst.
3538 pci=option[,option...] [PCI] various PCI subsystem options.
3540 Some options herein operate on a specific device
3541 or a set of devices (<pci_dev>). These are
3542 specified in one of the following formats:
3544 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3545 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3547 Note: the first format specifies a PCI
3548 bus/device/function address which may change
3549 if new hardware is inserted, if motherboard
3550 firmware changes, or due to changes caused
3551 by other kernel parameters. If the
3552 domain is left unspecified, it is
3553 taken to be zero. Optionally, a path
3554 to a device through multiple device/function
3555 addresses can be specified after the base
3556 address (this is more robust against
3557 renumbering issues). The second format
3558 selects devices using IDs from the
3559 configuration space which may match multiple
3560 devices in the system.
3562 earlydump dump PCI config space before the kernel
3564 off [X86] don't probe for the PCI bus
3565 bios [X86-32] force use of PCI BIOS, don't access
3566 the hardware directly. Use this if your machine
3567 has a non-standard PCI host bridge.
3568 nobios [X86-32] disallow use of PCI BIOS, only direct
3569 hardware access methods are allowed. Use this
3570 if you experience crashes upon bootup and you
3571 suspect they are caused by the BIOS.
3572 conf1 [X86] Force use of PCI Configuration Access
3573 Mechanism 1 (config address in IO port 0xCF8,
3574 data in IO port 0xCFC, both 32-bit).
3575 conf2 [X86] Force use of PCI Configuration Access
3576 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3577 the function, IO port 0xCFA, also 8-bit, sets
3578 bus number. The config space is then accessed
3579 through ports 0xC000-0xCFFF).
3580 See http://wiki.osdev.org/PCI for more info
3581 on the configuration access mechanisms.
3582 noaer [PCIE] If the PCIEAER kernel config parameter is
3583 enabled, this kernel boot option can be used to
3584 disable the use of PCIE advanced error reporting.
3585 nodomains [PCI] Disable support for multiple PCI
3586 root domains (aka PCI segments, in ACPI-speak).
3587 nommconf [X86] Disable use of MMCONFIG for PCI
3589 check_enable_amd_mmconf [X86] check for and enable
3590 properly configured MMIO access to PCI
3591 config space on AMD family 10h CPU
3592 nomsi [MSI] If the PCI_MSI kernel config parameter is
3593 enabled, this kernel boot option can be used to
3594 disable the use of MSI interrupts system-wide.
3595 noioapicquirk [APIC] Disable all boot interrupt quirks.
3596 Safety option to keep boot IRQs enabled. This
3597 should never be necessary.
3598 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3599 primary IO-APIC for bridges that cannot disable
3600 boot IRQs. This fixes a source of spurious IRQs
3601 when the system masks IRQs.
3602 noioapicreroute [APIC] Disable workaround that uses the
3603 boot IRQ equivalent of an IRQ that connects to
3604 a chipset where boot IRQs cannot be disabled.
3605 The opposite of ioapicreroute.
3606 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3607 routing table. These calls are known to be buggy
3608 on several machines and they hang the machine
3609 when used, but on other computers it's the only
3610 way to get the interrupt routing table. Try
3611 this option if the kernel is unable to allocate
3612 IRQs or discover secondary PCI buses on your
3614 rom [X86] Assign address space to expansion ROMs.
3615 Use with caution as certain devices share
3616 address decoders between ROMs and other
3618 norom [X86] Do not assign address space to
3619 expansion ROMs that do not already have
3620 BIOS assigned address ranges.
3621 nobar [X86] Do not assign address space to the
3622 BARs that weren't assigned by the BIOS.
3623 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3624 assigned automatically to PCI devices. You can
3625 make the kernel exclude IRQs of your ISA cards
3627 pirqaddr=0xAAAAA [X86] Specify the physical address
3628 of the PIRQ table (normally generated
3629 by the BIOS) if it is outside the
3630 F0000h-100000h range.
3631 lastbus=N [X86] Scan all buses thru bus #N. Can be
3632 useful if the kernel is unable to find your
3633 secondary buses and you want to tell it
3634 explicitly which ones they are.
3635 assign-busses [X86] Always assign all PCI bus
3636 numbers ourselves, overriding
3637 whatever the firmware may have done.
3638 usepirqmask [X86] Honor the possible IRQ mask stored
3639 in the BIOS $PIR table. This is needed on
3640 some systems with broken BIOSes, notably
3641 some HP Pavilion N5400 and Omnibook XE3
3642 notebooks. This will have no effect if ACPI
3643 IRQ routing is enabled.
3644 noacpi [X86] Do not use ACPI for IRQ routing
3645 or for PCI scanning.
3646 use_crs [X86] Use PCI host bridge window information
3647 from ACPI. On BIOSes from 2008 or later, this
3648 is enabled by default. If you need to use this,
3649 please report a bug.
3650 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3651 If you need to use this, please report a bug.
3652 routeirq Do IRQ routing for all PCI devices.
3653 This is normally done in pci_enable_device(),
3654 so this option is a temporary workaround
3655 for broken drivers that don't call it.
3656 skip_isa_align [X86] do not align io start addr, so can
3657 handle more pci cards
3658 noearly [X86] Don't do any early type 1 scanning.
3659 This might help on some broken boards which
3660 machine check when some devices' config space
3661 is read. But various workarounds are disabled
3662 and some IOMMU drivers will not work.
3663 bfsort Sort PCI devices into breadth-first order.
3664 This sorting is done to get a device
3665 order compatible with older (<= 2.4) kernels.
3666 nobfsort Don't sort PCI devices into breadth-first order.
3667 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3668 tuning and use the BIOS-configured MPS defaults.
3669 pcie_bus_safe Set every device's MPS to the largest value
3670 supported by all devices below the root complex.
3671 pcie_bus_perf Set device MPS to the largest allowable MPS
3672 based on its parent bus. Also set MRRS (Max
3673 Read Request Size) to the largest supported
3674 value (no larger than the MPS that the device
3675 or bus can support) for best performance.
3676 pcie_bus_peer2peer Set every device's MPS to 128B, which
3677 every device is guaranteed to support. This
3678 configuration allows peer-to-peer DMA between
3679 any pair of devices, possibly at the cost of
3680 reduced performance. This also guarantees
3681 that hot-added devices will work.
3682 cbiosize=nn[KMG] The fixed amount of bus space which is
3683 reserved for the CardBus bridge's IO window.
3684 The default value is 256 bytes.
3685 cbmemsize=nn[KMG] The fixed amount of bus space which is
3686 reserved for the CardBus bridge's memory
3687 window. The default value is 64 megabytes.
3690 [<order of align>@]<pci_dev>[; ...]
3691 Specifies alignment and device to reassign
3692 aligned memory resources. How to
3693 specify the device is described above.
3694 If <order of align> is not specified,
3695 PAGE_SIZE is used as alignment.
3696 A PCI-PCI bridge can be specified if resource
3697 windows need to be expanded.
3698 To specify the alignment for several
3699 instances of a device, the PCI vendor,
3700 device, subvendor, and subdevice may be
3701 specified, e.g., 12@pci:8086:9c22:103c:198f
3702 for 4096-byte alignment.
3703 ecrc= Enable/disable PCIe ECRC (transaction layer
3704 end-to-end CRC checking).
3705 bios: Use BIOS/firmware settings. This is the
3709 hpiosize=nn[KMG] The fixed amount of bus space which is
3710 reserved for hotplug bridge's IO window.
3711 Default size is 256 bytes.
3712 hpmmiosize=nn[KMG] The fixed amount of bus space which is
3713 reserved for hotplug bridge's MMIO window.
3714 Default size is 2 megabytes.
3715 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is
3716 reserved for hotplug bridge's MMIO_PREF window.
3717 Default size is 2 megabytes.
3718 hpmemsize=nn[KMG] The fixed amount of bus space which is
3719 reserved for hotplug bridge's MMIO and
3721 Default size is 2 megabytes.
3722 hpbussize=nn The minimum amount of additional bus numbers
3723 reserved for buses below a hotplug bridge.
3725 realloc= Enable/disable reallocating PCI bridge resources
3726 if allocations done by BIOS are too small to
3727 accommodate resources required by all child
3729 off: Turn realloc off
3731 realloc same as realloc=on
3732 noari do not use PCIe ARI.
3733 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
3734 do not use PCIe ATS (and IOMMU device IOTLB).
3735 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3736 only look for one device below a PCIe downstream
3738 big_root_window Try to add a big 64bit memory window to the PCIe
3739 root complex on AMD CPUs. Some GFX hardware
3740 can resize a BAR to allow access to all VRAM.
3741 Adding the window is slightly risky (it may
3742 conflict with unreported devices), so this
3744 disable_acs_redir=<pci_dev>[; ...]
3745 Specify one or more PCI devices (in the format
3746 specified above) separated by semicolons.
3747 Each device specified will have the PCI ACS
3748 redirect capabilities forced off which will
3749 allow P2P traffic between devices through
3750 bridges without forcing it upstream. Note:
3751 this removes isolation between devices and
3752 may put more devices in an IOMMU group.
3753 force_floating [S390] Force usage of floating interrupts.
3754 nomio [S390] Do not use MIO instructions.
3755 norid [S390] ignore the RID field and force use of
3756 one PCI domain per PCI function
3758 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3761 force Enable ASPM even on devices that claim not to support it.
3762 WARNING: Forcing ASPM on may cause system lockups.
3764 pcie_ports= [PCIE] PCIe port services handling:
3765 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
3766 even if the platform doesn't give the OS permission to
3767 use them. This may cause conflicts if the platform
3768 also tries to use these services.
3769 dpc-native Use native PCIe service for DPC only. May
3770 cause conflicts if firmware uses AER or DPC.
3771 compat Disable native PCIe services (PME, AER, DPC, PCIe
3774 pcie_port_pm= [PCIE] PCIe port power management handling:
3775 off Disable power management of all PCIe ports
3776 force Forcibly enable power management of all PCIe ports
3778 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3779 nomsi Do not use MSI for native PCIe PME signaling (this makes
3780 all PCIe root ports use INTx for all services).
3782 pcmv= [HW,PCMCIA] BadgePAD 4
3786 Keep all power-domains already enabled by bootloader on,
3787 even if no driver has claimed them. This is useful
3788 for debug and development, but should not be
3789 needed on a platform with proper driver support.
3792 See Documentation/admin-guide/blockdev/paride.rst.
3794 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3797 See arch/parisc/kernel/pdc_chassis.c
3799 percpu_alloc= Select which percpu first chunk allocator to use.
3800 Currently supported values are "embed" and "page".
3801 Archs may support subset or none of the selections.
3802 See comments in mm/percpu.c for details on each
3803 allocator. This parameter is primarily for debugging
3804 and performance comparison.
3807 See Documentation/admin-guide/blockdev/paride.rst.
3810 See Documentation/admin-guide/blockdev/paride.rst.
3812 pirq= [SMP,APIC] Manual mp-table setup
3813 See Documentation/x86/i386/IO-APIC.rst.
3815 plip= [PPT,NET] Parallel port network link
3816 Format: { parport<nr> | timid | 0 }
3817 See also Documentation/admin-guide/parport.rst.
3819 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3820 Override pmtimer IOPort with a hex value.
3823 pm_debug_messages [SUSPEND,KNL]
3824 Enable suspend/resume debug messages during boot up.
3827 Enable PNP debug messages (depends on the
3828 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3829 via /sys/module/pnp/parameters/debug. We always show
3830 current resource usage; turning this on also shows
3831 possible settings and some assignment information.
3837 { on | off | curr | res | no-curr | no-res }
3840 [ISAPNP] Exclude IRQs for the autoconfiguration
3843 [ISAPNP] Exclude DMAs for the autoconfiguration
3845 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3846 Ranges are in pairs (I/O port base and size).
3849 [ISAPNP] Exclude memory regions for the
3851 Ranges are in pairs (memory base and size).
3853 ports= [IP_VS_FTP] IPVS ftp helper module
3855 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3857 Format: <port>,<port>....
3859 powersave=off [PPC] This option disables power saving features.
3860 It specifically disables cpuidle and sets the
3861 platform machine description specific power_save
3862 function to NULL. On Idle the CPU just reduces
3865 ppc_strict_facility_enable
3866 [PPC] This option catches any kernel floating point,
3867 Altivec, VSX and SPE outside of regions specifically
3868 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3869 There is some performance impact when enabling this.
3873 Disable Hardware Transactional Memory
3875 print-fatal-signals=
3876 [KNL] debug: print fatal signals
3878 If enabled, warn about various signal handling
3879 related application anomalies: too many signals,
3880 too many POSIX.1 timers, fatal signals causing a
3883 If you hit the warning due to signal overflow,
3884 you might want to try "ulimit -i unlimited".
3888 printk.always_kmsg_dump=
3889 Trigger kmsg_dump for cases other than kernel oops or
3891 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3894 printk.devkmsg={on,off,ratelimit}
3895 Control writing to /dev/kmsg.
3896 on - unlimited logging to /dev/kmsg from userspace
3897 off - logging to /dev/kmsg disabled
3898 ratelimit - ratelimit the logging
3901 printk.time= Show timing data prefixed to each printk message line
3902 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3904 processor.max_cstate= [HW,ACPI]
3905 Limit processor to maximum C-state
3906 max_cstate=9 overrides any DMI blacklist limit.
3908 processor.nocst [HW,ACPI]
3909 Ignore the _CST method to determine C-states,
3910 instead using the legacy FADT method
3912 profile= [KNL] Enable kernel profiling via /proc/profile
3913 Format: [<profiletype>,]<number>
3914 Param: <profiletype>: "schedule", "sleep", or "kvm"
3915 [defaults to kernel profiling]
3916 Param: "schedule" - profile schedule points.
3917 Param: "sleep" - profile D-state sleeping (millisecs).
3918 Requires CONFIG_SCHEDSTATS
3919 Param: "kvm" - profile VM exits.
3920 Param: <number> - step/bucket size as a power of 2 for
3921 statistical time based profiling.
3923 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3925 See Documentation/admin-guide/blockdev/ramdisk.rst.
3927 prot_virt= [S390] enable hosting protected virtual machines
3928 isolated from the hypervisor (if hardware supports
3932 psi= [KNL] Enable or disable pressure stall information
3936 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3937 probe for; one of (bare|imps|exps|lifebook|any).
3938 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3940 psmouse.resetafter= [HW,MOUSE]
3941 Try to reset the device after so many bad packets
3944 [HW,MOUSE] Set desired mouse resolution, in dpi.
3945 psmouse.smartscroll=
3946 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3947 0 = disabled, 1 = enabled (default).
3949 pstore.backend= Specify the name of the pstore backend to use
3952 See Documentation/admin-guide/blockdev/paride.rst.
3954 pti= [X86_64] Control Page Table Isolation of user and
3955 kernel address spaces. Disabling this feature
3956 removes hardening, but improves performance of
3957 system calls and interrupts.
3959 on - unconditionally enable
3960 off - unconditionally disable
3961 auto - kernel detects whether your CPU model is
3962 vulnerable to issues that PTI mitigates
3964 Not specifying this option is equivalent to pti=auto.
3967 Equivalent to pti=off
3970 [KNL] Number of legacy pty's. Overwrites compiled-in
3973 quiet [KNL] Disable most log messages
3978 See Documentation/admin-guide/md.rst.
3980 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3981 See Documentation/admin-guide/blockdev/ramdisk.rst.
3983 random.trust_cpu={on,off}
3984 [KNL] Enable or disable trusting the use of the
3985 CPU's random number generator (if available) to
3986 fully seed the kernel's CRNG. Default is controlled
3987 by CONFIG_RANDOM_TRUST_CPU.
3989 ras=option[,option,...] [KNL] RAS-specific options
3992 Disable the Correctable Errors Collector,
3993 see CONFIG_RAS_CEC help text.
3996 The argument is a cpu list, as described above,
3997 except that the string "all" can be used to
3998 specify every CPU on the system.
4000 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
4001 the specified list of CPUs to be no-callback CPUs.
4002 Invocation of these CPUs' RCU callbacks will be
4003 offloaded to "rcuox/N" kthreads created for that
4004 purpose, where "x" is "p" for RCU-preempt, and
4005 "s" for RCU-sched, and "N" is the CPU number.
4006 This reduces OS jitter on the offloaded CPUs,
4007 which can be useful for HPC and real-time
4008 workloads. It can also improve energy efficiency
4009 for asymmetric multiprocessors.
4012 Rather than requiring that offloaded CPUs
4013 (specified by rcu_nocbs= above) explicitly
4014 awaken the corresponding "rcuoN" kthreads,
4015 make these kthreads poll for callbacks.
4016 This improves the real-time response for the
4017 offloaded CPUs by relieving them of the need to
4018 wake up the corresponding kthread, but degrades
4019 energy efficiency by requiring that the kthreads
4020 periodically wake up to do the polling.
4022 rcutree.blimit= [KNL]
4023 Set maximum number of finished RCU callbacks to
4024 process in one batch.
4026 rcutree.dump_tree= [KNL]
4027 Dump the structure of the rcu_node combining tree
4028 out at early boot. This is used for diagnostic
4029 purposes, to verify correct tree setup.
4031 rcutree.gp_cleanup_delay= [KNL]
4032 Set the number of jiffies to delay each step of
4033 RCU grace-period cleanup.
4035 rcutree.gp_init_delay= [KNL]
4036 Set the number of jiffies to delay each step of
4037 RCU grace-period initialization.
4039 rcutree.gp_preinit_delay= [KNL]
4040 Set the number of jiffies to delay each step of
4041 RCU grace-period pre-initialization, that is,
4042 the propagation of recent CPU-hotplug changes up
4043 the rcu_node combining tree.
4045 rcutree.use_softirq= [KNL]
4046 If set to zero, move all RCU_SOFTIRQ processing to
4047 per-CPU rcuc kthreads. Defaults to a non-zero
4048 value, meaning that RCU_SOFTIRQ is used by default.
4049 Specify rcutree.use_softirq=0 to use rcuc kthreads.
4051 rcutree.rcu_fanout_exact= [KNL]
4052 Disable autobalancing of the rcu_node combining
4053 tree. This is used by rcutorture, and might
4054 possibly be useful for architectures having high
4055 cache-to-cache transfer latencies.
4057 rcutree.rcu_fanout_leaf= [KNL]
4058 Change the number of CPUs assigned to each
4059 leaf rcu_node structure. Useful for very
4060 large systems, which will choose the value 64,
4061 and for NUMA systems with large remote-access
4062 latencies, which will choose a value aligned
4063 with the appropriate hardware boundaries.
4065 rcutree.rcu_min_cached_objs= [KNL]
4066 Minimum number of objects which are cached and
4067 maintained per one CPU. Object size is equal
4068 to PAGE_SIZE. The cache allows to reduce the
4069 pressure to page allocator, also it makes the
4070 whole algorithm to behave better in low memory
4073 rcutree.jiffies_till_first_fqs= [KNL]
4074 Set delay from grace-period initialization to
4075 first attempt to force quiescent states.
4076 Units are jiffies, minimum value is zero,
4077 and maximum value is HZ.
4079 rcutree.jiffies_till_next_fqs= [KNL]
4080 Set delay between subsequent attempts to force
4081 quiescent states. Units are jiffies, minimum
4082 value is one, and maximum value is HZ.
4084 rcutree.jiffies_till_sched_qs= [KNL]
4085 Set required age in jiffies for a
4086 given grace period before RCU starts
4087 soliciting quiescent-state help from
4088 rcu_note_context_switch() and cond_resched().
4089 If not specified, the kernel will calculate
4090 a value based on the most recent settings
4091 of rcutree.jiffies_till_first_fqs
4092 and rcutree.jiffies_till_next_fqs.
4093 This calculated value may be viewed in
4094 rcutree.jiffies_to_sched_qs. Any attempt to set
4095 rcutree.jiffies_to_sched_qs will be cheerfully
4098 rcutree.kthread_prio= [KNL,BOOT]
4099 Set the SCHED_FIFO priority of the RCU per-CPU
4100 kthreads (rcuc/N). This value is also used for
4101 the priority of the RCU boost threads (rcub/N)
4102 and for the RCU grace-period kthreads (rcu_bh,
4103 rcu_preempt, and rcu_sched). If RCU_BOOST is
4104 set, valid values are 1-99 and the default is 1
4105 (the least-favored priority). Otherwise, when
4106 RCU_BOOST is not set, valid values are 0-99 and
4107 the default is zero (non-realtime operation).
4109 rcutree.rcu_nocb_gp_stride= [KNL]
4110 Set the number of NOCB callback kthreads in
4111 each group, which defaults to the square root
4112 of the number of CPUs. Larger numbers reduce
4113 the wakeup overhead on the global grace-period
4114 kthread, but increases that same overhead on
4115 each group's NOCB grace-period kthread.
4117 rcutree.qhimark= [KNL]
4118 Set threshold of queued RCU callbacks beyond which
4119 batch limiting is disabled.
4121 rcutree.qlowmark= [KNL]
4122 Set threshold of queued RCU callbacks below which
4123 batch limiting is re-enabled.
4125 rcutree.qovld= [KNL]
4126 Set threshold of queued RCU callbacks beyond which
4127 RCU's force-quiescent-state scan will aggressively
4128 enlist help from cond_resched() and sched IPIs to
4129 help CPUs more quickly reach quiescent states.
4130 Set to less than zero to make this be set based
4131 on rcutree.qhimark at boot time and to zero to
4132 disable more aggressive help enlistment.
4134 rcutree.rcu_idle_gp_delay= [KNL]
4135 Set wakeup interval for idle CPUs that have
4136 RCU callbacks (RCU_FAST_NO_HZ=y).
4138 rcutree.rcu_idle_lazy_gp_delay= [KNL]
4139 Set wakeup interval for idle CPUs that have
4140 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
4141 Lazy RCU callbacks are those which RCU can
4142 prove do nothing more than free memory.
4144 rcutree.rcu_kick_kthreads= [KNL]
4145 Cause the grace-period kthread to get an extra
4146 wake_up() if it sleeps three times longer than
4147 it should at force-quiescent-state time.
4148 This wake_up() will be accompanied by a
4149 WARN_ONCE() splat and an ftrace_dump().
4151 rcutree.sysrq_rcu= [KNL]
4152 Commandeer a sysrq key to dump out Tree RCU's
4153 rcu_node tree with an eye towards determining
4154 why a new grace period has not yet started.
4156 rcuperf.gp_async= [KNL]
4157 Measure performance of asynchronous
4158 grace-period primitives such as call_rcu().
4160 rcuperf.gp_async_max= [KNL]
4161 Specify the maximum number of outstanding
4162 callbacks per writer thread. When a writer
4163 thread exceeds this limit, it invokes the
4164 corresponding flavor of rcu_barrier() to allow
4165 previously posted callbacks to drain.
4167 rcuperf.gp_exp= [KNL]
4168 Measure performance of expedited synchronous
4169 grace-period primitives.
4171 rcuperf.holdoff= [KNL]
4172 Set test-start holdoff period. The purpose of
4173 this parameter is to delay the start of the
4174 test until boot completes in order to avoid
4177 rcuperf.kfree_rcu_test= [KNL]
4178 Set to measure performance of kfree_rcu() flooding.
4180 rcuperf.kfree_nthreads= [KNL]
4181 The number of threads running loops of kfree_rcu().
4183 rcuperf.kfree_alloc_num= [KNL]
4184 Number of allocations and frees done in an iteration.
4186 rcuperf.kfree_loops= [KNL]
4187 Number of loops doing rcuperf.kfree_alloc_num number
4188 of allocations and frees.
4190 rcuperf.nreaders= [KNL]
4191 Set number of RCU readers. The value -1 selects
4192 N, where N is the number of CPUs. A value
4193 "n" less than -1 selects N-n+1, where N is again
4194 the number of CPUs. For example, -2 selects N
4195 (the number of CPUs), -3 selects N+1, and so on.
4196 A value of "n" less than or equal to -N selects
4199 rcuperf.nwriters= [KNL]
4200 Set number of RCU writers. The values operate
4201 the same as for rcuperf.nreaders.
4202 N, where N is the number of CPUs
4204 rcuperf.perf_type= [KNL]
4205 Specify the RCU implementation to test.
4207 rcuperf.shutdown= [KNL]
4208 Shut the system down after performance tests
4209 complete. This is useful for hands-off automated
4212 rcuperf.verbose= [KNL]
4213 Enable additional printk() statements.
4215 rcuperf.writer_holdoff= [KNL]
4216 Write-side holdoff between grace periods,
4217 in microseconds. The default of zero says
4220 rcutorture.fqs_duration= [KNL]
4221 Set duration of force_quiescent_state bursts
4224 rcutorture.fqs_holdoff= [KNL]
4225 Set holdoff time within force_quiescent_state bursts
4228 rcutorture.fqs_stutter= [KNL]
4229 Set wait time between force_quiescent_state bursts
4232 rcutorture.fwd_progress= [KNL]
4233 Enable RCU grace-period forward-progress testing
4234 for the types of RCU supporting this notion.
4236 rcutorture.fwd_progress_div= [KNL]
4237 Specify the fraction of a CPU-stall-warning
4238 period to do tight-loop forward-progress testing.
4240 rcutorture.fwd_progress_holdoff= [KNL]
4241 Number of seconds to wait between successive
4242 forward-progress tests.
4244 rcutorture.fwd_progress_need_resched= [KNL]
4245 Enclose cond_resched() calls within checks for
4246 need_resched() during tight-loop forward-progress
4249 rcutorture.gp_cond= [KNL]
4250 Use conditional/asynchronous update-side
4251 primitives, if available.
4253 rcutorture.gp_exp= [KNL]
4254 Use expedited update-side primitives, if available.
4256 rcutorture.gp_normal= [KNL]
4257 Use normal (non-expedited) asynchronous
4258 update-side primitives, if available.
4260 rcutorture.gp_sync= [KNL]
4261 Use normal (non-expedited) synchronous
4262 update-side primitives, if available. If all
4263 of rcutorture.gp_cond=, rcutorture.gp_exp=,
4264 rcutorture.gp_normal=, and rcutorture.gp_sync=
4265 are zero, rcutorture acts as if is interpreted
4266 they are all non-zero.
4268 rcutorture.n_barrier_cbs= [KNL]
4269 Set callbacks/threads for rcu_barrier() testing.
4271 rcutorture.nfakewriters= [KNL]
4272 Set number of concurrent RCU writers. These just
4273 stress RCU, they don't participate in the actual
4274 test, hence the "fake".
4276 rcutorture.nreaders= [KNL]
4277 Set number of RCU readers. The value -1 selects
4278 N-1, where N is the number of CPUs. A value
4279 "n" less than -1 selects N-n-2, where N is again
4280 the number of CPUs. For example, -2 selects N
4281 (the number of CPUs), -3 selects N+1, and so on.
4283 rcutorture.object_debug= [KNL]
4284 Enable debug-object double-call_rcu() testing.
4286 rcutorture.onoff_holdoff= [KNL]
4287 Set time (s) after boot for CPU-hotplug testing.
4289 rcutorture.onoff_interval= [KNL]
4290 Set time (jiffies) between CPU-hotplug operations,
4291 or zero to disable CPU-hotplug testing.
4293 rcutorture.read_exit= [KNL]
4294 Set the number of read-then-exit kthreads used
4295 to test the interaction of RCU updaters and
4296 task-exit processing.
4298 rcutorture.read_exit_burst= [KNL]
4299 The number of times in a given read-then-exit
4300 episode that a set of read-then-exit kthreads
4303 rcutorture.read_exit_delay= [KNL]
4304 The delay, in seconds, between successive
4305 read-then-exit testing episodes.
4307 rcutorture.shuffle_interval= [KNL]
4308 Set task-shuffle interval (s). Shuffling tasks
4309 allows some CPUs to go into dyntick-idle mode
4310 during the rcutorture test.
4312 rcutorture.shutdown_secs= [KNL]
4313 Set time (s) after boot system shutdown. This
4314 is useful for hands-off automated testing.
4316 rcutorture.stall_cpu= [KNL]
4317 Duration of CPU stall (s) to test RCU CPU stall
4318 warnings, zero to disable.
4320 rcutorture.stall_cpu_block= [KNL]
4321 Sleep while stalling if set. This will result
4322 in warnings from preemptible RCU in addition
4323 to any other stall-related activity.
4325 rcutorture.stall_cpu_holdoff= [KNL]
4326 Time to wait (s) after boot before inducing stall.
4328 rcutorture.stall_cpu_irqsoff= [KNL]
4329 Disable interrupts while stalling if set.
4331 rcutorture.stall_gp_kthread= [KNL]
4332 Duration (s) of forced sleep within RCU
4333 grace-period kthread to test RCU CPU stall
4334 warnings, zero to disable. If both stall_cpu
4335 and stall_gp_kthread are specified, the
4336 kthread is starved first, then the CPU.
4338 rcutorture.stat_interval= [KNL]
4339 Time (s) between statistics printk()s.
4341 rcutorture.stutter= [KNL]
4342 Time (s) to stutter testing, for example, specifying
4343 five seconds causes the test to run for five seconds,
4344 wait for five seconds, and so on. This tests RCU's
4345 ability to transition abruptly to and from idle.
4347 rcutorture.test_boost= [KNL]
4348 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
4349 "Maybe" means test if the RCU implementation
4350 under test support RCU priority boosting.
4352 rcutorture.test_boost_duration= [KNL]
4353 Duration (s) of each individual boost test.
4355 rcutorture.test_boost_interval= [KNL]
4356 Interval (s) between each boost test.
4358 rcutorture.test_no_idle_hz= [KNL]
4359 Test RCU's dyntick-idle handling. See also the
4360 rcutorture.shuffle_interval parameter.
4362 rcutorture.torture_type= [KNL]
4363 Specify the RCU implementation to test.
4365 rcutorture.verbose= [KNL]
4366 Enable additional printk() statements.
4368 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
4369 Dump ftrace buffer after reporting RCU CPU
4372 rcupdate.rcu_cpu_stall_suppress= [KNL]
4373 Suppress RCU CPU stall warning messages.
4375 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL]
4376 Suppress RCU CPU stall warning messages and
4377 rcutorture writer stall warnings that occur
4378 during early boot, that is, during the time
4379 before the init task is spawned.
4381 rcupdate.rcu_cpu_stall_timeout= [KNL]
4382 Set timeout for RCU CPU stall warning messages.
4384 rcupdate.rcu_expedited= [KNL]
4385 Use expedited grace-period primitives, for
4386 example, synchronize_rcu_expedited() instead
4387 of synchronize_rcu(). This reduces latency,
4388 but can increase CPU utilization, degrade
4389 real-time latency, and degrade energy efficiency.
4390 No effect on CONFIG_TINY_RCU kernels.
4392 rcupdate.rcu_normal= [KNL]
4393 Use only normal grace-period primitives,
4394 for example, synchronize_rcu() instead of
4395 synchronize_rcu_expedited(). This improves
4396 real-time latency, CPU utilization, and
4397 energy efficiency, but can expose users to
4398 increased grace-period latency. This parameter
4399 overrides rcupdate.rcu_expedited. No effect on
4400 CONFIG_TINY_RCU kernels.
4402 rcupdate.rcu_normal_after_boot= [KNL]
4403 Once boot has completed (that is, after
4404 rcu_end_inkernel_boot() has been invoked), use
4405 only normal grace-period primitives. No effect
4406 on CONFIG_TINY_RCU kernels.
4408 rcupdate.rcu_task_ipi_delay= [KNL]
4409 Set time in jiffies during which RCU tasks will
4410 avoid sending IPIs, starting with the beginning
4411 of a given grace period. Setting a large
4412 number avoids disturbing real-time workloads,
4413 but lengthens grace periods.
4415 rcupdate.rcu_task_stall_timeout= [KNL]
4416 Set timeout in jiffies for RCU task stall warning
4417 messages. Disable with a value less than or equal
4420 rcupdate.rcu_self_test= [KNL]
4421 Run the RCU early boot self tests
4425 Run specified binary instead of /init from the ramdisk,
4426 used for early userspace startup. See initrd.
4429 force - Override the decision by the kernel to hide the
4430 advertisement of RDRAND support (this affects
4431 certain AMD processors because of buggy BIOS
4432 support, specifically around the suspend/resume
4436 Turn on/off individual RDT features. List is:
4437 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
4439 E.g. to turn on cmt and turn off mba use:
4443 Format (x86 or x86_64):
4444 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
4446 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
4448 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
4449 (prefix with 'panic_' to set mode for panic
4451 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
4452 reboot_force is either force or not specified,
4453 reboot_cpu is s[mp]#### with #### being the processor
4454 to be used for rebooting.
4456 refscale.holdoff= [KNL]
4457 Set test-start holdoff period. The purpose of
4458 this parameter is to delay the start of the
4459 test until boot completes in order to avoid
4462 refscale.loops= [KNL]
4463 Set the number of loops over the synchronization
4464 primitive under test. Increasing this number
4465 reduces noise due to loop start/end overhead,
4466 but the default has already reduced the per-pass
4467 noise to a handful of picoseconds on ca. 2020
4470 refscale.nreaders= [KNL]
4471 Set number of readers. The default value of -1
4472 selects N, where N is roughly 75% of the number
4473 of CPUs. A value of zero is an interesting choice.
4475 refscale.nruns= [KNL]
4476 Set number of runs, each of which is dumped onto
4479 refscale.readdelay= [KNL]
4480 Set the read-side critical-section duration,
4481 measured in microseconds.
4483 refscale.scale_type= [KNL]
4484 Specify the read-protection implementation to test.
4486 refscale.shutdown= [KNL]
4487 Shut down the system at the end of the performance
4488 test. This defaults to 1 (shut it down) when
4489 rcuperf is built into the kernel and to 0 (leave
4490 it running) when rcuperf is built as a module.
4492 refscale.verbose= [KNL]
4493 Enable additional printk() statements.
4496 [KNL, SMP] Set scheduler's default relax_domain_level.
4497 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
4499 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
4500 Format: <base1>,<size1>[,<base2>,<size2>,...]
4501 Reserve I/O ports or memory so the kernel won't use
4502 them. If <base> is less than 0x10000, the region
4503 is assumed to be I/O ports; otherwise it is memory.
4505 reservetop= [X86-32]
4507 Reserves a hole at the top of the kernel virtual
4512 Set the amount of memory to reserve for BIOS at
4513 the bottom of the address space.
4515 reset_devices [KNL] Force drivers to reset the underlying device
4516 during initialization.
4519 Specify the partition device for software suspend
4521 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
4523 resume_offset= [SWSUSP]
4524 Specify the offset from the beginning of the partition
4525 given by "resume=" at which the swap header is located,
4526 in <PAGE_SIZE> units (needed only for swap files).
4527 See Documentation/power/swsusp-and-swap-files.rst
4529 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
4530 read the resume files
4532 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
4533 Useful for devices that are detected asynchronously
4534 (e.g. USB and MMC devices).
4536 hibernate= [HIBERNATION]
4537 noresume Don't check if there's a hibernation image
4538 present during boot.
4539 nocompress Don't compress/decompress hibernation images.
4540 no Disable hibernation and resume.
4541 protect_image Turn on image protection during restoration
4542 (that will set all pages holding image data
4543 during restoration read-only).
4545 retain_initrd [RAM] Keep initrd memory after extraction
4547 rfkill.default_state=
4548 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
4549 etc. communication is blocked by default.
4552 rfkill.master_switch_mode=
4553 0 The "airplane mode" button does nothing.
4554 1 The "airplane mode" button toggles between everything
4555 blocked and the previous configuration.
4556 2 The "airplane mode" button toggles between everything
4557 blocked and everything unblocked.
4559 rhash_entries= [KNL,NET]
4560 Set number of hash buckets for route cache
4563 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4566 ro [KNL] Mount root device read-only on boot
4569 on Mark read-only kernel memory as read-only (default).
4570 off Leave read-only kernel memory writable for debugging.
4573 Enable the uart passthrough on the designated usb port
4574 on Rockchip SoCs. When active, the signals of the
4575 debug-uart get routed to the D+ and D- pins of the usb
4576 port and the regular usb controller gets disabled.
4578 root= [KNL] Root filesystem
4579 See name_to_dev_t comment in init/do_mounts.c.
4581 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4582 mount the root filesystem
4584 rootflags= [KNL] Set root filesystem mount option string
4586 rootfstype= [KNL] Set root filesystem type
4588 rootwait [KNL] Wait (indefinitely) for root device to show up.
4589 Useful for devices that are detected asynchronously
4590 (e.g. USB and MMC devices).
4592 rproc_mem=nn[KMG][@address]
4593 [KNL,ARM,CMA] Remoteproc physical memory block.
4594 Memory area to be used by remote processor image,
4597 rw [KNL] Mount root device read-write on boot
4599 S [KNL] Run init in single mode
4601 s390_iommu= [HW,S390]
4602 Set s390 IOTLB flushing mode
4604 With strict flushing every unmap operation will result in
4605 an IOTLB flush. Default is lazy flushing before reuse,
4609 See drivers/net/irda/sa1100_ir.c.
4611 sbni= [NET] Granch SBNI12 leased line adapter
4613 sched_debug [KNL] Enables verbose scheduler debug messages.
4615 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4616 Allowed values are enable and disable. This feature
4617 incurs a small amount of overhead in the scheduler
4618 but is useful for debugging and performance tuning.
4620 sched_thermal_decay_shift=
4621 [KNL, SMP] Set a decay shift for scheduler thermal
4622 pressure signal. Thermal pressure signal follows the
4623 default decay period of other scheduler pelt
4624 signals(usually 32 ms but configurable). Setting
4625 sched_thermal_decay_shift will left shift the decay
4626 period for the thermal pressure signal by the shift
4628 i.e. with the default pelt decay period of 32 ms
4629 sched_thermal_decay_shift thermal pressure decay pr
4633 Format: integer between 0 and 10
4636 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
4637 xtime_lock contention on larger systems, and/or RCU lock
4638 contention on all systems with CONFIG_MAXSMP set.
4639 Format: { "0" | "1" }
4640 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
4642 Note: increases power consumption, thus should only be
4643 enabled if running jitter sensitive (HPC/RT) workloads.
4645 security= [SECURITY] Choose a legacy "major" security module to
4646 enable at boot. This has been deprecated by the
4649 selinux= [SELINUX] Disable or enable SELinux at boot time.
4650 Format: { "0" | "1" }
4651 See security/selinux/Kconfig help text.
4656 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
4657 Format: { "0" | "1" }
4658 See security/apparmor/Kconfig help text
4661 Default value is set via kernel config option.
4663 serialnumber [BUGS=X86-32]
4666 Maximal number of shapers.
4674 Disable merging of slabs with similar size. May be
4675 necessary if there is some reason to distinguish
4676 allocs to different slabs, especially in hardened
4677 environments where the risk of heap overflows and
4678 layout control by attackers can usually be
4679 frustrated by disabling merging. This will reduce
4680 most of the exposure of a heap attack to a single
4681 cache (risks via metadata attacks are mostly
4682 unchanged). Debug options disable merging on their
4684 For more information see Documentation/vm/slub.rst.
4686 slab_max_order= [MM, SLAB]
4687 Determines the maximum allowed order for slabs.
4688 A high setting may cause OOMs due to memory
4689 fragmentation. Defaults to 1 for systems with
4690 more than 32MB of RAM, 0 otherwise.
4692 slub_debug[=options[,slabs]] [MM, SLUB]
4693 Enabling slub_debug allows one to determine the
4694 culprit if slab objects become corrupted. Enabling
4695 slub_debug can create guard zones around objects and
4696 may poison objects when not in use. Also tracks the
4697 last alloc / free. For more information see
4698 Documentation/vm/slub.rst.
4700 slub_memcg_sysfs= [MM, SLUB]
4701 Determines whether to enable sysfs directories for
4702 memory cgroup sub-caches. 1 to enable, 0 to disable.
4703 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
4704 Enabling this can lead to a very high number of debug
4705 directories and files being created under
4708 slub_max_order= [MM, SLUB]
4709 Determines the maximum allowed order for slabs.
4710 A high setting may cause OOMs due to memory
4711 fragmentation. For more information see
4712 Documentation/vm/slub.rst.
4714 slub_min_objects= [MM, SLUB]
4715 The minimum number of objects per slab. SLUB will
4716 increase the slab order up to slub_max_order to
4717 generate a sufficiently large slab able to contain
4718 the number of objects indicated. The higher the number
4719 of objects the smaller the overhead of tracking slabs
4720 and the less frequently locks need to be acquired.
4721 For more information see Documentation/vm/slub.rst.
4723 slub_min_order= [MM, SLUB]
4724 Determines the minimum page order for slabs. Must be
4725 lower than slub_max_order.
4726 For more information see Documentation/vm/slub.rst.
4728 slub_nomerge [MM, SLUB]
4729 Same with slab_nomerge. This is supported for legacy.
4730 See slab_nomerge for more information.
4733 Format: <io1>[,<io2>[,...,<io8>]]
4735 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
4736 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4737 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4738 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4739 smsc-ircc2.ircc_irq= [HW] IRQ line
4740 smsc-ircc2.ircc_dma= [HW] DMA channel
4741 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4742 0: Toshiba Satellite 1800 (GP data pin select)
4743 1: Fast pin select (default)
4746 smt [KNL,S390] Set the maximum number of threads (logical
4747 CPUs) to use per physical CPU on systems capable of
4748 symmetric multithreading (SMT). Will be capped to the
4749 actual hardware limit.
4751 Default: -1 (no limit)
4754 [KNL] Should the soft-lockup detector generate panics.
4757 A value of 1 instructs the soft-lockup detector
4758 to panic the machine when a soft-lockup occurs. It is
4759 also controlled by the kernel.softlockup_panic sysctl
4760 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the
4761 respective build-time switch to that functionality.
4763 softlockup_all_cpu_backtrace=
4764 [KNL] Should the soft-lockup detector generate
4765 backtraces on all cpus.
4768 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4769 See Documentation/admin-guide/laptops/sonypi.rst
4771 spectre_v2= [X86] Control mitigation of Spectre variant 2
4772 (indirect branch speculation) vulnerability.
4773 The default operation protects the kernel from
4776 on - unconditionally enable, implies
4778 off - unconditionally disable, implies
4780 auto - kernel detects whether your CPU model is
4783 Selecting 'on' will, and 'auto' may, choose a
4784 mitigation method at run time according to the
4785 CPU, the available microcode, the setting of the
4786 CONFIG_RETPOLINE configuration option, and the
4787 compiler with which the kernel was built.
4789 Selecting 'on' will also enable the mitigation
4790 against user space to user space task attacks.
4792 Selecting 'off' will disable both the kernel and
4793 the user space protections.
4795 Specific mitigations can also be selected manually:
4797 retpoline - replace indirect branches
4798 retpoline,generic - google's original retpoline
4799 retpoline,amd - AMD-specific minimal thunk
4801 Not specifying this option is equivalent to
4805 [X86] Control mitigation of Spectre variant 2
4806 (indirect branch speculation) vulnerability between
4809 on - Unconditionally enable mitigations. Is
4810 enforced by spectre_v2=on
4812 off - Unconditionally disable mitigations. Is
4813 enforced by spectre_v2=off
4815 prctl - Indirect branch speculation is enabled,
4816 but mitigation can be enabled via prctl
4817 per thread. The mitigation control state
4818 is inherited on fork.
4821 - Like "prctl" above, but only STIBP is
4822 controlled per thread. IBPB is issued
4823 always when switching between different user
4827 - Same as "prctl" above, but all seccomp
4828 threads will enable the mitigation unless
4829 they explicitly opt out.
4832 - Like "seccomp" above, but only STIBP is
4833 controlled per thread. IBPB is issued
4834 always when switching between different
4835 user space processes.
4837 auto - Kernel selects the mitigation depending on
4838 the available CPU features and vulnerability.
4841 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4843 Not specifying this option is equivalent to
4844 spectre_v2_user=auto.
4846 spec_store_bypass_disable=
4847 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4848 (Speculative Store Bypass vulnerability)
4850 Certain CPUs are vulnerable to an exploit against a
4851 a common industry wide performance optimization known
4852 as "Speculative Store Bypass" in which recent stores
4853 to the same memory location may not be observed by
4854 later loads during speculative execution. The idea
4855 is that such stores are unlikely and that they can
4856 be detected prior to instruction retirement at the
4857 end of a particular speculation execution window.
4859 In vulnerable processors, the speculatively forwarded
4860 store can be used in a cache side channel attack, for
4861 example to read memory to which the attacker does not
4862 directly have access (e.g. inside sandboxed code).
4864 This parameter controls whether the Speculative Store
4865 Bypass optimization is used.
4867 On x86 the options are:
4869 on - Unconditionally disable Speculative Store Bypass
4870 off - Unconditionally enable Speculative Store Bypass
4871 auto - Kernel detects whether the CPU model contains an
4872 implementation of Speculative Store Bypass and
4873 picks the most appropriate mitigation. If the
4874 CPU is not vulnerable, "off" is selected. If the
4875 CPU is vulnerable the default mitigation is
4876 architecture and Kconfig dependent. See below.
4877 prctl - Control Speculative Store Bypass per thread
4878 via prctl. Speculative Store Bypass is enabled
4879 for a process by default. The state of the control
4880 is inherited on fork.
4881 seccomp - Same as "prctl" above, but all seccomp threads
4882 will disable SSB unless they explicitly opt out.
4884 Default mitigations:
4885 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4887 On powerpc the options are:
4889 on,auto - On Power8 and Power9 insert a store-forwarding
4890 barrier on kernel entry and exit. On Power7
4891 perform a software flush on kernel entry and
4895 Not specifying this option is equivalent to
4896 spec_store_bypass_disable=auto.
4898 spia_io_base= [HW,MTD]
4904 [X86] Enable split lock detection
4906 When enabled (and if hardware support is present), atomic
4907 instructions that access data across cache line
4908 boundaries will result in an alignment check exception.
4912 warn - the kernel will emit rate limited warnings
4913 about applications triggering the #AC
4914 exception. This mode is the default on CPUs
4915 that supports split lock detection.
4917 fatal - the kernel will send SIGBUS to applications
4918 that trigger the #AC exception.
4920 If an #AC exception is hit in the kernel or in
4921 firmware (i.e. not while executing in user mode)
4922 the kernel will oops in either "warn" or "fatal"
4926 Control the Special Register Buffer Data Sampling
4929 Certain CPUs are vulnerable to an MDS-like
4930 exploit which can leak bits from the random
4933 By default, this issue is mitigated by
4934 microcode. However, the microcode fix can cause
4935 the RDRAND and RDSEED instructions to become
4936 much slower. Among other effects, this will
4937 result in reduced throughput from /dev/urandom.
4939 The microcode mitigation can be disabled with
4940 the following option:
4942 off: Disable mitigation and remove
4943 performance impact to RDRAND and RDSEED
4945 srcutree.counter_wrap_check [KNL]
4946 Specifies how frequently to check for
4947 grace-period sequence counter wrap for the
4948 srcu_data structure's ->srcu_gp_seq_needed field.
4949 The greater the number of bits set in this kernel
4950 parameter, the less frequently counter wrap will
4951 be checked for. Note that the bottom two bits
4954 srcutree.exp_holdoff [KNL]
4955 Specifies how many nanoseconds must elapse
4956 since the end of the last SRCU grace period for
4957 a given srcu_struct until the next normal SRCU
4958 grace period will be considered for automatic
4959 expediting. Set to zero to disable automatic
4963 Speculative Store Bypass Disable control
4965 On CPUs that are vulnerable to the Speculative
4966 Store Bypass vulnerability and offer a
4967 firmware based mitigation, this parameter
4968 indicates how the mitigation should be used:
4970 force-on: Unconditionally enable mitigation for
4971 for both kernel and userspace
4972 force-off: Unconditionally disable mitigation for
4973 for both kernel and userspace
4974 kernel: Always enable mitigation in the
4975 kernel, and offer a prctl interface
4976 to allow userspace to register its
4977 interest in being mitigated too.
4979 stack_guard_gap= [MM]
4980 override the default stack gap protection. The value
4981 is in page units and it defines how many pages prior
4982 to (for stacks growing down) resp. after (for stacks
4983 growing up) the main stack are reserved for no other
4984 mapping. Default value is 256 pages.
4987 Enabled the stack tracer on boot up.
4989 stacktrace_filter=[function-list]
4990 [FTRACE] Limit the functions that the stack tracer
4991 will trace at boot up. function-list is a comma separated
4992 list of functions. This list can be changed at run
4993 time by the stack_trace_filter file in the debugfs
4994 tracing directory. Note, this enables stack tracing
4995 and the stacktrace above is not needed.
4999 Set the STI (builtin display/keyboard on the HP-PARISC
5000 machines) console (graphic card) which should be used
5001 as the initial boot-console.
5002 See also comment in drivers/video/console/sticore.c.
5005 See comment in drivers/video/console/sticore.c.
5008 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
5010 sunrpc.min_resvport=
5011 sunrpc.max_resvport=
5013 SunRPC servers often require that client requests
5014 originate from a privileged port (i.e. a port in the
5015 range 0 < portnr < 1024).
5016 An administrator who wishes to reserve some of these
5017 ports for other uses may adjust the range that the
5018 kernel's sunrpc client considers to be privileged
5019 using these two parameters to set the minimum and
5020 maximum port values.
5022 sunrpc.svc_rpc_per_connection_limit=
5024 Limit the number of requests that the server will
5025 process in parallel from a single connection.
5026 The default value is 0 (no limit).
5030 Control how the NFS server code allocates CPUs to
5031 service thread pools. Depending on how many NICs
5032 you have and where their interrupts are bound, this
5033 option will affect which CPUs will do NFS serving.
5034 Note: this parameter cannot be changed while the
5035 NFS server is running.
5037 auto the server chooses an appropriate mode
5038 automatically using heuristics
5039 global a single global pool contains all CPUs
5040 percpu one pool for each CPU
5041 pernode one pool for each NUMA node (equivalent
5042 to global on non-NUMA machines)
5044 sunrpc.tcp_slot_table_entries=
5045 sunrpc.udp_slot_table_entries=
5047 Sets the upper limit on the number of simultaneous
5048 RPC calls that can be sent from the client to a
5049 server. Increasing these values may allow you to
5050 improve throughput, but will also increase the
5051 amount of memory reserved for use by the client.
5053 suspend.pm_test_delay=
5055 Sets the number of seconds to remain in a suspend test
5056 mode before resuming the system (see
5057 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
5058 is set. Default value is 5.
5061 Format: { on | off | y | n | 1 | 0 }
5062 This parameter controls use of the Protected
5063 Execution Facility on pSeries.
5066 [KNL] Enable accounting of swap in memory resource
5067 controller if no parameter or 1 is given or disable
5068 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst)
5070 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
5071 Format: { <int> | force | noforce }
5072 <int> -- Number of I/O TLB slabs
5073 force -- force using of bounce buffers even if they
5074 wouldn't be automatically used by the kernel
5075 noforce -- Never use bounce buffers (for debugging)
5080 Set a sysctl parameter, right before loading the init
5081 process, as if the value was written to the respective
5082 /proc/sys/... file. Both '.' and '/' are recognized as
5083 separators. Unrecognized parameters and invalid values
5084 are reported in the kernel log. Sysctls registered
5085 later by a loaded module cannot be set this way.
5086 Example: sysctl.vm.swappiness=40
5088 sysfs.deprecated=0|1 [KNL]
5089 Enable/disable old style sysfs layout for old udev
5090 on older distributions. When this option is enabled
5091 very new udev will not work anymore. When this option
5092 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
5093 in older udev will not work anymore.
5094 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
5095 the kernel configuration.
5097 sysrq_always_enabled
5099 Ignore sysrq setting - this boot parameter will
5100 neutralize any effect of /proc/sys/kernel/sysrq.
5101 Useful for debugging.
5103 tcpmhash_entries= [KNL,NET]
5104 Set the number of tcp_metrics_hash slots.
5105 Default value is 8192 or 16384 depending on total
5106 ram pages. This is used to specify the TCP metrics
5107 cache size. See Documentation/networking/ip-sysctl.rst
5108 "tcp_no_metrics_save" section for more details.
5112 test_suspend= [SUSPEND][,N]
5113 Specify "mem" (for Suspend-to-RAM) or "standby" (for
5114 standby suspend) or "freeze" (for suspend type freeze)
5115 as the system sleep state during system startup with
5116 the optional capability to repeat N number of times.
5117 The system is woken from this state using a
5118 wakeup-capable RTC alarm.
5120 thash_entries= [KNL,NET]
5121 Set number of hash buckets for TCP connection
5123 thermal.act= [HW,ACPI]
5124 -1: disable all active trip points in all thermal zones
5125 <degrees C>: override all lowest active trip points
5127 thermal.crt= [HW,ACPI]
5128 -1: disable all critical trip points in all thermal zones
5129 <degrees C>: override all critical trip points
5131 thermal.nocrt= [HW,ACPI]
5132 Set to disable actions on ACPI thermal zone
5133 critical and hot trip points.
5135 thermal.off= [HW,ACPI]
5136 1: disable ACPI thermal control
5138 thermal.psv= [HW,ACPI]
5139 -1: disable all passive trip points
5140 <degrees C>: override all passive trip points to this
5143 thermal.tzp= [HW,ACPI]
5144 Specify global default ACPI thermal zone polling rate
5145 <deci-seconds>: poll all this frequency
5146 0: no polling (default)
5149 Force threading of all interrupt handlers except those
5150 marked explicitly IRQF_NO_THREAD.
5154 Specify if the kernel should make use of the cpu
5155 topology information if the hardware supports this.
5156 The scheduler will make use of this information and
5157 e.g. base its process migration decisions on it.
5160 topology_updates= [KNL, PPC, NUMA]
5162 Specify if the kernel should ignore (off)
5163 topology updates sent by the hypervisor to this
5166 torture.disable_onoff_at_boot= [KNL]
5167 Prevent the CPU-hotplug component of torturing
5168 until after init has spawned.
5170 torture.ftrace_dump_at_shutdown= [KNL]
5171 Dump the ftrace buffer at torture-test shutdown,
5172 even if there were no errors. This can be a
5173 very costly operation when many torture tests
5174 are running concurrently, especially on systems
5175 with rotating-rust storage.
5179 tpm_suspend_pcr=[HW,TPM]
5180 Format: integer pcr id
5181 Specify that at suspend time, the tpm driver
5182 should extend the specified pcr with zeros,
5183 as a workaround for some chips which fail to
5184 flush the last written pcr on TPM_SaveState.
5185 This will guarantee that all the other pcrs
5188 trace_buf_size=nn[KMG]
5189 [FTRACE] will set tracing buffer size on each cpu.
5191 trace_event=[event-list]
5192 [FTRACE] Set and start specified trace events in order
5193 to facilitate early boot debugging. The event-list is a
5194 comma separated list of trace events to enable. See
5195 also Documentation/trace/events.rst
5197 trace_options=[option-list]
5198 [FTRACE] Enable or disable tracer options at boot.
5199 The option-list is a comma delimited list of options
5200 that can be enabled or disabled just as if you were
5201 to echo the option name into
5203 /sys/kernel/debug/tracing/trace_options
5205 For example, to enable stacktrace option (to dump the
5206 stack trace of each event), add to the command line:
5208 trace_options=stacktrace
5210 See also Documentation/trace/ftrace.rst "trace options"
5214 Have the tracepoints sent to printk as well as the
5215 tracing ring buffer. This is useful for early boot up
5216 where the system hangs or reboots and does not give the
5217 option for reading the tracing buffer or performing a
5218 ftrace_dump_on_oops.
5220 To turn off having tracepoints sent to printk,
5221 echo 0 > /proc/sys/kernel/tracepoint_printk
5222 Note, echoing 1 into this file without the
5223 tracepoint_printk kernel cmdline option has no effect.
5227 Having tracepoints sent to printk() and activating high
5228 frequency tracepoints such as irq or sched, can cause
5229 the system to live lock.
5232 [FTRACE] enable this option to disable tracing when a
5233 warning is hit. This turns off "tracing_on". Tracing can
5234 be enabled again by echoing '1' into the "tracing_on"
5235 file located in /sys/kernel/debug/tracing/
5237 This option is useful, as it disables the trace before
5238 the WARNING dump is called, which prevents the trace to
5239 be filled with content caused by the warning output.
5241 This option can also be set at run time via the sysctl
5242 option: kernel/traceoff_on_warning
5244 transparent_hugepage=
5246 Format: [always|madvise|never]
5247 Can be used to control the default behavior of the system
5248 with respect to transparent hugepages.
5249 See Documentation/admin-guide/mm/transhuge.rst
5252 tsc= Disable clocksource stability checks for TSC.
5254 [x86] reliable: mark tsc clocksource as reliable, this
5255 disables clocksource verification at runtime, as well
5256 as the stability checks done at bootup. Used to enable
5257 high-resolution timer mode on older hardware, and in
5258 virtualized environment.
5259 [x86] noirqtime: Do not use TSC to do irq accounting.
5260 Used to run time disable IRQ_TIME_ACCOUNTING on any
5261 platforms where RDTSC is slow and this accounting
5263 [x86] unstable: mark the TSC clocksource as unstable, this
5264 marks the TSC unconditionally unstable at bootup and
5265 avoids any further wobbles once the TSC watchdog notices.
5266 [x86] nowatchdog: disable clocksource watchdog. Used
5267 in situations with strict latency requirements (where
5268 interruptions from clocksource watchdog are not
5271 tsc_early_khz= [X86] Skip early TSC calibration and use the given
5272 value instead. Useful when the early TSC frequency discovery
5273 procedure is not reliable, such as on overclocked systems
5274 with CPUID.16h support and partial CPUID.15h support.
5275 Format: <unsigned int>
5277 tsx= [X86] Control Transactional Synchronization
5278 Extensions (TSX) feature in Intel processors that
5279 support TSX control.
5281 This parameter controls the TSX feature. The options are:
5283 on - Enable TSX on the system. Although there are
5284 mitigations for all known security vulnerabilities,
5285 TSX has been known to be an accelerator for
5286 several previous speculation-related CVEs, and
5287 so there may be unknown security risks associated
5288 with leaving it enabled.
5290 off - Disable TSX on the system. (Note that this
5291 option takes effect only on newer CPUs which are
5292 not vulnerable to MDS, i.e., have
5293 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
5294 the new IA32_TSX_CTRL MSR through a microcode
5295 update. This new MSR allows for the reliable
5296 deactivation of the TSX functionality.)
5298 auto - Disable TSX if X86_BUG_TAA is present,
5299 otherwise enable TSX on the system.
5301 Not specifying this option is equivalent to tsx=off.
5303 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5306 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
5307 Abort (TAA) vulnerability.
5309 Similar to Micro-architectural Data Sampling (MDS)
5310 certain CPUs that support Transactional
5311 Synchronization Extensions (TSX) are vulnerable to an
5312 exploit against CPU internal buffers which can forward
5313 information to a disclosure gadget under certain
5316 In vulnerable processors, the speculatively forwarded
5317 data can be used in a cache side channel attack, to
5318 access data to which the attacker does not have direct
5321 This parameter controls the TAA mitigation. The
5324 full - Enable TAA mitigation on vulnerable CPUs
5327 full,nosmt - Enable TAA mitigation and disable SMT on
5328 vulnerable CPUs. If TSX is disabled, SMT
5329 is not disabled because CPU is not
5330 vulnerable to cross-thread TAA attacks.
5331 off - Unconditionally disable TAA mitigation
5333 On MDS-affected machines, tsx_async_abort=off can be
5334 prevented by an active MDS mitigation as both vulnerabilities
5335 are mitigated with the same mechanism so in order to disable
5336 this mitigation, you need to specify mds=off too.
5338 Not specifying this option is equivalent to
5339 tsx_async_abort=full. On CPUs which are MDS affected
5340 and deploy MDS mitigation, TAA mitigation is not
5341 required and doesn't provide any additional
5345 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
5347 turbografx.map[2|3]= [HW,JOY]
5348 TurboGraFX parallel port interface
5350 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
5351 See also Documentation/input/devices/joystick-parport.rst
5353 udbg-immortal [PPC] When debugging early kernel crashes that
5354 happen after console_init() and before a proper
5355 console driver takes over, this boot options might
5356 help "seeing" what's going on.
5358 uhash_entries= [KNL,NET]
5359 Set number of hash buckets for UDP/UDP-Lite connections
5362 [USB] Ignore overcurrent events (default N).
5363 Some badly-designed motherboards generate lots of
5364 bogus events, for ports that aren't wired to
5365 anything. Set this parameter to avoid log spamming.
5366 Note that genuine overcurrent events won't be
5370 [X86] Cause panic on unknown NMI.
5372 usbcore.authorized_default=
5373 [USB] Default USB device authorization:
5374 (default -1 = authorized except for wireless USB,
5375 0 = not authorized, 1 = authorized, 2 = authorized
5376 if device connected to internal port)
5378 usbcore.autosuspend=
5379 [USB] The autosuspend time delay (in seconds) used
5380 for newly-detected USB devices (default 2). This
5381 is the time required before an idle device will be
5382 autosuspended. Devices for which the delay is set
5383 to a negative value won't be autosuspended at all.
5385 usbcore.usbfs_snoop=
5386 [USB] Set to log all usbfs traffic (default 0 = off).
5388 usbcore.usbfs_snoop_max=
5389 [USB] Maximum number of bytes to snoop in each URB
5392 usbcore.blinkenlights=
5393 [USB] Set to cycle leds on hubs (default 0 = off).
5395 usbcore.old_scheme_first=
5396 [USB] Start with the old device initialization
5397 scheme (default 0 = off).
5399 usbcore.usbfs_memory_mb=
5400 [USB] Memory limit (in MB) for buffers allocated by
5401 usbfs (default = 16, 0 = max = 2047).
5403 usbcore.use_both_schemes=
5404 [USB] Try the other device initialization scheme
5405 if the first one fails (default 1 = enabled).
5407 usbcore.initial_descriptor_timeout=
5408 [USB] Specifies timeout for the initial 64-byte
5409 USB_REQ_GET_DESCRIPTOR request in milliseconds
5410 (default 5000 = 5.0 seconds).
5412 usbcore.nousb [USB] Disable the USB subsystem
5415 [USB] A list of quirk entries to augment the built-in
5416 usb core quirk list. List entries are separated by
5417 commas. Each entry has the form
5418 VendorID:ProductID:Flags. The IDs are 4-digit hex
5419 numbers and Flags is a set of letters. Each letter
5420 will change the built-in quirk; setting it if it is
5421 clear and clearing it if it is set. The letters have
5422 the following meanings:
5423 a = USB_QUIRK_STRING_FETCH_255 (string
5424 descriptors must not be fetched using
5426 b = USB_QUIRK_RESET_RESUME (device can't resume
5427 correctly so reset it instead);
5428 c = USB_QUIRK_NO_SET_INTF (device can't handle
5429 Set-Interface requests);
5430 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
5431 handle its Configuration or Interface
5433 e = USB_QUIRK_RESET (device can't be reset
5434 (e.g morph devices), don't use reset);
5435 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
5436 more interface descriptions than the
5437 bNumInterfaces count, and can't handle
5438 talking to these interfaces);
5439 g = USB_QUIRK_DELAY_INIT (device needs a pause
5440 during initialization, after we read
5441 the device descriptor);
5442 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
5443 high speed and super speed interrupt
5444 endpoints, the USB 2.0 and USB 3.0 spec
5445 require the interval in microframes (1
5446 microframe = 125 microseconds) to be
5447 calculated as interval = 2 ^
5449 Devices with this quirk report their
5450 bInterval as the result of this
5451 calculation instead of the exponent
5452 variable used in the calculation);
5453 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
5454 handle device_qualifier descriptor
5456 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
5457 generates spurious wakeup, ignore
5458 remote wakeup capability);
5459 k = USB_QUIRK_NO_LPM (device can't handle Link
5461 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
5462 (Device reports its bInterval as linear
5463 frames instead of the USB 2.0
5465 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
5466 to be disconnected before suspend to
5467 prevent spurious wakeup);
5468 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
5469 pause after every control message);
5470 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
5471 delay after resetting its port);
5472 Example: quirks=0781:5580:bk,0a5c:5834:gij
5475 [USBHID] The interval which mice are to be polled at.
5478 [USBHID] The interval which joysticks are to be polled at.
5481 [USBHID] The interval which keyboards are to be polled at.
5483 usb-storage.delay_use=
5484 [UMS] The delay in seconds before a new device is
5485 scanned for Logical Units (default 1).
5488 [UMS] A list of quirks entries to supplement or
5489 override the built-in unusual_devs list. List
5490 entries are separated by commas. Each entry has
5491 the form VID:PID:Flags where VID and PID are Vendor
5492 and Product ID values (4-digit hex numbers) and
5493 Flags is a set of characters, each corresponding
5494 to a common usb-storage quirk flag as follows:
5495 a = SANE_SENSE (collect more than 18 bytes
5496 of sense data, not on uas);
5497 b = BAD_SENSE (don't collect more than 18
5498 bytes of sense data, not on uas);
5499 c = FIX_CAPACITY (decrease the reported
5500 device capacity by one sector);
5501 d = NO_READ_DISC_INFO (don't use
5502 READ_DISC_INFO command, not on uas);
5503 e = NO_READ_CAPACITY_16 (don't use
5504 READ_CAPACITY_16 command);
5505 f = NO_REPORT_OPCODES (don't use report opcodes
5507 g = MAX_SECTORS_240 (don't transfer more than
5508 240 sectors at a time, uas only);
5509 h = CAPACITY_HEURISTICS (decrease the
5510 reported device capacity by one
5511 sector if the number is odd);
5512 i = IGNORE_DEVICE (don't bind to this
5514 j = NO_REPORT_LUNS (don't use report luns
5516 l = NOT_LOCKABLE (don't try to lock and
5517 unlock ejectable media, not on uas);
5518 m = MAX_SECTORS_64 (don't transfer more
5519 than 64 sectors = 32 KB at a time,
5521 n = INITIAL_READ10 (force a retry of the
5522 initial READ(10) command, not on uas);
5523 o = CAPACITY_OK (accept the capacity
5524 reported by the device, not on uas);
5525 p = WRITE_CACHE (the device cache is ON
5526 by default, not on uas);
5527 r = IGNORE_RESIDUE (the device reports
5528 bogus residue values, not on uas);
5529 s = SINGLE_LUN (the device has only one
5531 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
5532 commands, uas only);
5533 u = IGNORE_UAS (don't bind to the uas driver);
5534 w = NO_WP_DETECT (don't test whether the
5535 medium is write-protected).
5536 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
5537 even if the device claims no cache,
5539 Example: quirks=0419:aaf5:rl,0421:0433:rc
5541 user_debug= [KNL,ARM]
5543 See arch/arm/Kconfig.debug help text.
5544 1 - undefined instruction events
5546 4 - invalid data aborts
5549 Example: user_debug=31
5552 [X86] Flags controlling user PTE allocations.
5554 nohigh = do not allocate PTE pages in
5555 HIGHMEM regardless of setting
5559 On X86_32, this is an alias for vdso32=. Otherwise:
5561 vdso=1: enable VDSO (the default)
5562 vdso=0: disable VDSO mapping
5564 vdso32= [X86] Control the 32-bit vDSO
5565 vdso32=1: enable 32-bit VDSO
5566 vdso32=0 or vdso32=2: disable 32-bit VDSO
5568 See the help text for CONFIG_COMPAT_VDSO for more
5569 details. If CONFIG_COMPAT_VDSO is set, the default is
5570 vdso32=0; otherwise, the default is vdso32=1.
5572 For compatibility with older kernels, vdso32=2 is an
5575 Try vdso32=0 if you encounter an error that says:
5576 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
5579 vector=percpu: enable percpu vector domain
5581 video= [FB] Frame buffer configuration
5582 See Documentation/fb/modedb.rst.
5584 video.brightness_switch_enabled= [0,1]
5585 If set to 1, on receiving an ACPI notify event
5586 generated by hotkey, video driver will adjust brightness
5587 level and then send out the event to user space through
5588 the allocated input device; If set to 0, video driver
5589 will only send out the event without touching backlight
5594 [VMMIO] Memory mapped virtio (platform) device.
5596 <size>@<baseaddr>:<irq>[:<id>]
5598 <size> := size (can use standard suffixes
5600 <baseaddr> := physical base address
5601 <irq> := interrupt number (as passed to
5603 <id> := (optional) platform device id
5605 virtio_mmio.device=1K@0x100b0000:48:7
5607 Can be used multiple times for multiple devices.
5609 vga= [BOOT,X86-32] Select a particular video mode
5610 See Documentation/x86/boot.rst and
5611 Documentation/admin-guide/svga.rst.
5612 Use vga=ask for menu.
5613 This is actually a boot loader parameter; the value is
5614 passed to the kernel using a special protocol.
5616 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
5617 May slow down system boot speed, especially when
5618 enabled on systems with a large amount of memory.
5619 All options are enabled by default, and this
5620 interface is meant to allow for selectively
5621 enabling or disabling specific virtual memory
5624 Available options are:
5625 P Enable page structure init time poisoning
5626 - Disable all of the above options
5628 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
5629 size of <nn>. This can be used to increase the
5630 minimum size (128MB on x86). It can also be used to
5631 decrease the size and leave more room for directly
5634 vmcp_cma=nn[MG] [KNL,S390]
5635 Sets the memory size reserved for contiguous memory
5636 allocations for the vmcp device driver.
5638 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
5641 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
5644 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
5648 Controls the behavior of vsyscalls (i.e. calls to
5649 fixed addresses of 0xffffffffff600x00 from legacy
5650 code). Most statically-linked binaries and older
5651 versions of glibc use these calls. Because these
5652 functions are at fixed addresses, they make nice
5653 targets for exploits that can control RIP.
5655 emulate [default] Vsyscalls turn into traps and are
5656 emulated reasonably safely. The vsyscall
5659 xonly Vsyscalls turn into traps and are
5660 emulated reasonably safely. The vsyscall
5661 page is not readable.
5663 none Vsyscalls don't work at all. This makes
5664 them quite hard to use for exploits but
5665 might break your system.
5667 vt.color= [VT] Default text color.
5668 Format: 0xYX, X = foreground, Y = background.
5669 Default: 0x07 = light gray on black.
5671 vt.cur_default= [VT] Default cursor shape.
5672 Format: 0xCCBBAA, where AA, BB, and CC are the same as
5673 the parameters of the <Esc>[?A;B;Cc escape sequence;
5674 see VGA-softcursor.txt. Default: 2 = underline.
5676 vt.default_blu= [VT]
5677 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
5678 Change the default blue palette of the console.
5679 This is a 16-member array composed of values
5682 vt.default_grn= [VT]
5683 Format: <green0>,<green1>,<green2>,...,<green15>
5684 Change the default green palette of the console.
5685 This is a 16-member array composed of values
5688 vt.default_red= [VT]
5689 Format: <red0>,<red1>,<red2>,...,<red15>
5690 Change the default red palette of the console.
5691 This is a 16-member array composed of values
5697 Set system-wide default UTF-8 mode for all tty's.
5698 Default is 1, i.e. UTF-8 mode is enabled for all
5699 newly opened terminals.
5701 vt.global_cursor_default=
5704 Set system-wide default for whether a cursor
5705 is shown on new VTs. Default is -1,
5706 i.e. cursors will be created by default unless
5707 overridden by individual drivers. 0 will hide
5708 cursors, 1 will display them.
5710 vt.italic= [VT] Default color for italic text; 0-15.
5713 vt.underline= [VT] Default color for underlined text; 0-15.
5716 watchdog timers [HW,WDT] For information on watchdog timers,
5717 see Documentation/watchdog/watchdog-parameters.rst
5718 or other driver-specific files in the
5719 Documentation/watchdog/ directory.
5723 Set the hard lockup detector stall duration
5724 threshold in seconds. The soft lockup detector
5725 threshold is set to twice the value. A value of 0
5726 disables both lockup detectors. Default is 10
5729 workqueue.watchdog_thresh=
5730 If CONFIG_WQ_WATCHDOG is configured, workqueue can
5731 warn stall conditions and dump internal state to
5732 help debugging. 0 disables workqueue stall
5733 detection; otherwise, it's the stall threshold
5734 duration in seconds. The default value is 30 and
5735 it can be updated at runtime by writing to the
5736 corresponding sysfs file.
5738 workqueue.disable_numa
5739 By default, all work items queued to unbound
5740 workqueues are affine to the NUMA nodes they're
5741 issued on, which results in better behavior in
5742 general. If NUMA affinity needs to be disabled for
5743 whatever reason, this option can be used. Note
5744 that this also can be controlled per-workqueue for
5745 workqueues visible under /sys/bus/workqueue/.
5747 workqueue.power_efficient
5748 Per-cpu workqueues are generally preferred because
5749 they show better performance thanks to cache
5750 locality; unfortunately, per-cpu workqueues tend to
5751 be more power hungry than unbound workqueues.
5753 Enabling this makes the per-cpu workqueues which
5754 were observed to contribute significantly to power
5755 consumption unbound, leading to measurably lower
5756 power usage at the cost of small performance
5759 The default value of this parameter is determined by
5760 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
5762 workqueue.debug_force_rr_cpu
5763 Workqueue used to implicitly guarantee that work
5764 items queued without explicit CPU specified are put
5765 on the local CPU. This guarantee is no longer true
5766 and while local CPU is still preferred work items
5767 may be put on foreign CPUs. This debug option
5768 forces round-robin CPU selection to flush out
5769 usages which depend on the now broken guarantee.
5770 When enabled, memory and cache locality will be
5773 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
5774 default x2apic cluster mode on platforms
5777 x86_intel_mid_timer= [X86-32,APBT]
5778 Choose timer option for x86 Intel MID platform.
5779 Two valid options are apbt timer only and lapic timer
5780 plus one apbt timer for broadcast timer.
5781 x86_intel_mid_timer=apbt_only | lapic_and_apbt
5783 xen_512gb_limit [KNL,X86-64,XEN]
5784 Restricts the kernel running paravirtualized under Xen
5785 to use only up to 512 GB of RAM. The reason to do so is
5786 crash analysis tools and Xen tools for doing domain
5787 save/restore/migration must be enabled to handle larger
5790 xen_emul_unplug= [HW,X86,XEN]
5791 Unplug Xen emulated devices
5792 Format: [unplug0,][unplug1]
5793 ide-disks -- unplug primary master IDE devices
5794 aux-ide-disks -- unplug non-primary-master IDE devices
5795 nics -- unplug network devices
5796 all -- unplug all emulated devices (NICs and IDE disks)
5797 unnecessary -- unplugging emulated devices is
5798 unnecessary even if the host did not respond to
5800 never -- do not unplug even if version check succeeds
5802 xen_legacy_crash [X86,XEN]
5803 Crash from Xen panic notifier, without executing late
5804 panic() code such as dumping handler.
5806 xen_nopvspin [X86,XEN]
5807 Disables the ticketlock slowpath using Xen PV
5811 Disables the PV optimizations forcing the HVM guest to
5812 run as generic HVM guest with no PV drivers.
5813 This option is obsoleted by the "nopv" option, which
5814 has equivalent effect for XEN platform.
5816 xen_scrub_pages= [XEN]
5817 Boolean option to control scrubbing pages before giving them back
5818 to Xen, for use by other domains. Can be also changed at runtime
5819 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
5820 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
5822 xen_timer_slop= [X86-64,XEN]
5823 Set the timer slop (in nanoseconds) for the virtual Xen
5824 timers (default is 100000). This adjusts the minimum
5825 delta of virtualized Xen timers, where lower values
5826 improve timer resolution at the expense of processing
5827 more timer interrupts.
5829 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
5830 Disables the PV optimizations forcing the guest to run
5831 as generic guest with no PV drivers. Currently support
5832 XEN HVM, KVM, HYPER_V and VMWARE guest.
5834 xirc2ps_cs= [NET,PCMCIA]
5836 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
5839 By default on POWER9 and above, the kernel will
5840 natively use the XIVE interrupt controller. This option
5841 allows the fallback firmware mode to be used:
5843 off Fallback to firmware control of XIVE interrupt
5844 controller on both pseries and powernv
5845 platforms. Only useful on POWER9 and above.
5847 xhci-hcd.quirks [USB,KNL]
5848 A hex value specifying bitmask with supplemental xhci
5849 host controller quirks. Meaning of each bit can be
5850 consulted in header drivers/usb/host/xhci.h.
5853 Format: { early | on | rw | ro | off }
5854 Controls if xmon debugger is enabled. Default is off.
5855 Passing only "xmon" is equivalent to "xmon=early".
5856 early Call xmon as early as possible on boot; xmon
5857 debugger is called from setup_arch().
5858 on xmon debugger hooks will be installed so xmon
5859 is only called on a kernel crash. Default mode,
5860 i.e. either "ro" or "rw" mode, is controlled
5861 with CONFIG_XMON_DEFAULT_RO_MODE.
5862 rw xmon debugger hooks will be installed so xmon
5863 is called only on a kernel crash, mode is write,
5864 meaning SPR registers, memory and, other data
5865 can be written using xmon commands.
5866 ro same as "rw" option above but SPR registers,
5867 memory, and other data can't be written using
5869 off xmon is disabled.