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]
27 If set to vendor, prefer vendor specific driver
28 (e.g. thinkpad_acpi, sony_acpi, etc.) instead
29 of the ACPI video.ko driver.
31 acpi_force_32bit_fadt_addr
32 force FADT to use 32 bit addresses rather than the
33 64 bit X_* addresses. Some firmware have broken 64
34 bit addresses for force ACPI ignore these and use
35 the older legacy 32 bit addresses.
37 acpica_no_return_repair [HW, ACPI]
38 Disable AML predefined validation mechanism
39 This mechanism can repair the evaluation result to make
40 the return objects more ACPI specification compliant.
41 This option is useful for developers to identify the
42 root cause of an AML interpreter issue when the issue
43 has something to do with the repair mechanism.
45 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG]
46 acpi.debug_level= [HW,ACPI,ACPI_DEBUG]
48 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI
49 debug output. Bits in debug_layer correspond to a
50 _COMPONENT in an ACPI source file, e.g.,
51 #define _COMPONENT ACPI_PCI_COMPONENT
52 Bits in debug_level correspond to a level in
53 ACPI_DEBUG_PRINT statements, e.g.,
54 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ...
55 The debug_level mask defaults to "info". See
56 Documentation/firmware-guide/acpi/debug.rst for more information about
57 debug layers and levels.
59 Enable processor driver info messages:
60 acpi.debug_layer=0x20000000
61 Enable PCI/PCI interrupt routing info messages:
62 acpi.debug_layer=0x400000
63 Enable AML "Debug" output, i.e., stores to the Debug
64 object while interpreting AML:
65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2
66 Enable all messages related to ACPI hardware:
67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff
69 Some values produce so much output that the system is
70 unusable. The "log_buf_len" parameter may be useful
71 if you need to capture more output.
73 acpi_enforce_resources= [ACPI]
75 Check for resource conflicts between native drivers
76 and ACPI OperationRegions (SystemIO and SystemMemory
77 only). IO ports and memory declared in ACPI might be
78 used by the ACPI subsystem in arbitrary AML code and
79 can interfere with legacy drivers.
80 strict (default): access to resources claimed by ACPI
81 is denied; legacy drivers trying to access reserved
82 resources will fail to bind to device using them.
83 lax: access to resources claimed by ACPI is allowed;
84 legacy drivers trying to access reserved resources
85 will bind successfully but a warning message is logged.
86 no: ACPI OperationRegions are not marked as reserved,
87 no further checks are performed.
89 acpi_force_table_verification [HW,ACPI]
90 Enable table checksum verification during early stage.
91 By default, this is disabled due to x86 early mapping
94 acpi_irq_balance [HW,ACPI]
95 ACPI will balance active IRQs
98 acpi_irq_nobalance [HW,ACPI]
99 ACPI will not move active IRQs (default)
102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA
103 Format: <irq>,<irq>...
105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for
107 Format: <irq>,<irq>...
109 acpi_mask_gpe= [HW,ACPI]
110 Due to the existence of _Lxx/_Exx, some GPEs triggered
111 by unsupported hardware/firmware features can result in
112 GPE floodings that cannot be automatically disabled by
114 This facility can be used to prevent such uncontrolled
118 acpi_no_auto_serialize [HW,ACPI]
119 Disable auto-serialization of AML methods
120 AML control methods that contain the opcodes to create
121 named objects will be marked as "Serialized" by the
122 auto-serialization feature.
123 This feature is enabled by default.
124 This option allows to turn off the feature.
126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump
129 acpi_no_static_ssdt [HW,ACPI]
130 Disable installation of static SSDTs at early boot time
131 By default, SSDTs contained in the RSDT/XSDT will be
132 installed automatically and they will appear under
133 /sys/firmware/acpi/tables.
134 This option turns off this feature.
135 Note that specifying this option does not affect
136 dynamic table installation which will install SSDT
137 tables to /sys/firmware/acpi/tables/dynamic.
139 acpi_rsdp= [ACPI,EFI,KEXEC]
140 Pass the RSDP address to the kernel, mostly used
141 on machines running EFI runtime service to boot the
142 second kernel for kdump.
144 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS
145 Format: To spoof as Windows 98: ="Microsoft Windows"
147 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead
148 of 2 which is mandated by ACPI 6) as the supported ACPI
149 specification revision (when using this switch, it may
150 be necessary to carry out a cold reboot _twice_ in a
151 row to make it take effect on the platform firmware).
153 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings
154 acpi_osi="string1" # add string1
155 acpi_osi="!string2" # remove string2
156 acpi_osi=!* # remove all strings
157 acpi_osi=! # disable all built-in OS vendor
159 acpi_osi=!! # enable all built-in OS vendor
161 acpi_osi= # disable all strings
163 'acpi_osi=!' can be used in combination with single or
164 multiple 'acpi_osi="string1"' to support specific OS
165 vendor string(s). Note that such command can only
166 affect the default state of the OS vendor strings, thus
167 it cannot affect the default state of the feature group
168 strings and the current state of the OS vendor strings,
169 specifying it multiple times through kernel command line
170 is meaningless. This command is useful when one do not
171 care about the state of the feature group strings which
172 should be controlled by the OSPM.
174 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent
175 to 'acpi_osi="Windows 2000" acpi_osi=!', they all
176 can make '_OSI("Windows 2000")' TRUE.
178 'acpi_osi=' cannot be used in combination with other
179 'acpi_osi=' command lines, the _OSI method will not
180 exist in the ACPI namespace. NOTE that such command can
181 only affect the _OSI support state, thus specifying it
182 multiple times through kernel command line is also
185 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)'
188 'acpi_osi=!*' can be used in combination with single or
189 multiple 'acpi_osi="string1"' to support specific
190 string(s). Note that such command can affect the
191 current state of both the OS vendor strings and the
192 feature group strings, thus specifying it multiple times
193 through kernel command line is meaningful. But it may
194 still not able to affect the final state of a string if
195 there are quirks related to this string. This command
196 is useful when one want to control the state of the
197 feature group strings to debug BIOS issues related to
200 1. 'acpi_osi="Module Device" acpi_osi=!*' can make
201 '_OSI("Module Device")' FALSE.
202 2. 'acpi_osi=!* acpi_osi="Module Device"' can make
203 '_OSI("Module Device")' TRUE.
204 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is
206 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"'
208 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!',
209 they all will make '_OSI("Windows 2000")' TRUE.
212 Override the pmtimer bug detection: force the kernel
213 to assume that this machine's pmtimer latches its value
214 and always returns good values.
216 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode
217 Format: { level | edge | high | low }
219 acpi_skip_timer_override [HW,ACPI]
220 Recognize and ignore IRQ0/pin2 Interrupt Override.
221 For broken nForce2 BIOS resulting in XT-PIC timer.
223 acpi_sleep= [HW,ACPI] Sleep options
224 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig,
225 old_ordering, nonvs, sci_force_enable, nobl }
226 See Documentation/power/video.rst for information on
228 s3_beep is for debugging; it makes the PC's speaker beep
229 as soon as the kernel's real-mode entry point is called.
230 s4_nohwsig prevents ACPI hardware signature from being
231 used during resume from hibernation.
232 old_ordering causes the ACPI 1.0 ordering of the _PTS
233 control method, with respect to putting devices into
234 low power states, to be enforced (the ACPI 2.0 ordering
235 of _PTS is used by default).
236 nonvs prevents the kernel from saving/restoring the
237 ACPI NVS memory during suspend/hibernation and resume.
238 sci_force_enable causes the kernel to set SCI_EN directly
239 on resume from S1/S3 (which is against the ACPI spec,
240 but some broken systems don't work without it).
241 nobl causes the internal blacklist of systems known to
242 behave incorrectly in some ways with respect to system
243 suspend and resume to be ignored (use wisely).
245 acpi_use_timer_override [HW,ACPI]
246 Use timer override. For some broken Nvidia NF5 boards
247 that require a timer override, but don't have HPET
249 add_efi_memmap [EFI; X86] Include EFI memory map in
250 kernel's map of available physical RAM.
253 { off | try_unsupported }
254 off: disable AGP support
255 try_unsupported: try to drive unsupported chipsets
256 (may crash computer or cause data corruption)
259 See Documentation/sound/alsa-configuration.rst
262 Allow the default userspace alignment fault handler
263 behaviour to be specified. Bit 0 enables warnings,
264 bit 1 enables fixups, and bit 2 sends a segfault.
266 align_va_addr= [X86-64]
267 Align virtual addresses by clearing slice [14:12] when
268 allocating a VMA at process creation time. This option
269 gives you up to 3% performance improvement on AMD F15h
270 machines (where it is enabled by default) for a
271 CPU-intensive style benchmark, and it can vary highly in
272 a microbenchmark depending on workload and compiler.
274 32: only for 32-bit processes
275 64: only for 64-bit processes
276 on: enable for both 32- and 64-bit processes
277 off: disable for both 32- and 64-bit processes
279 alloc_snapshot [FTRACE]
280 Allocate the ftrace snapshot buffer on boot up when the
281 main buffer is allocated. This is handy if debugging
282 and you need to use tracing_snapshot() on boot up, and
283 do not want to use tracing_snapshot_alloc() as it needs
284 to be done where GFP_KERNEL allocations are allowed.
286 amd_iommu= [HW,X86-64]
287 Pass parameters to the AMD IOMMU driver in the system.
289 fullflush - enable flushing of IO/TLB entries when
290 they are unmapped. Otherwise they are
291 flushed before they will be reused, which
293 off - do not initialize any AMD IOMMU found in
295 force_isolation - Force device isolation for all
296 devices. The IOMMU driver is not
297 allowed anymore to lift isolation
298 requirements as needed. This option
299 does not override iommu=pt
301 amd_iommu_dump= [HW,X86-64]
302 Enable AMD IOMMU driver option to dump the ACPI table
303 for AMD IOMMU. With this option enabled, AMD IOMMU
304 driver will print ACPI tables for AMD IOMMU during
305 IOMMU initialization.
307 amd_iommu_intr= [HW,X86-64]
308 Specifies one of the following AMD IOMMU interrupt
310 legacy - Use legacy interrupt remapping mode.
311 vapic - Use virtual APIC mode, which allows IOMMU
312 to inject interrupts directly into guest.
313 This mode requires kvm-amd.avic=1.
314 (Default when IOMMU HW support is present.)
316 amijoy.map= [HW,JOY] Amiga joystick support
317 Map of devices attached to JOY0DAT and JOY1DAT
319 See also Documentation/input/joydev/joystick.rst
321 analog.map= [HW,JOY] Analog joystick and gamepad support
322 Specifies type or capabilities of an analog joystick
323 connected to one of 16 gameports
324 Format: <type1>,<type2>,..<type16>
327 Power management functions (SPARCstation-4/5 + deriv.)
329 Disable APC CPU standby support. SPARCstation-Fox does
330 not play well with APC CPU idle - disable it if you have
331 APC and your system crashes randomly.
333 apic= [APIC,X86] Advanced Programmable Interrupt Controller
334 Change the output verbosity while booting
335 Format: { quiet (default) | verbose | debug }
336 Change the amount of debugging information output
337 when initialising the APIC and IO-APIC components.
338 For X86-32, this can also be used to specify an APIC
340 Format: apic=driver_name
341 Examples: apic=bigsmp
343 apic_extnmi= [APIC,X86] External NMI delivery setting
344 Format: { bsp (default) | all | none }
345 bsp: External NMI is delivered only to CPU 0
346 all: External NMIs are broadcast to all CPUs as a
348 none: External NMI is masked for all CPUs. This is
349 useful so that a dump capture kernel won't be
353 See Documentation/networking/ipv6.txt.
355 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller
356 Limit apic dumping. The parameter defines the maximal
357 number of local apics being dumped. Also it is possible
358 to set it to "all" by meaning -- no limit here.
359 Format: { 1 (default) | 2 | ... | all }.
360 The parameter valid if only apic=debug or
361 apic=verbose is specified.
362 Example: apic=debug show_lapic=all
364 apm= [APM] Advanced Power Management
365 See header of arch/x86/kernel/apm_32.c.
367 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards
368 Format: <io>,<irq>,<nodeID>
372 atarimouse= [HW,MOUSE] Atari Mouse
374 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess,
375 EzKey and similar keyboards
377 atkbd.reset= [HW] Reset keyboard during initialization
379 atkbd.set= [HW] Select keyboard code set
380 Format: <int> (2 = AT (default), 3 = PS/2)
382 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar
385 atkbd.softraw= [HW] Choose between synthetic and real raw mode
386 Format: <bool> (0 = real, 1 = synthetic (default))
388 atkbd.softrepeat= [HW]
389 Use software keyboard repeat
391 audit= [KNL] Enable the audit sub-system
392 Format: { "0" | "1" | "off" | "on" }
393 0 | off - kernel audit is disabled and can not be
394 enabled until the next reboot
395 unset - kernel audit is initialized but disabled and
396 will be fully enabled by the userspace auditd.
397 1 | on - kernel audit is initialized and partially
398 enabled, storing at most audit_backlog_limit
399 messages in RAM until it is fully enabled by the
403 audit_backlog_limit= [KNL] Set the audit queue size limit.
404 Format: <int> (must be >=0)
407 bau= [X86_UV] Enable the BAU on SGI UV. The default
408 behavior is to disable the BAU (i.e. bau=0).
409 Format: { "0" | "1" }
412 unset - Disable the BAU.
414 baycom_epp= [HW,AX25]
417 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem
419 See header of drivers/net/hamradio/baycom_par.c.
421 baycom_ser_fdx= [HW,AX25]
422 BayCom Serial Port AX.25 Modem (Full Duplex Mode)
423 Format: <io>,<irq>,<mode>[,<baud>]
424 See header of drivers/net/hamradio/baycom_ser_fdx.c.
426 baycom_ser_hdx= [HW,AX25]
427 BayCom Serial Port AX.25 Modem (Half Duplex Mode)
428 Format: <io>,<irq>,<mode>
429 See header of drivers/net/hamradio/baycom_ser_hdx.c.
431 blkdevparts= Manual partition parsing of block device(s) for
432 embedded devices based on command line input.
433 See Documentation/block/cmdline-partition.rst
435 boot_delay= Milliseconds to delay each printk during boot.
436 Values larger than 10 seconds (10000) are changed to
440 bootmem_debug [KNL] Enable bootmem allocator debug messages.
443 Disable BERT OS support on buggy BIOSes.
445 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards)
446 bttv.radio= Most important insmod options are available as
448 bttv.pll= See Documentation/media/v4l-drivers/bttv.rst
451 bulk_remove=off [PPC] This parameter disables the use of the pSeries
452 firmware feature for flushing multiple hpte entries
455 c101= [NET] Moxa C101 synchronous serial card
457 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection.
458 Sometimes CPU hardware bugs make them report the cache
459 size incorrectly. The kernel will attempt work arounds
460 to fix known problems, but for some CPUs it is not
461 possible to determine what the correct size should be.
462 This option provides an override for these situations.
465 [NET] Specifies amount of time (in seconds) that
466 the kernel should wait for a network carrier. By default
467 it waits 120 seconds.
469 ca_keys= [KEYS] This parameter identifies a specific key(s) on
470 the system trusted keyring to be used for certificate
472 format: { id:<keyid> | builtin }
474 cca= [MIPS] Override the kernel pages' cache coherency
475 algorithm. Accepted values range from 0 to 7
476 inclusive. See arch/mips/include/asm/pgtable-bits.h
477 for platform specific values (SB1, Loongson3 and
480 ccw_timeout_log [S390]
481 See Documentation/s390/common_io.rst for details.
483 cgroup_disable= [KNL] Disable a particular controller
484 Format: {name of the controller(s) to disable}
485 The effects of cgroup_disable=foo are:
486 - foo isn't auto-mounted if you mount all cgroups in
488 - foo isn't visible as an individually mountable
490 {Currently only "memory" controller deal with this and
491 cut the overhead, others just disable the usage. So
492 only cgroup_disable=memory is actually worthy}
494 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1
495 Format: { { controller | "all" | "named" }
496 [,{ controller | "all" | "named" }...] }
497 Like cgroup_disable, but only applies to cgroup v1;
498 the blacklisted controllers remain available in cgroup2.
499 "all" blacklists all controllers and "named" disables
500 named mounts. Specifying both "all" and "named" disables
503 cgroup.memory= [KNL] Pass options to the cgroup memory controller.
505 nosocket -- Disable socket memory accounting.
506 nokmem -- Disable kernel memory accounting.
508 checkreqprot [SELINUX] Set initial checkreqprot flag value.
509 Format: { "0" | "1" }
510 See security/selinux/Kconfig help text.
511 0 -- check protection applied by kernel (includes
512 any implied execute protection).
513 1 -- check protection requested by application.
514 Default value is set via a kernel config option.
515 Value can be changed at runtime via
516 /selinux/checkreqprot.
519 See Documentation/s390/common_io.rst for details.
522 Prevents the clock framework from automatically gating
523 clocks that have not been explicitly enabled by a Linux
524 device driver but are enabled in hardware at reset or
525 by the bootloader/firmware. Note that this does not
526 force such clocks to be always-on nor does it reserve
527 those clocks in any way. This parameter is useful for
528 debug and development, but should not be needed on a
529 platform with proper driver support. For more
530 information, see Documentation/driver-api/clk.rst.
532 clock= [BUGS=X86-32, HW] gettimeofday clocksource override.
534 Forces specified clocksource (if available) to be used
535 when calculating gettimeofday(). If specified
536 clocksource is not available, it defaults to PIT.
537 Format: { pit | tsc | cyclone | pmtmr }
539 clocksource= Override the default clocksource
541 Override the default clocksource and use the clocksource
542 with the name specified.
543 Some clocksource names to choose from, depending on
545 [all] jiffies (this is the base, fallback clocksource)
547 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2,
548 pxa_timer,timer3,32k_counter,timer0_1
549 [X86-32] pit,hpet,tsc;
550 scx200_hrt on Geode; cyclone on IBM x440
558 clocksource.arm_arch_timer.evtstrm=
561 Enable/disable the eventstream feature of the ARM
562 architected timer so that code using WFE-based polling
563 loops can be debugged more effectively on production
566 clearcpuid=BITNUM [X86]
567 Disable CPUID feature X for the kernel. See
568 arch/x86/include/asm/cpufeatures.h for the valid bit
569 numbers. Note the Linux specific bits are not necessarily
570 stable over kernel options, but the vendor specific
572 Also note that user programs calling CPUID directly
573 or using the feature without checking anything
574 will still see it. This just prevents it from
575 being used by the kernel or shown in /proc/cpuinfo.
576 Also note the kernel might malfunction if you disable
579 cma=nn[MG]@[start[MG][-end[MG]]]
581 Sets the size of kernel global memory area for
582 contiguous memory allocations and optionally the
583 placement constraint by the physical address range of
584 memory allocations. A value of 0 disables CMA
585 altogether. For more information, see
586 include/linux/dma-contiguous.h
588 cmo_free_hint= [PPC] Format: { yes | no }
589 Specify whether pages are marked as being inactive
590 when they are freed. This is used in CMO environments
591 to determine OS memory pressure for page stealing by
595 coherent_pool=nn[KMG] [ARM,KNL]
596 Sets the size of memory pool for coherent, atomic dma
597 allocations, by default set to 256K.
599 com20020= [HW,NET] ARCnet - COM20020 chipset
601 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]]
603 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers)
607 ARCnet - COM90xx chipset (memory-mapped buffers)
608 Format: <io>[,<irq>[,<memstart>]]
610 condev= [HW,S390] console device
613 console= [KNL] Output console device and options.
615 tty<n> Use the virtual console device <n>.
619 Use the specified serial port. The options are of
620 the form "bbbbpnf", where "bbbb" is the baud rate,
621 "p" is parity ("n", "o", or "e"), "n" is number of
622 bits, and "f" is flow control ("r" for RTS or
623 omit it). Default is "9600n8".
625 See Documentation/admin-guide/serial-console.rst for more
627 Documentation/networking/netconsole.txt for an
630 uart[8250],io,<addr>[,options]
631 uart[8250],mmio,<addr>[,options]
632 uart[8250],mmio16,<addr>[,options]
633 uart[8250],mmio32,<addr>[,options]
634 uart[8250],0x<addr>[,options]
635 Start an early, polled-mode console on the 8250/16550
636 UART at the specified I/O port or MMIO address,
637 switching to the matching ttyS device later.
638 MMIO inter-register address stride is either 8-bit
639 (mmio), 16-bit (mmio16), or 32-bit (mmio32).
640 If none of [io|mmio|mmio16|mmio32], <addr> is assumed
641 to be equivalent to 'mmio'. 'options' are specified in
642 the same format described for ttyS above; if unspecified,
643 the h/w is not re-initialized.
645 hvc<n> Use the hypervisor console device <n>. This is for
646 both Xen and PowerPC hypervisors.
648 If the device connected to the port is not a TTY but a braille
649 device, prepend "brl," before the device type, for instance
651 For now, only VisioBraille is supported.
654 [KNL] Change console messages format
656 By default we print messages on consoles in
657 "[time stamp] text\n" format (time stamp may not be
658 printed, depending on CONFIG_PRINTK_TIME or
659 `printk_time' param).
661 Switch to syslog format: "<%u>[time stamp] text\n"
662 IOW, each message will have a facility and loglevel
663 prefix. The format is similar to one used by syslog()
664 syscall, or to executing "dmesg -S --raw" or to reading
667 consoleblank= [KNL] The console blank (screen saver) timeout in
668 seconds. A value of 0 disables the blank timer.
672 [KNL] Change the default value for
673 /proc/<pid>/coredump_filter.
674 See also Documentation/filesystems/proc.txt.
676 coresight_cpu_debug.enable
679 Enable/disable the CPU sampling based debugging.
680 0: default value, disable debugging
681 1: enable debugging at boot time
683 cpuidle.off=1 [CPU_IDLE]
684 disable the cpuidle sub-system
687 [CPU_IDLE] Name of the cpuidle governor to use.
689 cpufreq.off=1 [CPU_FREQ]
690 disable the cpufreq sub-system
693 [X86] Delay for N microsec between assert and de-assert
694 of APIC INIT to start processors. This delay occurs
695 on every CPU online, such as boot, and resume from suspend.
698 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver
700 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>]
702 crashkernel=size[KMG][@offset[KMG]]
703 [KNL] Using kexec, Linux can switch to a 'crash kernel'
704 upon panic. This parameter reserves the physical
705 memory region [offset, offset + size] for that kernel
706 image. If '@offset' is omitted, then a suitable offset
707 is selected automatically.
708 [KNL, x86_64] select a region under 4G first, and
709 fall back to reserve region above 4G when '@offset'
710 hasn't been specified.
711 See Documentation/admin-guide/kdump/kdump.rst for further details.
713 crashkernel=range1:size1[,range2:size2,...][@offset]
714 [KNL] Same as above, but depends on the memory
715 in the running system. The syntax of range is
716 start-[end] where start and end are both
717 a memory unit (amount[KMG]). See also
718 Documentation/admin-guide/kdump/kdump.rst for an example.
720 crashkernel=size[KMG],high
721 [KNL, x86_64] range could be above 4G. Allow kernel
722 to allocate physical memory region from top, so could
723 be above 4G if system have more than 4G ram installed.
724 Otherwise memory region will be allocated below 4G, if
726 It will be ignored if crashkernel=X is specified.
727 crashkernel=size[KMG],low
728 [KNL, x86_64] range under 4G. When crashkernel=X,high
729 is passed, kernel could allocate physical memory region
730 above 4G, that cause second kernel crash on system
731 that require some amount of low memory, e.g. swiotlb
732 requires at least 64M+32K low memory, also enough extra
733 low memory is needed to make sure DMA buffers for 32-bit
734 devices won't run out. Kernel would try to allocate at
735 at least 256M below 4G automatically.
736 This one let user to specify own low range under 4G
737 for second kernel instead.
738 0: to disable low allocation.
739 It will be ignored when crashkernel=X,high is not used
740 or memory reserved is below 4G.
743 [KNL] Disable crypto self-tests
748 cs89x0_media= [HW,NET]
749 Format: { rj45 | aui | bnc }
752 See header of drivers/s390/block/dasd_devmap.c.
754 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port
755 (one device per port)
756 Format: <port#>,<type>
757 See also Documentation/input/devices/joystick-parport.rst
759 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot
761 Documentation/admin-guide/dynamic-debug-howto.rst for
762 details. Deprecated, see dyndbg.
764 debug [KNL] Enable kernel debugging (events log level).
767 [KNL] Enable printing [hashed] pointers early in the
768 boot sequence. If enabled, we use a weak hash instead
769 of siphash to hash pointers. Use this option if you are
770 seeing instances of '(___ptrval___)') and need to see a
771 value (hashed pointer) instead. Cryptographically
772 insecure, please do not use on production kernels.
775 [KNL] verbose self-tests
777 Print debugging info while doing the locking API
779 We default to 0 (no extra messages), setting it to
780 1 will print _a lot_ more information - normally
781 only useful to kernel developers.
783 debug_objects [KNL] Enable object debugging
786 [KNL] Disable object debugging
788 debug_guardpage_minorder=
789 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this
790 parameter allows control of the order of pages that will
791 be intentionally kept free (and hence protected) by the
792 buddy allocator. Bigger value increase the probability
793 of catching random memory corruption, but reduce the
794 amount of memory for normal system use. The maximum
795 possible value is MAX_ORDER/2. Setting this parameter
796 to 1 or 2 should be enough to identify most random
797 memory corruption problems caused by bugs in kernel or
798 driver code when a CPU writes to (or reads from) a
799 random memory location. Note that there exists a class
800 of memory corruptions problems caused by buggy H/W or
801 F/W or by drivers badly programing DMA (basically when
802 memory is written at bus level and the CPU MMU is
803 bypassed) which are not detectable by
804 CONFIG_DEBUG_PAGEALLOC, hence this option will not help
805 tracking down these problems.
808 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter
809 enables the feature at boot time. By default, it is
810 disabled and the system will work mostly the same as a
811 kernel built without CONFIG_DEBUG_PAGEALLOC.
812 Note: to get most of debug_pagealloc error reports, it's
813 useful to also enable the page_owner functionality.
814 on: enable the feature
816 debugpat [X86] Enable PAT debugging
818 decnet.addr= [HW,NET]
819 Format: <area>[,<node>]
820 See also Documentation/networking/decnet.txt.
823 [same as hugepagesz=] The size of the default
824 HugeTLB page size. This is the size represented by
825 the legacy /proc/ hugepages APIs, used for SHM, and
826 default size when mounting hugetlbfs filesystems.
827 Defaults to the default architecture's huge page size
830 deferred_probe_timeout=
831 [KNL] Debugging option to set a timeout in seconds for
832 deferred probe to give up waiting on dependencies to
833 probe. Only specific dependencies (subsystems or
834 drivers) that have opted in will be ignored. A timeout of 0
835 will timeout at the end of initcalls. This option will also
836 dump out devices still on the deferred probe list after
840 Set number of hash buckets for dentry cache.
842 disable_1tb_segments [PPC]
843 Disables the use of 1TB hash page table segments. This
844 causes the kernel to fall back to 256MB segments which
845 can be useful when debugging issues that require an SLB
849 See Documentation/networking/ipv6.txt.
852 [KNL] Under CONFIG_HARDENED_USERCOPY, whether
853 hardening is enabled for this boot. Hardened
854 usercopy checking is used to protect the kernel
855 from reading or writing beyond known memory
856 allocation boundaries as a proactive defense
857 against bounds-checking flaws in the kernel's
858 copy_to_user()/copy_from_user() interface.
859 on Perform hardened usercopy checks (default).
860 off Disable hardened usercopy checks.
863 Disable RADIX MMU mode on POWER9
866 Disable TLBIE instruction. Currently does not work
867 with KVM, with HASH MMU, or with coherent accelerators.
869 disable_cpu_apicid= [X86,APIC,SMP]
871 The number of initial APIC ID for the
872 corresponding CPU to be disabled at boot,
873 mostly used for the kdump 2nd kernel to
874 disable BSP to wake up multiple CPUs without
875 causing system reset or hang due to sending
878 perf_v4_pmi= [X86,INTEL]
880 Disable Intel PMU counter freezing feature.
881 The feature only exists starting from
882 Arch Perfmon v4 (Skylake and newer).
884 disable_ddw [PPC/PSERIES]
885 Disable Dynamic DMA Window support. Use this if
886 to workaround buggy firmware.
889 See Documentation/networking/ipv6.txt.
891 disable_mtrr_cleanup [X86]
892 The kernel tries to adjust MTRR layout from continuous
893 to discrete, to make X server driver able to add WB
894 entry later. This parameter disables that.
896 disable_mtrr_trim [X86, Intel and AMD only]
897 By default the kernel will trim any uncacheable
898 memory out of your available memory pool based on
899 MTRR settings. This parameter disables that behavior,
900 possibly causing your machine to run very slowly.
902 disable_timer_pin_1 [X86]
903 Disable PIN 1 of APIC timer
904 Can be useful to work around chipset bugs.
906 dis_ucode_ldr [X86] Disable the microcode loader.
908 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support,
909 this option disables the debugging code at boot.
911 dma_debug_entries=<number>
912 This option allows to tune the number of preallocated
913 entries for DMA-API debugging code. One entry is
914 required per DMA-API allocation. Use this if the
915 DMA-API debugging code disables itself because the
916 architectural default is too low.
918 dma_debug_driver=<driver_name>
919 With this option the DMA-API debugging driver
920 filter feature can be enabled at boot time. Just
921 pass the driver to filter for as the parameter.
922 The filter can be disabled or changed to another
923 driver later using sysfs.
925 driver_async_probe= [KNL]
926 List of driver names to be probed asynchronously.
927 Format: <driver_name1>,<driver_name2>...
929 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>]
930 Broken monitors, graphic adapters, KVMs and EDIDless
931 panels may send no or incorrect EDID data sets.
932 This parameter allows to specify an EDID data sets
933 in the /lib/firmware directory that are used instead.
934 Generic built-in EDID data sets are used, if one of
935 edid/1024x768.bin, edid/1280x1024.bin,
936 edid/1680x1050.bin, or edid/1920x1080.bin is given
937 and no file with the same name exists. Details and
938 instructions how to build your own EDID data are
939 available in Documentation/driver-api/edid.rst. An EDID
940 data set will only be used for a particular connector,
941 if its name and a colon are prepended to the EDID
942 name. Each connector may use a unique EDID data
943 set by separating the files with a comma. An EDID
944 data set with no connector name will be used for
945 any connectors not explicitly specified.
950 Format: {"off" | "known"}
951 Control how the dt_cpu_ftrs device-tree binding is
952 used for CPU feature discovery and setup (if it
954 off: Do not use it, fall back to legacy cpu table.
955 known: Do not pass through unknown features to guests
956 or userspace, only those that the kernel is aware of.
958 dump_apple_properties [X86]
959 Dump name and content of EFI device properties on
960 x86 Macs. Useful for driver authors to determine
961 what data is available or for reverse-engineering.
963 dyndbg[="val"] [KNL,DYNAMIC_DEBUG]
964 module.dyndbg[="val"]
965 Enable debug messages at boot time. See
966 Documentation/admin-guide/dynamic-debug-howto.rst
969 nompx [X86] Disables Intel Memory Protection Extensions.
970 See Documentation/x86/intel_mpx.rst for more
971 information about the feature.
973 nopku [X86] Disable Memory Protection Keys CPU feature found
976 module.async_probe [KNL]
977 Enable asynchronous probe on this module.
979 early_ioremap_debug [KNL]
980 Enable debug messages in early_ioremap support. This
981 is useful for tracking down temporary early mappings
982 which are not unmapped.
984 earlycon= [KNL] Output early console device and options.
986 [ARM64] The early console is determined by the
987 stdout-path property in device tree's chosen node,
988 or determined by the ACPI SPCR table.
990 [X86] When used with no options the early console is
991 determined by the ACPI SPCR table.
993 cdns,<addr>[,options]
994 Start an early, polled-mode console on a Cadence
995 (xuartps) serial port at the specified address. Only
996 supported option is baud rate. If baud rate is not
997 specified, the serial port must already be setup and
1000 uart[8250],io,<addr>[,options]
1001 uart[8250],mmio,<addr>[,options]
1002 uart[8250],mmio32,<addr>[,options]
1003 uart[8250],mmio32be,<addr>[,options]
1004 uart[8250],0x<addr>[,options]
1005 Start an early, polled-mode console on the 8250/16550
1006 UART at the specified I/O port or MMIO address.
1007 MMIO inter-register address stride is either 8-bit
1008 (mmio) or 32-bit (mmio32 or mmio32be).
1009 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed
1010 to be equivalent to 'mmio'. 'options' are specified
1011 in the same format described for "console=ttyS<n>"; if
1012 unspecified, the h/w is not initialized.
1016 Start an early, polled-mode console on a pl011 serial
1017 port at the specified address. The pl011 serial port
1018 must already be setup and configured. Options are not
1019 yet supported. If 'mmio32' is specified, then only
1020 the driver will use only 32-bit accessors to read/write
1021 the device registers.
1024 Start an early, polled-mode console on a meson serial
1025 port at the specified address. The serial port must
1026 already be setup and configured. Options are not yet
1030 Start an early, polled-mode console on an msm serial
1031 port at the specified address. The serial port
1032 must already be setup and configured. Options are not
1035 msm_serial_dm,<addr>
1036 Start an early, polled-mode console on an msm serial
1037 dm port at the specified address. The serial port
1038 must already be setup and configured. Options are not
1042 Start an early, polled-mode console on a serial port
1043 of an Actions Semi SoC, such as S500 or S900, at the
1044 specified address. The serial port must already be
1045 setup and configured. Options are not yet supported.
1048 Start an early, polled-mode console on a serial port
1049 of an RDA Micro SoC, such as RDA8810PL, at the
1050 specified address. The serial port must already be
1051 setup and configured. Options are not yet supported.
1054 Use RISC-V SBI (Supervisor Binary Interface) for early
1057 smh Use ARM semihosting calls for early console.
1065 Use early console provided by serial driver available
1066 on Samsung SoCs, requires selecting proper type and
1067 a correct base address of the selected UART port. The
1068 serial port must already be setup and configured.
1069 Options are not yet supported.
1072 Start an early, polled-mode console on a lantiq serial
1073 (lqasc) port at the specified address. The serial port
1074 must already be setup and configured. Options are not
1079 Use early console provided by Freescale LP UART driver
1080 found on Freescale Vybrid and QorIQ LS1021A processors.
1081 A valid base address must be provided, and the serial
1082 port must already be setup and configured.
1085 Start an early, polled-mode console on the
1086 Armada 3700 serial port at the specified
1087 address. The serial port must already be setup
1088 and configured. Options are not yet supported.
1091 Start an early, polled-mode console on a Qualcomm
1092 Generic Interface (GENI) based serial port at the
1093 specified address. The serial port must already be
1094 setup and configured. Options are not yet supported.
1097 Start an early, unaccelerated console on the EFI
1098 memory mapped framebuffer (if available). On cache
1099 coherent non-x86 systems that use system memory for
1100 the framebuffer, pass the 'ram' option so that it is
1101 mapped with the correct attributes.
1104 Use early console provided by Freescale LinFlex UART
1105 serial driver for NXP S32V234 SoCs. A valid base
1106 address must be provided, and the serial port must
1107 already be setup and configured.
1109 earlyprintk= [X86,SH,ARM,M68k,S390]
1113 earlyprintk=serial[,ttySn[,baudrate]]
1114 earlyprintk=serial[,0x...[,baudrate]]
1115 earlyprintk=ttySn[,baudrate]
1116 earlyprintk=dbgp[debugController#]
1117 earlyprintk=pciserial[,force],bus:device.function[,baudrate]
1118 earlyprintk=xdbc[xhciController#]
1120 earlyprintk is useful when the kernel crashes before
1121 the normal console is initialized. It is not enabled by
1122 default because it has some cosmetic problems.
1124 Append ",keep" to not disable it when the real console
1127 Only one of vga, efi, serial, or usb debug port can
1130 Currently only ttyS0 and ttyS1 may be specified by
1131 name. Other I/O ports may be explicitly specified
1132 on some architectures (x86 and arm at least) by
1133 replacing ttySn with an I/O port address, like this:
1134 earlyprintk=serial,0x1008,115200
1135 You can find the port for a given device in
1136 /proc/tty/driver/serial:
1137 2: uart:ST16650V2 port:00001008 irq:18 ...
1139 Interaction with the standard serial driver is not
1142 The VGA and EFI output is eventually overwritten by
1145 The xen output can only be used by Xen PV guests.
1147 The sclp output can only be used on s390.
1149 The optional "force" to "pciserial" enables use of a
1150 PCI device even when its classcode is not of the
1153 edac_report= [HW,EDAC] Control how to report EDAC event
1154 Format: {"on" | "off" | "force"}
1155 on: enable EDAC to report H/W event. May be overridden
1156 by other higher priority error reporting module.
1157 off: disable H/W event reporting through EDAC.
1158 force: enforce the use of EDAC to report H/W event.
1161 ekgdboc= [X86,KGDB] Allow early kernel console debugging
1164 This is designed to be used in conjunction with
1165 the boot argument: earlyprintk=vga
1168 Format: {"off" | "on" | "skip[mbr]"}
1171 Format: { "old_map", "nochunk", "noruntime", "debug" }
1172 old_map [X86-64]: switch to the old ioremap-based EFI
1173 runtime services mapping. 32-bit still uses this one by
1175 nochunk: disable reading files in "chunks" in the EFI
1176 boot stub, as chunking can cause problems with some
1177 firmware implementations.
1178 noruntime : disable EFI runtime services support
1179 debug: enable misc debug output
1181 efi_no_storage_paranoia [EFI; X86]
1182 Using this parameter you can use more than 50% of
1183 your efi variable storage. Use this parameter only if
1184 you are really sure that your UEFI does sane gc and
1185 fulfills the spec otherwise your board may brick.
1187 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86]
1188 Add arbitrary attribute to specific memory range by
1189 updating original EFI memory map.
1190 Region of memory which aa attribute is added to is
1192 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000
1193 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000)
1194 attribute is added to range 0x100000000-0x180000000 and
1195 0x10a0000000-0x1120000000.
1197 Using this parameter you can do debugging of EFI memmap
1198 related feature. For example, you can do debugging of
1199 Address Range Mirroring feature even if your box
1202 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT
1203 that is to be dynamically loaded by Linux. If there are
1204 multiple variables with the same name but with different
1205 vendor GUIDs, all of them will be loaded. See
1206 Documentation/admin-guide/acpi/ssdt-overlays.rst for details.
1209 eisa_irq_edge= [PARISC,HW]
1210 See header of drivers/parisc/eisa.c.
1213 See comment before function elanfreq_setup() in
1214 arch/x86/kernel/cpu/cpufreq/elanfreq.c.
1216 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390]
1217 Specifies physical address of start of kernel core
1218 image elf header and optionally the size. Generally
1219 kexec loader will pass this option to capture kernel.
1220 See Documentation/admin-guide/kdump/kdump.rst for details.
1222 enable_mtrr_cleanup [X86]
1223 The kernel tries to adjust MTRR layout from continuous
1224 to discrete, to make X server driver able to add WB
1225 entry later. This parameter enables that.
1227 enable_timer_pin_1 [X86]
1228 Enable PIN 1 of APIC timer
1229 Can be useful to work around chipset bugs
1230 (in particular on some ATI chipsets).
1231 The kernel tries to set a reasonable default.
1233 enforcing [SELINUX] Set initial enforcing status.
1235 See security/selinux/Kconfig help text.
1236 0 -- permissive (log only, no denials).
1237 1 -- enforcing (deny and log).
1239 Value can be changed at runtime via /selinux/enforce.
1242 Disable Error Record Serialization Table (ERST)
1245 ether= [HW,NET] Ethernet cards parameters
1246 This option is obsoleted by the "netdev=" option, which
1247 has equivalent usage. See its documentation for details.
1251 Permit 'security.evm' to be updated regardless of
1252 current integrity status.
1256 fail_make_request=[KNL]
1257 General fault injection mechanism.
1258 Format: <interval>,<probability>,<space>,<times>
1259 See also Documentation/fault-injection/.
1262 See Documentation/admin-guide/blockdev/floppy.rst.
1264 force_pal_cache_flush
1265 [IA-64] Avoid check_sal_cache_flush which may hang on
1266 buggy SAL_CACHE_FLUSH implementations. Using this
1267 parameter will force ia64_sal_cache_flush to call
1268 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
1271 Forcefully enable Physical Address Extension (PAE).
1272 Many Pentium M systems disable PAE but may have a
1273 functionally usable PAE implementation.
1274 Warning: use of this parameter will taint the kernel
1275 and may cause unknown problems.
1278 [FTRACE] will set and start the specified tracer
1279 as early as possible in order to facilitate early
1282 ftrace_dump_on_oops[=orig_cpu]
1283 [FTRACE] will dump the trace buffers on oops.
1284 If no parameter is passed, ftrace will dump
1285 buffers of all CPUs, but if you pass orig_cpu, it will
1286 dump only the buffer of the CPU that triggered the
1289 ftrace_filter=[function-list]
1290 [FTRACE] Limit the functions traced by the function
1291 tracer at boot up. function-list is a comma separated
1292 list of functions. This list can be changed at run
1293 time by the set_ftrace_filter file in the debugfs
1296 ftrace_notrace=[function-list]
1297 [FTRACE] Do not trace the functions specified in
1298 function-list. This list can be changed at run time
1299 by the set_ftrace_notrace file in the debugfs
1302 ftrace_graph_filter=[function-list]
1303 [FTRACE] Limit the top level callers functions traced
1304 by the function graph tracer at boot up.
1305 function-list is a comma separated list of functions
1306 that can be changed at run time by the
1307 set_graph_function file in the debugfs tracing directory.
1309 ftrace_graph_notrace=[function-list]
1310 [FTRACE] Do not trace from the functions specified in
1311 function-list. This list is a comma separated list of
1312 functions that can be changed at run time by the
1313 set_graph_notrace file in the debugfs tracing directory.
1315 ftrace_graph_max_depth=<uint>
1316 [FTRACE] Used with the function graph tracer. This is
1317 the max depth it will trace into a function. This value
1318 can be changed at run time by the max_graph_depth file
1319 in the tracefs tracing directory. default: 0 (no limit)
1322 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad
1323 support via parallel port (up to 5 devices per port)
1324 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5>
1325 See also Documentation/input/devices/joystick-parport.rst
1329 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART
1333 gcov_persist= [GCOV] When non-zero (default), profiling data for
1334 kernel modules is saved and remains accessible via
1335 debugfs, even when the module is unloaded/reloaded.
1336 When zero, profiling data is discarded and associated
1337 debugfs files are removed at module unload time.
1339 goldfish [X86] Enable the goldfish android emulator platform.
1340 Don't use this when you are not running on the
1343 gpt [EFI] Forces disk with valid GPT signature but
1344 invalid Protective MBR to be treated as GPT. If the
1345 primary GPT is corrupted, it enables the backup/alternate
1346 GPT to be used instead.
1348 grcan.enable0= [HW] Configuration of physical interface 0. Determines
1349 the "Enable 0" bit of the configuration register.
1352 grcan.enable1= [HW] Configuration of physical interface 1. Determines
1353 the "Enable 0" bit of the configuration register.
1356 grcan.select= [HW] Select which physical interface to use.
1359 grcan.txsize= [HW] Sets the size of the tx buffer.
1360 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0.
1362 grcan.rxsize= [HW] Sets the size of the rx buffer.
1363 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0.
1366 gpio-mockup.gpio_mockup_ranges
1367 [HW] Sets the ranges of gpiochip of for this device.
1368 Format: <start1>,<end1>,<start2>,<end2>...
1370 hardlockup_all_cpu_backtrace=
1371 [KNL] Should the hard-lockup detector generate
1372 backtraces on all cpus.
1375 hashdist= [KNL,NUMA] Large hashes allocated during boot
1376 are distributed across NUMA nodes. Defaults on
1377 for 64-bit NUMA, off otherwise.
1378 Format: 0 | 1 (for off | on)
1380 hcl= [IA-64] SGI's Hardware Graph compatibility layer
1382 hd= [EIDE] (E)IDE hard drive subsystem geometry
1383 Format: <cyl>,<head>,<sect>
1386 Disable Hardware Error Source Table (HEST) support;
1387 corresponding firmware-first mode error processing
1388 logic will be disabled.
1390 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact
1391 size of <nn>. This works even on boxes that have no
1392 highmem otherwise. This also works to reduce highmem
1393 size on bigger boxes.
1395 highres= [KNL] Enable/disable high resolution timer mode.
1396 Valid parameters: "on", "off"
1401 hpet= [X86-32,HPET] option to control HPET usage
1402 Format: { enable (default) | disable | force |
1404 disable: disable HPET and use PIT instead
1405 force: allow force enabled of undocumented chips (ICH4,
1407 verbose: show contents of HPET registers during setup
1409 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET
1410 registers. Default set by CONFIG_HPET_MMAP_DEFAULT.
1412 hugepages= [HW,X86-32,IA-64] HugeTLB pages to allocate at boot.
1413 hugepagesz= [HW,IA-64,PPC,X86-64] The size of the HugeTLB pages.
1414 On x86-64 and powerpc, this option can be specified
1415 multiple times interleaved with hugepages= to reserve
1416 huge pages of different sizes. Valid pages sizes on
1417 x86-64 are 2M (when the CPU supports "pse") and 1G
1418 (when the CPU supports the "pdpe1gb" cpuinfo flag).
1421 [KNL] Should the hung task detector generate panics.
1424 A nonzero value instructs the kernel to panic when a
1425 hung task is detected. The default value is controlled
1426 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time
1427 option. The value selected by this boot parameter can
1428 be changed later by the kernel.hung_task_panic sysctl.
1430 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC)
1431 terminal devices. Valid values: 0..8
1432 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs.
1433 If specified, z/VM IUCV HVC accepts connections
1434 from listed z/VM user IDs only.
1436 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations
1437 which allow the hypervisor to 'idle' the
1438 guest on lock contention.
1441 Do not unregister boot console at start. This is only
1442 useful for debugging when something happens in the window
1443 between unregistering the boot console and initializing
1446 i2c_bus= [HW] Override the default board specific I2C bus speed
1447 or register an additional I2C bus that is not
1448 registered from board initialization code.
1452 i8042.debug [HW] Toggle i8042 debug mode
1453 i8042.unmask_kbd_data
1454 [HW] Enable printing of interrupt data from the KBD port
1455 (disabled by default, and as a pre-condition
1456 requires that i8042.debug=1 be enabled)
1457 i8042.direct [HW] Put keyboard port into non-translated mode
1458 i8042.dumbkbd [HW] Pretend that controller can only read data from
1459 keyboard and cannot control its state
1460 (Don't attempt to blink the leds)
1461 i8042.noaux [HW] Don't check for auxiliary (== mouse) port
1462 i8042.nokbd [HW] Don't check/create keyboard port
1463 i8042.noloop [HW] Disable the AUX Loopback command while probing
1465 i8042.nomux [HW] Don't check presence of an active multiplexing
1467 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX
1469 i8042.notimeout [HW] Ignore timeout condition signalled by controller
1470 i8042.reset [HW] Reset the controller during init, cleanup and
1471 suspend-to-ram transitions, only during s2r
1472 transitions, or never reset
1473 Format: { 1 | Y | y | 0 | N | n }
1474 1, Y, y: always reset controller
1475 0, N, n: don't ever reset controller
1476 Default: only on s2r transitions on x86; most other
1477 architectures force reset to be always executed
1478 i8042.unlock [HW] Unlock (ignore) the keylock
1479 i8042.kbdreset [HW] Reset device connected to KBD port
1483 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
1484 indicates that the driver is running on unsupported
1486 i8k.force [HW] Activate i8k driver even if SMM BIOS signature
1487 does not match list of supported models.
1489 [HW] Report power status in /proc/i8k
1490 (disabled by default)
1491 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN
1494 i915.invert_brightness=
1495 [DRM] Invert the sense of the variable that is used to
1496 set the brightness of the panel backlight. Normally a
1497 brightness value of 0 indicates backlight switched off,
1498 and the maximum of the brightness value sets the backlight
1499 to maximum brightness. If this parameter is set to 0
1500 (default) and the machine requires it, or this parameter
1501 is set to 1, a brightness value of 0 sets the backlight
1502 to maximum brightness, and the maximum of the brightness
1503 value switches the backlight off.
1504 -1 -- never invert brightness
1505 0 -- machine default
1506 1 -- force brightness inversion
1509 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
1511 ide-core.nodma= [HW] (E)IDE subsystem
1512 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc
1513 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr
1514 .cdrom .chs .ignore_cable are additional options
1515 See Documentation/ide/ide.rst.
1517 ide-generic.probe-mask= [HW] (E)IDE subsystem
1519 Probe mask for legacy ISA IDE ports. Depending on
1520 platform up to 6 ports are supported, enabled by
1521 setting corresponding bits in the mask to 1. The
1522 default value is 0x0, which has a special meaning.
1523 On systems that have PCI, it triggers scanning the
1524 PCI bus for the first and the second port, which
1525 are then probed. On systems without PCI the value
1526 of 0x0 enables probing the two first ports as if it
1529 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem
1530 Claim all unknown PCI IDE storage controllers.
1533 Format: idle=poll, idle=halt, idle=nomwait
1534 Poll forces a polling idle loop that can slightly
1535 improve the performance of waking up a idle CPU, but
1536 will use a lot of power and make the system run hot.
1538 idle=halt: Halt is forced to be used for CPU idle.
1539 In such case C2/C3 won't be used again.
1540 idle=nomwait: Disable mwait for CPU C-states
1542 ieee754= [MIPS] Select IEEE Std 754 conformance mode
1543 Format: { strict | legacy | 2008 | relaxed }
1546 Choose which programs will be accepted for execution
1547 based on the IEEE 754 NaN encoding(s) supported by
1548 the FPU and the NaN encoding requested with the value
1549 of an ELF file header flag individually set by each
1550 binary. Hardware implementations are permitted to
1551 support either or both of the legacy and the 2008 NaN
1554 Available settings are as follows:
1555 strict accept binaries that request a NaN encoding
1556 supported by the FPU
1557 legacy only accept legacy-NaN binaries, if supported
1559 2008 only accept 2008-NaN binaries, if supported
1561 relaxed accept any binaries regardless of whether
1562 supported by the FPU
1564 The FPU emulator is always able to support both NaN
1565 encodings, so if no FPU hardware is present or it has
1566 been disabled with 'nofpu', then the settings of
1567 'legacy' and '2008' strap the emulator accordingly,
1568 'relaxed' straps the emulator for both legacy-NaN and
1569 2008-NaN, whereas 'strict' enables legacy-NaN only on
1570 legacy processors and both NaN encodings on MIPS32 or
1573 The setting for ABS.fmt/NEG.fmt instruction execution
1574 mode generally follows that for the NaN encoding,
1575 except where unsupported by hardware.
1577 ignore_loglevel [KNL]
1578 Ignore loglevel setting - this will print /all/
1579 kernel messages to the console. Useful for debugging.
1580 We also add it as printk module parameter, so users
1581 could change it dynamically, usually by
1582 /sys/module/printk/parameters/ignore_loglevel.
1585 Ignore RLIMIT_DATA setting for data mappings,
1586 print warning at first misuse. Can be changed via
1587 /sys/module/kernel/parameters/ignore_rlimit_data.
1589 ihash_entries= [KNL]
1590 Set number of hash buckets for inode cache.
1592 ima_appraise= [IMA] appraise integrity measurements
1593 Format: { "off" | "enforce" | "fix" | "log" }
1596 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead.
1597 The builtin appraise policy appraises all files
1600 ima_canonical_fmt [IMA]
1601 Use the canonical format for the binary runtime
1602 measurements, instead of host native format.
1605 Format: { md5 | sha1 | rmd160 | sha256 | sha384
1609 The list of supported hash algorithms is defined
1610 in crypto/hash_info.h.
1613 The builtin policies to load during IMA setup.
1614 Format: "tcb | appraise_tcb | secure_boot |
1617 The "tcb" policy measures all programs exec'd, files
1618 mmap'd for exec, and all files opened with the read
1619 mode bit set by either the effective uid (euid=0) or
1622 The "appraise_tcb" policy appraises the integrity of
1623 all files owned by root.
1625 The "secure_boot" policy appraises the integrity
1626 of files (eg. kexec kernel image, kernel modules,
1627 firmware, policy, etc) based on file signatures.
1629 The "fail_securely" policy forces file signature
1630 verification failure also on privileged mounted
1631 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE
1634 ima_tcb [IMA] Deprecated. Use ima_policy= instead.
1635 Load a policy which meets the needs of the Trusted
1636 Computing Base. This means IMA will measure all
1637 programs exec'd, files mmap'd for exec, and all files
1638 opened for read by uid=0.
1641 Select one of defined IMA measurements template formats.
1642 Formats: { "ima" | "ima-ng" | "ima-sig" }
1646 [IMA] Define a custom template format.
1647 Format: { "field1|...|fieldN" }
1649 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage
1650 Format: <min_file_size>
1651 Set the minimal file size for using asynchronous hash.
1652 If left unspecified, ahash usage is disabled.
1654 ahash performance varies for different data sizes on
1655 different crypto accelerators. This option can be used
1656 to achieve the best performance for a particular HW.
1658 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size
1660 Set hashing buffer size. Default: 4k.
1662 ahash performance varies for different chunk sizes on
1663 different crypto accelerators. This option can be used
1664 to achieve best performance for particular HW.
1668 Run specified binary instead of /sbin/init as init
1671 initcall_debug [KNL] Trace initcalls as they are executed. Useful
1672 for working out where the kernel is dying during
1675 initcall_blacklist= [KNL] Do not execute a comma-separated list of
1676 initcall functions. Useful for debugging built-in
1677 modules and initcalls.
1679 initrd= [BOOT] Specify the location of the initial ramdisk
1681 init_on_alloc= [MM] Fill newly allocated pages and heap objects with
1684 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON.
1686 init_on_free= [MM] Fill freed pages and heap objects with zeroes.
1688 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON.
1690 init_pkru= [x86] Specify the default memory protection keys rights
1691 register contents for all processes. 0x55555554 by
1692 default (disallow access to all but pkey 0). Can
1693 override in debugfs after boot.
1695 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver
1698 int_pln_enable [x86] Enable power limit notification interrupt
1700 integrity_audit=[IMA]
1701 Format: { "0" | "1" }
1702 0 -- basic integrity auditing messages. (Default)
1703 1 -- additional integrity auditing messages.
1705 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option
1707 Enable intel iommu driver.
1709 Disable intel iommu driver.
1710 igfx_off [Default Off]
1711 By default, gfx is mapped as normal device. If a gfx
1712 device has a dedicated DMAR unit, the DMAR unit is
1713 bypassed by not enabling DMAR with this option. In
1714 this case, gfx device will use physical address for
1717 With this option iommu will not optimize to look
1718 for io virtual address below 32-bit forcing dual
1719 address cycle on pci bus for cards supporting greater
1720 than 32-bit addressing. The default is to look
1721 for translation below 32-bit and if not available
1722 then look in the higher range.
1723 strict [Default Off]
1724 With this option on every unmap_single operation will
1725 result in a hardware IOTLB flush operation as opposed
1726 to batching them for performance.
1727 sp_off [Default Off]
1728 By default, super page will be supported if Intel IOMMU
1729 has the capability. With this option, super page will
1732 By default, scalable mode will be disabled even if the
1733 hardware advertises that it has support for the scalable
1734 mode translation. With this option set, scalable mode
1735 will be used on hardware which claims to support it.
1736 tboot_noforce [Default Off]
1737 Do not force the Intel IOMMU enabled under tboot.
1738 By default, tboot will force Intel IOMMU on, which
1739 could harm performance of some high-throughput
1740 devices like 40GBit network cards, even if identity
1742 Note that using this option lowers the security
1743 provided by tboot because it makes the system
1744 vulnerable to DMA attacks.
1745 nobounce [Default off]
1746 Disable bounce buffer for unstrusted devices such as
1747 the Thunderbolt devices. This will treat the untrusted
1748 devices as the trusted ones, hence might expose security
1749 risks of DMA attacks.
1751 intel_idle.max_cstate= [KNL,HW,ACPI,X86]
1752 0 disables intel_idle and fall back on acpi_idle.
1753 1 to 9 specify maximum depth of C-state.
1757 Do not enable intel_pstate as the default
1758 scaling driver for the supported processors
1760 Use intel_pstate as a scaling driver, but configure it
1761 to work with generic cpufreq governors (instead of
1762 enabling its internal governor). This mode cannot be
1763 used along with the hardware-managed P-states (HWP)
1766 Enable intel_pstate on systems that prohibit it by default
1767 in favor of acpi-cpufreq. Forcing the intel_pstate driver
1768 instead of acpi-cpufreq may disable platform features, such
1769 as thermal controls and power capping, that rely on ACPI
1770 P-States information being indicated to OSPM and therefore
1771 should be used with caution. This option does not work with
1772 processors that aren't supported by the intel_pstate driver
1773 or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
1775 Do not enable hardware P state control (HWP)
1778 Only load intel_pstate on systems which support
1779 hardware P state control (HWP) if available.
1781 Enforce ACPI _PPC performance limits. If the Fixed ACPI
1782 Description Table, specifies preferred power management
1783 profile as "Enterprise Server" or "Performance Server",
1784 then this feature is turned on by default.
1786 Allow per-logical-CPU P-State performance control limits using
1787 cpufreq sysfs interface
1789 intremap= [X86-64, Intel-IOMMU]
1790 on enable Interrupt Remapping (default)
1791 off disable Interrupt Remapping
1792 nosid disable Source ID checking
1794 BIOS x2APIC opt-out request will be ignored
1795 nopost disable Interrupt Posting
1797 iomem= Disable strict checking of access to MMIO memory
1798 strict regions from userspace.
1813 nobypass [PPC/POWERNV]
1814 Disable IOMMU bypass, using IOMMU for PCI devices.
1816 iommu.strict= [ARM64] Configure TLB invalidation behaviour
1817 Format: { "0" | "1" }
1819 Request that DMA unmap operations use deferred
1820 invalidation of hardware TLBs, for increased
1821 throughput at the cost of reduced device isolation.
1822 Will fall back to strict mode if not supported by
1823 the relevant IOMMU driver.
1824 1 - Strict mode (default).
1825 DMA unmap operations invalidate IOMMU hardware TLBs
1829 [ARM64, X86] Configure DMA to bypass the IOMMU by default.
1830 Format: { "0" | "1" }
1831 0 - Use IOMMU translation for DMA.
1832 1 - Bypass the IOMMU for DMA.
1833 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH.
1835 io7= [HW] IO7 for Marvel based alpha systems
1836 See comment before marvel_specify_io7 in
1837 arch/alpha/kernel/core_marvel.c.
1839 io_delay= [X86] I/O delay method
1841 Standard port 0x80 based delay
1843 Alternate port 0xed based delay (needed on some systems)
1845 Simple two microseconds delay
1850 See Documentation/filesystems/nfs/nfsroot.txt.
1852 ipcmni_extend [KNL] Extend the maximum number of unique System V
1853 IPC identifiers from 32,768 to 16,777,216.
1855 irqaffinity= [SMP] Set the default irq affinity mask
1856 The argument is a cpu list, as described above.
1858 irqchip.gicv2_force_probe=
1861 Force the kernel to look for the second 4kB page
1862 of a GICv2 controller even if the memory range
1863 exposed by the device tree is too small.
1865 irqchip.gicv3_nolpi=
1867 Force the kernel to ignore the availability of
1868 LPIs (and by consequence ITSs). Intended for system
1869 that use the kernel as a bootloader, and thus want
1870 to let secondary kernels in charge of setting up
1873 irqchip.gicv3_pseudo_nmi= [ARM64]
1874 Enables support for pseudo-NMIs in the kernel. This
1875 requires the kernel to be built with
1876 CONFIG_ARM64_PSEUDO_NMI.
1879 When an interrupt is not handled search all handlers
1880 for it. Intended to get systems with badly broken
1884 When an interrupt is not handled search all handlers
1885 for it. Also check all handlers each timer
1886 interrupt. Intended to get systems with badly broken
1890 Format: <RDP>,<reset>,<pci_scan>,<verbosity>
1892 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance.
1893 [Deprecated - use cpusets instead]
1894 Format: [flag-list,]<cpu-list>
1896 Specify one or more CPUs to isolate from disturbances
1897 specified in the flag list (default: domain):
1900 Disable the tick when a single task runs.
1902 A residual 1Hz tick is offloaded to workqueues, which you
1903 need to affine to housekeeping through the global
1904 workqueue's affinity configured via the
1905 /sys/devices/virtual/workqueue/cpumask sysfs file, or
1906 by using the 'domain' flag described below.
1908 NOTE: by default the global workqueue runs on all CPUs,
1909 so to protect individual CPUs the 'cpumask' file has to
1910 be configured manually after bootup.
1913 Isolate from the general SMP balancing and scheduling
1914 algorithms. Note that performing domain isolation this way
1915 is irreversible: it's not possible to bring back a CPU to
1916 the domains once isolated through isolcpus. It's strongly
1917 advised to use cpusets instead to disable scheduler load
1918 balancing through the "cpuset.sched_load_balance" file.
1919 It offers a much more flexible interface where CPUs can
1920 move in and out of an isolated set anytime.
1922 You can move a process onto or off an "isolated" CPU via
1923 the CPU affinity syscalls or cpuset.
1924 <cpu number> begins at 0 and the maximum value is
1925 "number of CPUs in system - 1".
1927 The format of <cpu-list> is described above.
1933 ivrs_ioapic [HW,X86_64]
1934 Provide an override to the IOAPIC-ID<->DEVICE-ID
1935 mapping provided in the IVRS ACPI table. For
1936 example, to map IOAPIC-ID decimal 10 to
1937 PCI device 00:14.0 write the parameter as:
1938 ivrs_ioapic[10]=00:14.0
1940 ivrs_hpet [HW,X86_64]
1941 Provide an override to the HPET-ID<->DEVICE-ID
1942 mapping provided in the IVRS ACPI table. For
1943 example, to map HPET-ID decimal 0 to
1944 PCI device 00:14.0 write the parameter as:
1945 ivrs_hpet[0]=00:14.0
1947 ivrs_acpihid [HW,X86_64]
1948 Provide an override to the ACPI-HID:UID<->DEVICE-ID
1949 mapping provided in the IVRS ACPI table. For
1950 example, to map UART-HID:UID AMD0020:0 to
1951 PCI device 00:14.5 write the parameter as:
1952 ivrs_acpihid[00:14.5]=AMD0020:0
1954 js= [HW,JOY] Analog joystick
1955 See Documentation/input/joydev/joystick.rst.
1958 When CONFIG_RANDOMIZE_BASE is set, this disables
1959 kernel and module base offset ASLR (Address Space
1960 Layout Randomization).
1963 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print
1964 report on every invalid memory access. Without this
1965 parameter KASAN will print report only for the first
1970 kernelcore= [KNL,X86,IA-64,PPC]
1971 Format: nn[KMGTPE] | nn% | "mirror"
1972 This parameter specifies the amount of memory usable by
1973 the kernel for non-movable allocations. The requested
1974 amount is spread evenly throughout all nodes in the
1975 system as ZONE_NORMAL. The remaining memory is used for
1976 movable memory in its own zone, ZONE_MOVABLE. In the
1977 event, a node is too small to have both ZONE_NORMAL and
1978 ZONE_MOVABLE, kernelcore memory will take priority and
1979 other nodes will have a larger ZONE_MOVABLE.
1981 ZONE_MOVABLE is used for the allocation of pages that
1982 may be reclaimed or moved by the page migration
1983 subsystem. Note that allocations like PTEs-from-HighMem
1984 still use the HighMem zone if it exists, and the Normal
1985 zone if it does not.
1987 It is possible to specify the exact amount of memory in
1988 the form of "nn[KMGTPE]", a percentage of total system
1989 memory in the form of "nn%", or "mirror". If "mirror"
1990 option is specified, mirrored (reliable) memory is used
1991 for non-movable allocations and remaining memory is used
1992 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror"
1993 are exclusive, so you cannot specify multiple forms.
1995 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port.
1996 Format: <Controller#>[,poll interval]
1997 The controller # is the number of the ehci usb debug
1998 port as it is probed via PCI. The poll interval is
1999 optional and is the number seconds in between
2000 each poll cycle to the debug port in case you need
2001 the functionality for interrupting the kernel with
2002 gdb or control-c on the dbgp connection. When
2003 not using this parameter you use sysrq-g to break into
2004 the kernel debugger.
2006 kgdboc= [KGDB,HW] kgdb over consoles.
2007 Requires a tty driver that supports console polling,
2008 or a supported polling keyboard driver (non-usb).
2009 Serial only format: <serial_device>[,baud]
2010 keyboard only format: kbd
2011 keyboard and serial format: kbd,<serial_device>[,baud]
2012 Optional Kernel mode setting:
2013 kms, kbd format: kms,kbd
2014 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud]
2016 kgdbwait [KGDB] Stop kernel execution and enter the
2017 kernel debugger at the earliest opportunity.
2019 kmac= [MIPS] korina ethernet MAC address.
2020 Configure the RouterBoard 532 series on-chip
2021 Ethernet adapter MAC address.
2023 kmemleak= [KNL] Boot-time kmemleak enable/disable
2024 Valid arguments: on, off
2026 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y,
2029 kprobe_event=[probe-list]
2030 [FTRACE] Add kprobe events and enable at boot time.
2031 The probe-list is a semicolon delimited list of probe
2032 definitions. Each definition is same as kprobe_events
2033 interface, but the parameters are comma delimited.
2034 For example, to add a kprobe event on vfs_read with
2035 arg1 and arg2, add to the command line;
2037 kprobe_event=p,vfs_read,$arg1,$arg2
2039 See also Documentation/trace/kprobetrace.rst "Kernel
2040 Boot Parameter" section.
2042 kpti= [ARM64] Control page table isolation of user
2043 and kernel address spaces.
2044 Default: enabled on cores which need mitigation.
2048 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs.
2049 Default is 0 (don't ignore, but inject #GP)
2051 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface.
2052 Default is false (don't support).
2054 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit
2059 [KVM] Controls the software workaround for the
2060 X86_BUG_ITLB_MULTIHIT bug.
2061 force : Always deploy workaround.
2062 off : Never deploy workaround.
2063 auto : Deploy workaround based on the presence of
2064 X86_BUG_ITLB_MULTIHIT.
2068 If the software workaround is enabled for the host,
2069 guests do need not to enable it for nested guests.
2071 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM.
2072 Default is 1 (enabled)
2074 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU)
2076 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode.
2078 kvm-arm.vgic_v3_group0_trap=
2079 [KVM,ARM] Trap guest accesses to GICv3 group-0
2082 kvm-arm.vgic_v3_group1_trap=
2083 [KVM,ARM] Trap guest accesses to GICv3 group-1
2086 kvm-arm.vgic_v3_common_trap=
2087 [KVM,ARM] Trap guest accesses to GICv3 common
2090 kvm-arm.vgic_v4_enable=
2091 [KVM,ARM] Allow use of GICv4 for direct injection of
2094 kvm-intel.ept= [KVM,Intel] Disable extended page tables
2095 (virtualized MMU) support on capable Intel chips.
2096 Default is 1 (enabled)
2098 kvm-intel.emulate_invalid_guest_state=
2099 [KVM,Intel] Enable emulation of invalid guest states
2100 Default is 0 (disabled)
2102 kvm-intel.flexpriority=
2103 [KVM,Intel] Disable FlexPriority feature (TPR shadow).
2104 Default is 1 (enabled)
2107 [KVM,Intel] Enable VMX nesting (nVMX).
2108 Default is 0 (disabled)
2110 kvm-intel.unrestricted_guest=
2111 [KVM,Intel] Disable unrestricted guest feature
2112 (virtualized real and unpaged mode) on capable
2113 Intel chips. Default is 1 (enabled)
2115 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault
2118 Valid arguments: never, cond, always
2120 always: L1D cache flush on every VMENTER.
2121 cond: Flush L1D on VMENTER only when the code between
2122 VMEXIT and VMENTER can leak host memory.
2123 never: Disables the mitigation
2125 Default is cond (do L1 cache flush in specific instances)
2127 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification
2128 feature (tagged TLBs) on capable Intel chips.
2129 Default is 1 (enabled)
2131 l1tf= [X86] Control mitigation of the L1TF vulnerability on
2134 The kernel PTE inversion protection is unconditionally
2135 enabled and cannot be disabled.
2138 Provides all available mitigations for the
2139 L1TF vulnerability. Disables SMT and
2140 enables all mitigations in the
2141 hypervisors, i.e. unconditional L1D flush.
2143 SMT control and L1D flush control via the
2144 sysfs interface is still possible after
2145 boot. Hypervisors will issue a warning
2146 when the first VM is started in a
2147 potentially insecure configuration,
2148 i.e. SMT enabled or L1D flush disabled.
2151 Same as 'full', but disables SMT and L1D
2152 flush runtime control. Implies the
2153 'nosmt=force' command line option.
2154 (i.e. sysfs control of SMT is disabled.)
2157 Leaves SMT enabled and enables the default
2158 hypervisor mitigation, i.e. conditional
2161 SMT control and L1D flush control via the
2162 sysfs interface is still possible after
2163 boot. Hypervisors will issue a warning
2164 when the first VM is started in a
2165 potentially insecure configuration,
2166 i.e. SMT enabled or L1D flush disabled.
2170 Disables SMT and enables the default
2171 hypervisor mitigation.
2173 SMT control and L1D flush control via the
2174 sysfs interface is still possible after
2175 boot. Hypervisors will issue a warning
2176 when the first VM is started in a
2177 potentially insecure configuration,
2178 i.e. SMT enabled or L1D flush disabled.
2181 Same as 'flush', but hypervisors will not
2182 warn when a VM is started in a potentially
2183 insecure configuration.
2186 Disables hypervisor mitigations and doesn't
2188 It also drops the swap size and available
2189 RAM limit restriction on both hypervisor and
2194 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst
2200 lapic [X86-32,APIC] Enable the local APIC even if BIOS
2203 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline
2204 value for LAPIC timer one-shot implementation. Default
2205 back to the programmable timer unit in the LAPIC.
2207 lapic_timer_c2_ok [X86,APIC] trust the local apic timer
2210 libata.dma= [LIBATA] DMA control
2211 libata.dma=0 Disable all PATA and SATA DMA
2212 libata.dma=1 PATA and SATA Disk DMA only
2213 libata.dma=2 ATAPI (CDROM) DMA only
2214 libata.dma=4 Compact Flash DMA only
2215 Combinations also work, so libata.dma=3 enables DMA
2216 for disks and CDROMs, but not CFs.
2218 libata.ignore_hpa= [LIBATA] Ignore HPA limit
2219 libata.ignore_hpa=0 keep BIOS limits (default)
2220 libata.ignore_hpa=1 ignore limits, using full disk
2222 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume
2226 libata.force= [LIBATA] Force configurations. The format is comma
2227 separated list of "[ID:]VAL" where ID is
2228 PORT[.DEVICE]. PORT and DEVICE are decimal numbers
2229 matching port, link or device. Basically, it matches
2230 the ATA ID string printed on console by libata. If
2231 the whole ID part is omitted, the last PORT and DEVICE
2232 values are used. If ID hasn't been specified yet, the
2233 configuration applies to all ports, links and devices.
2235 If only DEVICE is omitted, the parameter applies to
2236 the port and all links and devices behind it. DEVICE
2237 number of 0 either selects the first device or the
2238 first fan-out link behind PMP device. It does not
2239 select the host link. DEVICE number of 15 selects the
2240 host link and device attached to it.
2242 The VAL specifies the configuration to force. As long
2243 as there's no ambiguity shortcut notation is allowed.
2244 For example, both 1.5 and 1.5G would work for 1.5Gbps.
2245 The following configurations can be forced.
2247 * Cable type: 40c, 80c, short40c, unk, ign or sata.
2248 Any ID with matching PORT is used.
2250 * SATA link speed limit: 1.5Gbps or 3.0Gbps.
2252 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7].
2253 udma[/][16,25,33,44,66,100,133] notation is also
2256 * [no]ncq: Turn on or off NCQ.
2258 * [no]ncqtrim: Turn off queued DSM TRIM.
2260 * nohrst, nosrst, norst: suppress hard, soft
2263 * rstonce: only attempt one reset during
2264 hot-unplug link recovery
2266 * dump_id: dump IDENTIFY data.
2268 * atapi_dmadir: Enable ATAPI DMADIR bridge support
2270 * disable: Disable this device.
2272 If there are multiple matching configurations changing
2273 the same attribute, the last one is used.
2275 memblock=debug [KNL] Enable memblock debug messages.
2277 load_ramdisk= [RAM] List of ramdisks to load from floppy
2278 See Documentation/admin-guide/blockdev/ramdisk.rst.
2280 lockd.nlm_grace_period=P [NFS] Assign grace period.
2283 lockd.nlm_tcpport=N [NFS] Assign TCP port.
2286 lockd.nlm_timeout=T [NFS] Assign timeout value.
2289 lockd.nlm_udpport=M [NFS] Assign UDP port.
2292 lockdown= [SECURITY]
2293 { integrity | confidentiality }
2294 Enable the kernel lockdown feature. If set to
2295 integrity, kernel features that allow userland to
2296 modify the running kernel are disabled. If set to
2297 confidentiality, kernel features that allow userland
2298 to extract confidential information from the kernel
2301 locktorture.nreaders_stress= [KNL]
2302 Set the number of locking read-acquisition kthreads.
2303 Defaults to being automatically set based on the
2304 number of online CPUs.
2306 locktorture.nwriters_stress= [KNL]
2307 Set the number of locking write-acquisition kthreads.
2309 locktorture.onoff_holdoff= [KNL]
2310 Set time (s) after boot for CPU-hotplug testing.
2312 locktorture.onoff_interval= [KNL]
2313 Set time (s) between CPU-hotplug operations, or
2314 zero to disable CPU-hotplug testing.
2316 locktorture.shuffle_interval= [KNL]
2317 Set task-shuffle interval (jiffies). Shuffling
2318 tasks allows some CPUs to go into dyntick-idle
2319 mode during the locktorture test.
2321 locktorture.shutdown_secs= [KNL]
2322 Set time (s) after boot system shutdown. This
2323 is useful for hands-off automated testing.
2325 locktorture.stat_interval= [KNL]
2326 Time (s) between statistics printk()s.
2328 locktorture.stutter= [KNL]
2329 Time (s) to stutter testing, for example,
2330 specifying five seconds causes the test to run for
2331 five seconds, wait for five seconds, and so on.
2332 This tests the locking primitive's ability to
2333 transition abruptly to and from idle.
2335 locktorture.torture_type= [KNL]
2336 Specify the locking implementation to test.
2338 locktorture.verbose= [KNL]
2339 Enable additional printk() statements.
2341 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver
2344 loglevel= All Kernel Messages with a loglevel smaller than the
2345 console loglevel will be printed to the console. It can
2346 also be changed with klogd or other programs. The
2347 loglevels are defined as follows:
2349 0 (KERN_EMERG) system is unusable
2350 1 (KERN_ALERT) action must be taken immediately
2351 2 (KERN_CRIT) critical conditions
2352 3 (KERN_ERR) error conditions
2353 4 (KERN_WARNING) warning conditions
2354 5 (KERN_NOTICE) normal but significant condition
2355 6 (KERN_INFO) informational
2356 7 (KERN_DEBUG) debug-level messages
2358 log_buf_len=n[KMG] Sets the size of the printk ring buffer,
2359 in bytes. n must be a power of two and greater
2360 than the minimal size. The minimal size is defined
2361 by LOG_BUF_SHIFT kernel config parameter. There is
2362 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter
2363 that allows to increase the default size depending on
2364 the number of CPUs. See init/Kconfig for more details.
2366 logo.nologo [FB] Disables display of the built-in Linux logo.
2367 This may be used to provide more screen space for
2368 kernel log messages and is useful when debugging
2369 kernel boot problems.
2371 lp=0 [LP] Specify parallel ports to use, e.g,
2372 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses
2373 lp=reset first parallel port). 'lp=0' disables the
2374 lp=auto printer driver. 'lp=reset' (which can be
2375 specified in addition to the ports) causes
2376 attached printers to be reset. Using
2377 lp=port1,port2,... specifies the parallel ports
2378 to associate lp devices with, starting with
2379 lp0. A port specification may be 'none' to skip
2380 that lp device, or a parport name such as
2381 'parport0'. Specifying 'lp=auto' instead of a
2382 port specification list means that device IDs
2383 from each port should be examined, to see if
2384 an IEEE 1284-compliant printer is attached; if
2385 so, the driver will manage that printer.
2386 See also header of drivers/char/lp.c.
2389 Sets loops_per_jiffy to given constant, thus avoiding
2390 time-consuming boot-time autodetection (up to 250 ms per
2391 CPU). 0 enables autodetection (default). To determine
2392 the correct value for your kernel, boot with normal
2393 autodetection and see what value is printed. Note that
2394 on SMP systems the preset will be applied to all CPUs,
2395 which is likely to cause problems if your CPUs need
2396 significantly divergent settings. An incorrect value
2397 will cause delays in the kernel to be wrong, leading to
2398 unpredictable I/O errors and other breakage. Although
2399 unlikely, in the extreme case this might damage your
2403 Format: <io>,<irq>,<dma>
2405 lsm.debug [SECURITY] Enable LSM initialization debugging output.
2408 [SECURITY] Choose order of LSM initialization. This
2409 overrides CONFIG_LSM, and the "security=" parameter.
2411 machvec= [IA-64] Force the use of a particular machine-vector
2412 (machvec) in a generic kernel.
2413 Example: machvec=hpzx1
2415 machtype= [Loongson] Share the same kernel image file between different
2417 Example: machtype=lemote-yeeloong-2f-7inch
2419 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater
2420 than or equal to this physical address is ignored.
2422 maxcpus= [SMP] Maximum number of processors that an SMP kernel
2423 will bring up during bootup. maxcpus=n : n >= 0 limits
2424 the kernel to bring up 'n' processors. Surely after
2425 bootup you can bring up the other plugged cpu by executing
2426 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus
2427 only takes effect during system bootup.
2428 While n=0 is a special case, it is equivalent to "nosmp",
2429 which also disables the IO APIC.
2431 max_loop= [LOOP] The number of loop block devices that get
2432 (loop.max_loop) unconditionally pre-created at init time. The default
2433 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead
2434 of statically allocating a predefined number, loop
2435 devices can be requested on-demand with the
2436 /dev/loop-control interface.
2438 mce [X86-32] Machine Check Exception
2440 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst
2442 md= [HW] RAID subsystems devices and level
2443 See Documentation/admin-guide/md.rst.
2446 Format: <first>,<last>
2447 Specifies range of consoles to be captured by the MDA.
2450 Control mitigation for the Micro-architectural Data
2451 Sampling (MDS) vulnerability.
2453 Certain CPUs are vulnerable to an exploit against CPU
2454 internal buffers which can forward information to a
2455 disclosure gadget under certain conditions.
2457 In vulnerable processors, the speculatively
2458 forwarded data can be used in a cache side channel
2459 attack, to access data to which the attacker does
2460 not have direct access.
2462 This parameter controls the MDS mitigation. The
2465 full - Enable MDS mitigation on vulnerable CPUs
2466 full,nosmt - Enable MDS mitigation and disable
2467 SMT on vulnerable CPUs
2468 off - Unconditionally disable MDS mitigation
2470 Not specifying this option is equivalent to
2473 For details see: Documentation/admin-guide/hw-vuln/mds.rst
2475 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
2476 Amount of memory to be used when the kernel is not able
2477 to see the whole system memory or for test.
2478 [X86] Work as limiting max address. Use together
2479 with memmap= to avoid physical address space collisions.
2480 Without memmap= PCI devices could be placed at addresses
2481 belonging to unused RAM.
2483 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel
2487 [KNL,SH] Allow user to override the default size for
2488 per-device physically contiguous DMA buffers.
2490 memhp_default_state=online/offline
2491 [KNL] Set the initial state for the memory hotplug
2492 onlining policy. If not specified, the default value is
2493 set according to the
2494 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config
2496 See Documentation/admin-guide/mm/memory-hotplug.rst.
2498 memmap=exactmap [KNL,X86] Enable setting of an exact
2499 E820 memory map, as specified by the user.
2500 Such memmap=exactmap lines can be constructed based on
2501 BIOS output or other requirements. See the memmap=nn@ss
2504 memmap=nn[KMG]@ss[KMG]
2505 [KNL] Force usage of a specific region of memory.
2506 Region of memory to be used is from ss to ss+nn.
2507 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG],
2508 which limits max address to nn[KMG].
2509 Multiple different regions can be specified,
2512 memmap=100M@2G,100M#3G,1G!1024G
2514 memmap=nn[KMG]#ss[KMG]
2515 [KNL,ACPI] Mark specific memory as ACPI data.
2516 Region of memory to be marked is from ss to ss+nn.
2518 memmap=nn[KMG]$ss[KMG]
2519 [KNL,ACPI] Mark specific memory as reserved.
2520 Region of memory to be reserved is from ss to ss+nn.
2521 Example: Exclude memory from 0x18690000-0x1869ffff
2522 memmap=64K$0x18690000
2524 memmap=0x10000$0x18690000
2525 Some bootloaders may need an escape character before '$',
2526 like Grub2, otherwise '$' and the following number
2529 memmap=nn[KMG]!ss[KMG]
2530 [KNL,X86] Mark specific memory as protected.
2531 Region of memory to be used, from ss to ss+nn.
2532 The memory region may be marked as e820 type 12 (0xc)
2533 and is NVDIMM or ADR memory.
2535 memmap=<size>%<offset>-<oldtype>+<newtype>
2536 [KNL,ACPI] Convert memory within the specified region
2537 from <oldtype> to <newtype>. If "-<oldtype>" is left
2538 out, the whole region will be marked as <newtype>,
2539 even if previously unavailable. If "+<newtype>" is left
2540 out, matching memory will be removed. Types are
2541 specified as e820 types, e.g., 1 = RAM, 2 = reserved,
2542 3 = ACPI, 12 = PRAM.
2544 memory_corruption_check=0/1 [X86]
2545 Some BIOSes seem to corrupt the first 64k of
2546 memory when doing things like suspend/resume.
2547 Setting this option will scan the memory
2548 looking for corruption. Enabling this will
2549 both detect corruption and prevent the kernel
2550 from using the memory being corrupted.
2551 However, its intended as a diagnostic tool; if
2552 repeatable BIOS-originated corruption always
2553 affects the same memory, you can use memmap=
2554 to prevent the kernel from using that memory.
2556 memory_corruption_check_size=size [X86]
2557 By default it checks for corruption in the low
2558 64k, making this memory unavailable for normal
2559 use. Use this parameter to scan for
2560 corruption in more or less memory.
2562 memory_corruption_check_period=seconds [X86]
2563 By default it checks for corruption every 60
2564 seconds. Use this parameter to check at some
2565 other rate. 0 disables periodic checking.
2567 memtest= [KNL,X86,ARM,PPC] Enable memtest
2569 default : 0 <disable>
2570 Specifies the number of memtest passes to be
2571 performed. Each pass selects another test
2572 pattern from a given set of patterns. Memtest
2573 fills the memory with this pattern, validates
2574 memory contents and reserves bad memory
2575 regions that are detected.
2577 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control
2578 Valid arguments: on, off
2579 Default (depends on kernel configuration option):
2580 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y)
2581 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n)
2582 mem_encrypt=on: Activate SME
2583 mem_encrypt=off: Do not activate SME
2585 Refer to Documentation/virt/kvm/amd-memory-encryption.rst
2586 for details on when memory encryption can be activated.
2588 mem_sleep_default= [SUSPEND] Default system suspend mode:
2589 s2idle - Suspend-To-Idle
2590 shallow - Power-On Suspend or equivalent (if supported)
2591 deep - Suspend-To-RAM or equivalent (if supported)
2592 See Documentation/admin-guide/pm/sleep-states.rst.
2594 meye.*= [HW] Set MotionEye Camera parameters
2595 See Documentation/media/v4l-drivers/meye.rst.
2597 mfgpt_irq= [IA-32] Specify the IRQ to use for the
2598 Multi-Function General Purpose Timers on AMD Geode
2601 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when
2602 the BIOS has incorrectly applied a workaround. TinyBIOS
2603 version 0.98 is known to be affected, 0.99 fixes the
2604 problem by letting the user disable the workaround.
2608 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this
2609 physical address is ignored.
2611 mini2440= [ARM,HW,KNL]
2612 Format:[0..2][b][c][t]
2614 MINI2440 configuration specification:
2615 0 - The attached screen is the 3.5" TFT
2616 1 - The attached screen is the 7" TFT
2617 2 - The VGA Shield is attached (1024x768)
2618 Leaving out the screen size parameter will not load
2619 the TFT driver, and the framebuffer will be left
2621 b - Enable backlight. The TFT backlight pin will be
2622 linked to the kernel VESA blanking code and a GPIO
2623 LED. This parameter is not necessary when using the
2625 c - Enable the s3c camera interface.
2626 t - Reserved for enabling touchscreen support. The
2627 touchscreen support is not enabled in the mainstream
2628 kernel as of 2.6.30, a preliminary port can be found
2629 in the "bleeding edge" mini2440 support kernel at
2630 http://repo.or.cz/w/linux-2.6/mini2440.git
2633 [X86,PPC,S390,ARM64] Control optional mitigations for
2634 CPU vulnerabilities. This is a set of curated,
2635 arch-independent options, each of which is an
2636 aggregation of existing arch-specific options.
2639 Disable all optional CPU mitigations. This
2640 improves system performance, but it may also
2641 expose users to several CPU vulnerabilities.
2642 Equivalent to: nopti [X86,PPC]
2644 nospectre_v1 [X86,PPC]
2646 nospectre_v2 [X86,PPC,S390,ARM64]
2647 spectre_v2_user=off [X86]
2648 spec_store_bypass_disable=off [X86,PPC]
2649 ssbd=force-off [ARM64]
2652 tsx_async_abort=off [X86]
2653 kvm.nx_huge_pages=off [X86]
2656 This does not have any effect on
2657 kvm.nx_huge_pages when
2658 kvm.nx_huge_pages=force.
2661 Mitigate all CPU vulnerabilities, but leave SMT
2662 enabled, even if it's vulnerable. This is for
2663 users who don't want to be surprised by SMT
2664 getting disabled across kernel upgrades, or who
2665 have other ways of avoiding SMT-based attacks.
2666 Equivalent to: (default behavior)
2669 Mitigate all CPU vulnerabilities, disabling SMT
2670 if needed. This is for users who always want to
2671 be fully mitigated, even if it means losing SMT.
2672 Equivalent to: l1tf=flush,nosmt [X86]
2673 mds=full,nosmt [X86]
2674 tsx_async_abort=full,nosmt [X86]
2677 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this
2678 parameter allows control of the logging verbosity for
2679 the additional memory initialisation checks. A value
2680 of 0 disables mminit logging and a level of 4 will
2681 log everything. Information is printed at KERN_DEBUG
2682 so loglevel=8 may also need to be specified.
2685 [KNL] When CONFIG_MODULE_SIG is set, this means that
2686 modules without (valid) signatures will fail to load.
2687 Note that if CONFIG_MODULE_SIG_FORCE is set, that
2688 is always true, so this option does nothing.
2690 module_blacklist= [KNL] Do not load a comma-separated list of
2691 modules. Useful for debugging problem modules.
2694 [MOUSE] Maximum time between finger touching and
2695 leaving touchpad surface for touch to be considered
2696 a tap and be reported as a left button click (for
2697 touchpads working in absolute mode only).
2699 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices
2700 reporting absolute coordinates, such as tablets
2701 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices
2702 reporting absolute coordinates, such as tablets
2704 movablecore= [KNL,X86,IA-64,PPC]
2705 Format: nn[KMGTPE] | nn%
2706 This parameter is the complement to kernelcore=, it
2707 specifies the amount of memory used for migratable
2708 allocations. If both kernelcore and movablecore is
2709 specified, then kernelcore will be at *least* the
2710 specified value but may be more. If movablecore on its
2711 own is specified, the administrator must be careful
2712 that the amount of memory usable for all allocations
2715 movable_node [KNL] Boot-time switch to make hotplugable memory
2716 NUMA nodes to be movable. This means that the memory
2717 of such nodes will be usable only for movable
2718 allocations which rules out almost all kernel
2719 allocations. Use with caution!
2721 MTD_Partition= [MTD]
2722 Format: <name>,<region-number>,<size>,<offset>
2724 MTD_Region= [MTD] Format:
2725 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>]
2728 See drivers/mtd/cmdlinepart.c.
2730 multitce=off [PPC] This parameter disables the use of the pSeries
2731 firmware feature for updating multiple TCE entries
2734 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration
2736 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock]
2738 boundary - index of last SLC block on Flex-OneNAND.
2739 The remaining blocks are configured as MLC blocks.
2740 lock - Configure if Flex-OneNAND boundary should be locked.
2741 Once locked, the boundary cannot be changed.
2742 1 indicates lock status, 0 indicates unlock status.
2745 ARM/S3C2412 JIVE boot control
2747 See arch/arm/mach-s3c2412/mach-jive.c
2749 mtouchusb.raw_coordinates=
2750 [HW] Make the MicroTouch USB driver use raw coordinates
2751 ('y', default) or cooked coordinates ('n')
2753 mtrr_chunk_size=nn[KMG] [X86]
2754 used for mtrr cleanup. It is largest continuous chunk
2755 that could hold holes aka. UC entries.
2757 mtrr_gran_size=nn[KMG] [X86]
2758 Used for mtrr cleanup. It is granularity of mtrr block.
2760 Large value could prevent small alignment from
2763 mtrr_spare_reg_nr=n [X86]
2765 Range: 0,7 : spare reg number
2767 Used for mtrr cleanup. It is spare mtrr entries number.
2768 Set to 2 or more if your graphical card needs more.
2770 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card
2772 netdev= [NET] Network devices parameters
2773 Format: <irq>,<io>,<mem_start>,<mem_end>,<name>
2774 Note that mem_start is often overloaded to mean
2775 something different and driver-specific.
2776 This usage is only documented in each driver source
2780 [NETFILTER] Enable connection tracking flow accounting
2781 0 to disable accounting
2782 1 to enable accounting
2785 nfsaddrs= [NFS] Deprecated. Use ip= instead.
2786 See Documentation/filesystems/nfs/nfsroot.txt.
2788 nfsroot= [NFS] nfs root filesystem for disk-less boxes.
2789 See Documentation/filesystems/nfs/nfsroot.txt.
2791 nfsrootdebug [NFS] enable nfsroot debugging messages.
2792 See Documentation/filesystems/nfs/nfsroot.txt.
2794 nfs.callback_nr_threads=
2795 [NFSv4] set the total number of threads that the
2796 NFS client will assign to service NFSv4 callback
2799 nfs.callback_tcpport=
2800 [NFS] set the TCP port on which the NFSv4 callback
2801 channel should listen.
2804 [NFS] sets the pathname to the program which is used
2805 to update the NFS client cache entries.
2807 nfs.cache_getent_timeout=
2808 [NFS] sets the timeout after which an attempt to
2809 update a cache entry is deemed to have failed.
2811 nfs.idmap_cache_timeout=
2812 [NFS] set the maximum lifetime for idmapper cache
2816 [NFS] enable 64-bit inode numbers.
2817 If zero, the NFS client will fake up a 32-bit inode
2818 number for the readdir() and stat() syscalls instead
2819 of returning the full 64-bit number.
2820 The default is to return 64-bit inode numbers.
2822 nfs.max_session_cb_slots=
2823 [NFSv4.1] Sets the maximum number of session
2824 slots the client will assign to the callback
2825 channel. This determines the maximum number of
2826 callbacks the client will process in parallel for
2827 a particular server.
2829 nfs.max_session_slots=
2830 [NFSv4.1] Sets the maximum number of session slots
2831 the client will attempt to negotiate with the server.
2832 This limits the number of simultaneous RPC requests
2833 that the client can send to the NFSv4.1 server.
2834 Note that there is little point in setting this
2835 value higher than the max_tcp_slot_table_limit.
2837 nfs.nfs4_disable_idmapping=
2838 [NFSv4] When set to the default of '1', this option
2839 ensures that both the RPC level authentication
2840 scheme and the NFS level operations agree to use
2841 numeric uids/gids if the mount is using the
2842 'sec=sys' security flavour. In effect it is
2843 disabling idmapping, which can make migration from
2844 legacy NFSv2/v3 systems to NFSv4 easier.
2845 Servers that do not support this mode of operation
2846 will be autodetected by the client, and it will fall
2847 back to using the idmapper.
2848 To turn off this behaviour, set the value to '0'.
2850 [NFS4] Specify an additional fixed unique ident-
2851 ification string that NFSv4 clients can insert into
2852 their nfs_client_id4 string. This is typically a
2853 UUID that is generated at system install time.
2855 nfs.send_implementation_id =
2856 [NFSv4.1] Send client implementation identification
2857 information in exchange_id requests.
2858 If zero, no implementation identification information
2860 The default is to send the implementation identification
2863 nfs.recover_lost_locks =
2864 [NFSv4] Attempt to recover locks that were lost due
2865 to a lease timeout on the server. Please note that
2866 doing this risks data corruption, since there are
2867 no guarantees that the file will remain unchanged
2868 after the locks are lost.
2869 If you want to enable the kernel legacy behaviour of
2870 attempting to recover these locks, then set this
2872 The default parameter value of '0' causes the kernel
2873 not to attempt recovery of lost locks.
2875 nfs4.layoutstats_timer =
2876 [NFSv4.2] Change the rate at which the kernel sends
2877 layoutstats to the pNFS metadata server.
2879 Setting this to value to 0 causes the kernel to use
2880 whatever value is the default set by the layout
2881 driver. A non-zero value sets the minimum interval
2882 in seconds between layoutstats transmissions.
2884 nfsd.nfs4_disable_idmapping=
2885 [NFSv4] When set to the default of '1', the NFSv4
2886 server will return only numeric uids and gids to
2887 clients using auth_sys, and will accept numeric uids
2888 and gids from such clients. This is intended to ease
2889 migration from NFSv2/v3.
2891 nmi_debug= [KNL,SH] Specify one or more actions to take
2892 when a NMI is triggered.
2893 Format: [state][,regs][,debounce][,die]
2895 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
2896 Format: [panic,][nopanic,][num]
2898 0 - turn hardlockup detector in nmi_watchdog off
2899 1 - turn hardlockup detector in nmi_watchdog on
2900 When panic is specified, panic when an NMI watchdog
2901 timeout occurs (or 'nopanic' to not panic on an NMI
2902 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
2903 To disable both hard and soft lockup detectors,
2904 please see 'nowatchdog'.
2905 This is useful when you use a panic=... timeout and
2906 need the box quickly up again.
2908 These settings can be accessed at runtime via
2909 the nmi_watchdog and hardlockup_panic sysctls.
2911 netpoll.carrier_timeout=
2912 [NET] Specifies amount of time (in seconds) that
2913 netpoll should wait for a carrier. By default netpoll
2916 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths
2917 emulation library even if a 387 maths coprocessor
2920 no5lvl [X86-64] Disable 5-level paging mode. Forces
2921 kernel to use 4-level paging instead.
2924 [HW] Never suspend the console
2925 Disable suspending of consoles during suspend and
2926 hibernate operations. Once disabled, debugging
2927 messages can reach various consoles while the rest
2928 of the system is being put to sleep (ie, while
2929 debugging driver suspend/resume hooks). This may
2930 not work reliably with all consoles, but is known
2931 to work with serial and VGA consoles.
2932 To facilitate more flexible debugging, we also add
2933 console_suspend, a printk module parameter to control
2934 it. Users could use console_suspend (usually
2935 /sys/module/printk/parameters/console_suspend) to
2936 turn on/off it dynamically.
2938 novmcoredd [KNL,KDUMP]
2939 Disable device dump. Device dump allows drivers to
2940 append dump data to vmcore so you can collect driver
2941 specified debug info. Drivers can append the data
2942 without any limit and this data is stored in memory,
2943 so this may cause significant memory stress. Disabling
2944 device dump can help save memory but the driver debug
2945 data will be no longer available. This parameter
2946 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP
2949 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien
2950 caches in the slab allocator. Saves per-node memory,
2951 but will impact performance.
2955 noaltinstr [S390] Disables alternative instructions patching
2956 (CPU alternatives feature).
2958 noapic [SMP,APIC] Tells the kernel to not make use of any
2959 IOAPICs that may be present in the system.
2961 noautogroup Disable scheduler automatic task group creation.
2963 nobats [PPC] Do not use BATs for mapping kernel lowmem
2964 on "Classic" PPC cores.
2968 noclflush [BUGS=X86] Don't use the CLFLUSH instruction
2970 nodelayacct [KNL] Disable per-task delay accounting
2972 nodsp [SH] Disable hardware DSP at boot time.
2974 noefi Disable EFI runtime services support.
2979 On X86-32 available only on PAE configured kernels.
2980 noexec=on: enable non-executable mappings (default)
2981 noexec=off: disable non-executable mappings
2984 Disable SMAP (Supervisor Mode Access Prevention)
2985 even if it is supported by processor.
2988 Disable SMEP (Supervisor Mode Execution Prevention)
2989 even if it is supported by processor.
2992 This affects only 32-bit executables.
2993 noexec32=on: enable non-executable mappings (default)
2994 read doesn't imply executable mappings
2995 noexec32=off: disable non-executable mappings
2996 read implies executable mappings
2998 nofpu [MIPS,SH] Disable hardware FPU at boot time.
3000 nofxsr [BUGS=X86-32] Disables x86 floating point extended
3001 register save and restore. The kernel will only save
3002 legacy floating-point registers on task switch.
3004 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings.
3006 nosmt [KNL,S390] Disable symmetric multithreading (SMT).
3007 Equivalent to smt=1.
3009 [KNL,x86] Disable symmetric multithreading (SMT).
3010 nosmt=force: Force disable SMT, cannot be undone
3011 via the sysfs control file.
3013 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1
3014 (bounds check bypass). With this option data leaks are
3015 possible in the system.
3017 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for
3018 the Spectre variant 2 (indirect branch prediction)
3019 vulnerability. System may allow data leaks with this
3022 nospec_store_bypass_disable
3023 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability
3025 noxsave [BUGS=X86] Disables x86 extended register state save
3026 and restore using xsave. The kernel will fallback to
3027 enabling legacy floating-point and sse state.
3029 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended
3030 register states. The kernel will fall back to use
3031 xsave to save the states. By using this parameter,
3032 performance of saving the states is degraded because
3033 xsave doesn't support modified optimization while
3034 xsaveopt supports it on xsaveopt enabled systems.
3036 noxsaves [X86] Disables xsaves and xrstors used in saving and
3037 restoring x86 extended register state in compacted
3038 form of xsave area. The kernel will fall back to use
3039 xsaveopt and xrstor to save and restore the states
3040 in standard form of xsave area. By using this
3041 parameter, xsave area per process might occupy more
3042 memory on xsaves enabled systems.
3044 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or
3045 wfi(ARM) instruction doesn't work correctly and not to
3046 use it. This is also useful when using JTAG debugger.
3048 no_file_caps Tells the kernel not to honor file capabilities. The
3049 only way then for a file to be executed with privilege
3050 is to be setuid root or executed by root.
3052 nohalt [IA-64] Tells the kernel not to use the power saving
3053 function PAL_HALT_LIGHT when idle. This increases
3054 power-consumption. On the positive side, it reduces
3055 interrupt wake-up latency, which may improve performance
3056 in certain environments such as networked servers or
3059 nohibernate [HIBERNATION] Disable hibernation and resume.
3061 nohz= [KNL] Boottime enable/disable dynamic ticks
3062 Valid arguments: on, off
3065 nohz_full= [KNL,BOOT,SMP,ISOL]
3066 The argument is a cpu list, as described above.
3067 In kernels built with CONFIG_NO_HZ_FULL=y, set
3068 the specified list of CPUs whose tick will be stopped
3069 whenever possible. The boot CPU will be forced outside
3070 the range to maintain the timekeeping. Any CPUs
3071 in this list will have their RCU callbacks offloaded,
3072 just as if they had also been called out in the
3073 rcu_nocbs= boot parameter.
3075 noiotrap [SH] Disables trapped I/O port accesses.
3077 noirqdebug [X86-32] Disables the code which attempts to detect and
3078 disable unhandled interrupt sources.
3080 no_timer_check [X86,APIC] Disables the code which tests for
3081 broken timer IRQ sources.
3083 noisapnp [ISAPNP] Disables ISA PnP code.
3085 noinitrd [RAM] Tells the kernel not to load any configured
3088 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt
3090 [Deprecated - use intremap=off]
3094 noinvpcid [X86] Disable the INVPCID cpu feature.
3096 nojitter [IA-64] Disables jitter checking for ITC timers.
3098 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver
3100 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page
3104 [X86,PV_OPS] Disable paravirtualized VMware scheduler
3105 clock and use the default one.
3107 no-steal-acc [X86,KVM] Disable paravirtualized steal time accounting.
3108 steal time is computed, but won't influence scheduler
3111 nolapic [X86-32,APIC] Do not enable or use the local APIC.
3113 nolapic_timer [X86-32,APIC] Do not use the local APIC timer.
3115 noltlbs [PPC] Do not use large page/tlb entries for kernel
3116 lowmem mapping on PPC40x and PPC8xx
3118 nomca [IA-64] Disable machine check abort handling
3120 nomce [X86-32] Disable Machine Check Exception
3122 nomfgpt [X86-32] Disable Multi-Function General Purpose
3123 Timer usage (for AMD Geode machines).
3125 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to
3126 shutdown the other cpus. Instead use the REBOOT_VECTOR
3129 nomodule Disable module load
3131 nopat [X86] Disable PAT (page attribute table extension of
3132 pagetables) support.
3134 nopcid [X86-64] Disable the PCID cpu feature.
3136 norandmaps Don't use address space randomization. Equivalent to
3137 echo 0 > /proc/sys/kernel/randomize_va_space
3139 noreplace-smp [X86-32,SMP] Don't replace SMP instructions
3140 with UP alternatives
3142 nordrand [X86] Disable kernel use of the RDRAND and
3143 RDSEED instructions even if they are supported
3144 by the processor. RDRAND and RDSEED are still
3145 available to user space applications.
3147 noresume [SWSUSP] Disables resume and restores original swap
3150 no-scroll [VGA] Disables scrollback.
3151 This is required for the Braillex ib80-piezo Braille
3152 reader made by F.H. Papenmeier (Germany).
3156 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support.
3158 nosmp [SMP] Tells an SMP kernel to act as a UP kernel,
3159 and disable the IO APIC. legacy for "maxcpus=0".
3161 nosoftlockup [KNL] Disable the soft-lockup detector.
3163 nosync [HW,M68K] Disables sync negotiation for all devices.
3165 nowatchdog [KNL] Disable both lockup detectors, i.e.
3166 soft-lockup and NMI watchdog (hard-lockup).
3170 nox2apic [X86-64,APIC] Do not enable x2APIC mode.
3172 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when
3173 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off.
3174 Some features depend on CPU0. Known dependencies are:
3175 1. Resume from suspend/hibernate depends on CPU0.
3176 Suspend/hibernate will fail if CPU0 is offline and you
3177 need to online CPU0 before suspend/hibernate.
3178 2. PIC interrupts also depend on CPU0. CPU0 can't be
3179 removed if a PIC interrupt is detected.
3180 It's said poweroff/reboot may depend on CPU0 on some
3181 machines although I haven't seen such issues so far
3182 after CPU0 is offline on a few tested machines.
3183 If the dependencies are under your control, you can
3184 turn on cpu0_hotplug.
3186 nps_mtm_hs_ctr= [KNL,ARC]
3187 This parameter sets the maximum duration, in
3188 cycles, each HW thread of the CTOP can run
3189 without interruptions, before HW switches it.
3190 The actual maximum duration is 16 times this
3192 Format: integer between 1 and 255
3195 nptcg= [IA-64] Override max number of concurrent global TLB
3196 purges which is reported from either PAL_VM_SUMMARY or
3199 nr_cpus= [SMP] Maximum number of processors that an SMP kernel
3200 could support. nr_cpus=n : n >= 1 limits the kernel to
3201 support 'n' processors. It could be larger than the
3202 number of already plugged CPU during bootup, later in
3203 runtime you can physically add extra cpu until it reaches
3204 n. So during boot up some boot time memory for per-cpu
3205 variables need be pre-allocated for later physical cpu
3208 nr_uarts= [SERIAL] maximum number of UARTs to be registered.
3210 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing.
3211 Allowed values are enable and disable
3213 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA.
3214 'node', 'default' can be specified
3215 This can be set from sysctl after boot.
3216 See Documentation/admin-guide/sysctl/vm.rst for details.
3218 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver.
3219 See Documentation/debugging-via-ohci1394.txt for more
3222 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands
3223 Rather than timing out after 20 ms if an EC
3224 command is not properly ACKed, override the length
3225 of the timeout. We have interrupts disabled while
3226 waiting for the ACK, so if this is set too high
3227 interrupts *may* be lost!
3229 omap_mux= [OMAP] Override bootloader pin multiplexing.
3230 Format: <mux_mode0.mode_name=value>...
3231 For example, to override I2C bus2:
3232 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100
3234 oprofile.timer= [HW]
3235 Use timer interrupt instead of performance counters
3237 oprofile.cpu_type= Force an oprofile cpu type
3238 This might be useful if you have an older oprofile
3239 userland or if you want common events.
3240 Format: { arch_perfmon }
3241 arch_perfmon: [X86] Force use of architectural
3242 perfmon on Intel CPUs instead of the
3243 CPU specific event set.
3244 timer: [X86] Force use of architectural NMI
3245 timer mode (see also oprofile.timer
3246 for generic hr timer mode)
3248 oops=panic Always panic on oopses. Default is to just kill the
3249 process, but there is a small probability of
3250 deadlocking the machine.
3251 This will also cause panics on machine check exceptions.
3252 Useful together with panic=30 to trigger a reboot.
3255 [KNL] Boolean flag to control whether the page allocator
3256 should randomize its free lists. The randomization may
3257 be automatically enabled if the kernel detects it is
3258 running on a platform with a direct-mapped memory-side
3259 cache, and this parameter can be used to
3260 override/disable that behavior. The state of the flag
3261 can be read from sysfs at:
3262 /sys/module/page_alloc/parameters/shuffle.
3264 page_owner= [KNL] Boot-time page_owner enabling option.
3265 Storage of the information about who allocated
3266 each page is disabled in default. With this switch,
3268 on: enable the feature
3270 page_poison= [KNL] Boot-time parameter changing the state of
3271 poisoning on the buddy allocator, available with
3272 CONFIG_PAGE_POISONING=y.
3273 off: turn off poisoning (default)
3274 on: turn on poisoning
3276 panic= [KNL] Kernel behaviour on panic: delay <timeout>
3277 timeout > 0: seconds before rebooting
3278 timeout = 0: wait forever
3279 timeout < 0: reboot immediately
3282 panic_print= Bitmask for printing system info when panic happens.
3283 User can chose combination of the following bits:
3284 bit 0: print all tasks info
3285 bit 1: print system memory info
3286 bit 2: print timer info
3287 bit 3: print locks info if CONFIG_LOCKDEP is on
3288 bit 4: print ftrace buffer
3289 bit 5: print all printk messages in buffer
3291 panic_on_warn panic() instead of WARN(). Useful to cause kdump
3294 crash_kexec_post_notifiers
3295 Run kdump after running panic-notifiers and dumping
3296 kmsg. This only for the users who doubt kdump always
3297 succeeds in any situation.
3298 Note that this also increases risks of kdump failure,
3299 because some panic notifiers can make the crashed
3300 kernel more unstable.
3302 parkbd.port= [HW] Parallel port number the keyboard adapter is
3303 connected to, default is 0.
3305 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation,
3306 0 for XT, 1 for AT (default is AT).
3309 parport= [HW,PPT] Specify parallel ports. 0 disables.
3310 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] }
3311 Use 'auto' to force the driver to use any
3312 IRQ/DMA settings detected (the default is to
3313 ignore detected IRQ/DMA settings because of
3314 possible conflicts). You can specify the base
3315 address, IRQ, and DMA settings; IRQ and DMA
3316 should be numbers, or 'auto' (for using detected
3317 settings on that particular port), or 'nofifo'
3318 (to avoid using a FIFO even if it is detected).
3319 Parallel ports are assigned in the order they
3320 are specified on the command line, starting
3323 parport_init_mode= [HW,PPT]
3324 Configure VIA parallel port to operate in
3325 a specific mode. This is necessary on Pegasos
3326 computer where firmware has no options for setting
3327 up parallel port mode and sets it to spp.
3328 Currently this function knows 686a and 8231 chips.
3329 Format: [spp|ps2|epp|ecp|ecpepp]
3332 Halt all CPUs after the first oops has been printed for
3333 the specified number of seconds. This is to be used if
3334 your oopses keep scrolling off the screen.
3339 See header of drivers/block/paride/pcd.c.
3340 See also Documentation/admin-guide/blockdev/paride.rst.
3342 pci=option[,option...] [PCI] various PCI subsystem options.
3344 Some options herein operate on a specific device
3345 or a set of devices (<pci_dev>). These are
3346 specified in one of the following formats:
3348 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]*
3349 pci:<vendor>:<device>[:<subvendor>:<subdevice>]
3351 Note: the first format specifies a PCI
3352 bus/device/function address which may change
3353 if new hardware is inserted, if motherboard
3354 firmware changes, or due to changes caused
3355 by other kernel parameters. If the
3356 domain is left unspecified, it is
3357 taken to be zero. Optionally, a path
3358 to a device through multiple device/function
3359 addresses can be specified after the base
3360 address (this is more robust against
3361 renumbering issues). The second format
3362 selects devices using IDs from the
3363 configuration space which may match multiple
3364 devices in the system.
3366 earlydump dump PCI config space before the kernel
3368 off [X86] don't probe for the PCI bus
3369 bios [X86-32] force use of PCI BIOS, don't access
3370 the hardware directly. Use this if your machine
3371 has a non-standard PCI host bridge.
3372 nobios [X86-32] disallow use of PCI BIOS, only direct
3373 hardware access methods are allowed. Use this
3374 if you experience crashes upon bootup and you
3375 suspect they are caused by the BIOS.
3376 conf1 [X86] Force use of PCI Configuration Access
3377 Mechanism 1 (config address in IO port 0xCF8,
3378 data in IO port 0xCFC, both 32-bit).
3379 conf2 [X86] Force use of PCI Configuration Access
3380 Mechanism 2 (IO port 0xCF8 is an 8-bit port for
3381 the function, IO port 0xCFA, also 8-bit, sets
3382 bus number. The config space is then accessed
3383 through ports 0xC000-0xCFFF).
3384 See http://wiki.osdev.org/PCI for more info
3385 on the configuration access mechanisms.
3386 noaer [PCIE] If the PCIEAER kernel config parameter is
3387 enabled, this kernel boot option can be used to
3388 disable the use of PCIE advanced error reporting.
3389 nodomains [PCI] Disable support for multiple PCI
3390 root domains (aka PCI segments, in ACPI-speak).
3391 nommconf [X86] Disable use of MMCONFIG for PCI
3393 check_enable_amd_mmconf [X86] check for and enable
3394 properly configured MMIO access to PCI
3395 config space on AMD family 10h CPU
3396 nomsi [MSI] If the PCI_MSI kernel config parameter is
3397 enabled, this kernel boot option can be used to
3398 disable the use of MSI interrupts system-wide.
3399 noioapicquirk [APIC] Disable all boot interrupt quirks.
3400 Safety option to keep boot IRQs enabled. This
3401 should never be necessary.
3402 ioapicreroute [APIC] Enable rerouting of boot IRQs to the
3403 primary IO-APIC for bridges that cannot disable
3404 boot IRQs. This fixes a source of spurious IRQs
3405 when the system masks IRQs.
3406 noioapicreroute [APIC] Disable workaround that uses the
3407 boot IRQ equivalent of an IRQ that connects to
3408 a chipset where boot IRQs cannot be disabled.
3409 The opposite of ioapicreroute.
3410 biosirq [X86-32] Use PCI BIOS calls to get the interrupt
3411 routing table. These calls are known to be buggy
3412 on several machines and they hang the machine
3413 when used, but on other computers it's the only
3414 way to get the interrupt routing table. Try
3415 this option if the kernel is unable to allocate
3416 IRQs or discover secondary PCI buses on your
3418 rom [X86] Assign address space to expansion ROMs.
3419 Use with caution as certain devices share
3420 address decoders between ROMs and other
3422 norom [X86] Do not assign address space to
3423 expansion ROMs that do not already have
3424 BIOS assigned address ranges.
3425 nobar [X86] Do not assign address space to the
3426 BARs that weren't assigned by the BIOS.
3427 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be
3428 assigned automatically to PCI devices. You can
3429 make the kernel exclude IRQs of your ISA cards
3431 pirqaddr=0xAAAAA [X86] Specify the physical address
3432 of the PIRQ table (normally generated
3433 by the BIOS) if it is outside the
3434 F0000h-100000h range.
3435 lastbus=N [X86] Scan all buses thru bus #N. Can be
3436 useful if the kernel is unable to find your
3437 secondary buses and you want to tell it
3438 explicitly which ones they are.
3439 assign-busses [X86] Always assign all PCI bus
3440 numbers ourselves, overriding
3441 whatever the firmware may have done.
3442 usepirqmask [X86] Honor the possible IRQ mask stored
3443 in the BIOS $PIR table. This is needed on
3444 some systems with broken BIOSes, notably
3445 some HP Pavilion N5400 and Omnibook XE3
3446 notebooks. This will have no effect if ACPI
3447 IRQ routing is enabled.
3448 noacpi [X86] Do not use ACPI for IRQ routing
3449 or for PCI scanning.
3450 use_crs [X86] Use PCI host bridge window information
3451 from ACPI. On BIOSes from 2008 or later, this
3452 is enabled by default. If you need to use this,
3453 please report a bug.
3454 nocrs [X86] Ignore PCI host bridge windows from ACPI.
3455 If you need to use this, please report a bug.
3456 routeirq Do IRQ routing for all PCI devices.
3457 This is normally done in pci_enable_device(),
3458 so this option is a temporary workaround
3459 for broken drivers that don't call it.
3460 skip_isa_align [X86] do not align io start addr, so can
3461 handle more pci cards
3462 noearly [X86] Don't do any early type 1 scanning.
3463 This might help on some broken boards which
3464 machine check when some devices' config space
3465 is read. But various workarounds are disabled
3466 and some IOMMU drivers will not work.
3467 bfsort Sort PCI devices into breadth-first order.
3468 This sorting is done to get a device
3469 order compatible with older (<= 2.4) kernels.
3470 nobfsort Don't sort PCI devices into breadth-first order.
3471 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size)
3472 tuning and use the BIOS-configured MPS defaults.
3473 pcie_bus_safe Set every device's MPS to the largest value
3474 supported by all devices below the root complex.
3475 pcie_bus_perf Set device MPS to the largest allowable MPS
3476 based on its parent bus. Also set MRRS (Max
3477 Read Request Size) to the largest supported
3478 value (no larger than the MPS that the device
3479 or bus can support) for best performance.
3480 pcie_bus_peer2peer Set every device's MPS to 128B, which
3481 every device is guaranteed to support. This
3482 configuration allows peer-to-peer DMA between
3483 any pair of devices, possibly at the cost of
3484 reduced performance. This also guarantees
3485 that hot-added devices will work.
3486 cbiosize=nn[KMG] The fixed amount of bus space which is
3487 reserved for the CardBus bridge's IO window.
3488 The default value is 256 bytes.
3489 cbmemsize=nn[KMG] The fixed amount of bus space which is
3490 reserved for the CardBus bridge's memory
3491 window. The default value is 64 megabytes.
3494 [<order of align>@]<pci_dev>[; ...]
3495 Specifies alignment and device to reassign
3496 aligned memory resources. How to
3497 specify the device is described above.
3498 If <order of align> is not specified,
3499 PAGE_SIZE is used as alignment.
3500 A PCI-PCI bridge can be specified if resource
3501 windows need to be expanded.
3502 To specify the alignment for several
3503 instances of a device, the PCI vendor,
3504 device, subvendor, and subdevice may be
3505 specified, e.g., 12@pci:8086:9c22:103c:198f
3506 for 4096-byte alignment.
3507 ecrc= Enable/disable PCIe ECRC (transaction layer
3508 end-to-end CRC checking).
3509 bios: Use BIOS/firmware settings. This is the
3513 hpiosize=nn[KMG] The fixed amount of bus space which is
3514 reserved for hotplug bridge's IO window.
3515 Default size is 256 bytes.
3516 hpmemsize=nn[KMG] The fixed amount of bus space which is
3517 reserved for hotplug bridge's memory window.
3518 Default size is 2 megabytes.
3519 hpbussize=nn The minimum amount of additional bus numbers
3520 reserved for buses below a hotplug bridge.
3522 realloc= Enable/disable reallocating PCI bridge resources
3523 if allocations done by BIOS are too small to
3524 accommodate resources required by all child
3526 off: Turn realloc off
3528 realloc same as realloc=on
3529 noari do not use PCIe ARI.
3530 noats [PCIE, Intel-IOMMU, AMD-IOMMU]
3531 do not use PCIe ATS (and IOMMU device IOTLB).
3532 pcie_scan_all Scan all possible PCIe devices. Otherwise we
3533 only look for one device below a PCIe downstream
3535 big_root_window Try to add a big 64bit memory window to the PCIe
3536 root complex on AMD CPUs. Some GFX hardware
3537 can resize a BAR to allow access to all VRAM.
3538 Adding the window is slightly risky (it may
3539 conflict with unreported devices), so this
3541 disable_acs_redir=<pci_dev>[; ...]
3542 Specify one or more PCI devices (in the format
3543 specified above) separated by semicolons.
3544 Each device specified will have the PCI ACS
3545 redirect capabilities forced off which will
3546 allow P2P traffic between devices through
3547 bridges without forcing it upstream. Note:
3548 this removes isolation between devices and
3549 may put more devices in an IOMMU group.
3550 force_floating [S390] Force usage of floating interrupts.
3551 nomio [S390] Do not use MIO instructions.
3553 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power
3556 force Enable ASPM even on devices that claim not to support it.
3557 WARNING: Forcing ASPM on may cause system lockups.
3559 pcie_ports= [PCIE] PCIe port services handling:
3560 native Use native PCIe services (PME, AER, DPC, PCIe hotplug)
3561 even if the platform doesn't give the OS permission to
3562 use them. This may cause conflicts if the platform
3563 also tries to use these services.
3564 compat Disable native PCIe services (PME, AER, DPC, PCIe
3567 pcie_port_pm= [PCIE] PCIe port power management handling:
3568 off Disable power management of all PCIe ports
3569 force Forcibly enable power management of all PCIe ports
3571 pcie_pme= [PCIE,PM] Native PCIe PME signaling options:
3572 nomsi Do not use MSI for native PCIe PME signaling (this makes
3573 all PCIe root ports use INTx for all services).
3575 pcmv= [HW,PCMCIA] BadgePAD 4
3579 Keep all power-domains already enabled by bootloader on,
3580 even if no driver has claimed them. This is useful
3581 for debug and development, but should not be
3582 needed on a platform with proper driver support.
3585 See Documentation/admin-guide/blockdev/paride.rst.
3587 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at
3590 See arch/parisc/kernel/pdc_chassis.c
3592 percpu_alloc= Select which percpu first chunk allocator to use.
3593 Currently supported values are "embed" and "page".
3594 Archs may support subset or none of the selections.
3595 See comments in mm/percpu.c for details on each
3596 allocator. This parameter is primarily for debugging
3597 and performance comparison.
3600 See Documentation/admin-guide/blockdev/paride.rst.
3603 See Documentation/admin-guide/blockdev/paride.rst.
3605 pirq= [SMP,APIC] Manual mp-table setup
3606 See Documentation/x86/i386/IO-APIC.rst.
3608 plip= [PPT,NET] Parallel port network link
3609 Format: { parport<nr> | timid | 0 }
3610 See also Documentation/admin-guide/parport.rst.
3612 pmtmr= [X86] Manual setup of pmtmr I/O Port.
3613 Override pmtimer IOPort with a hex value.
3617 Enable PNP debug messages (depends on the
3618 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time
3619 via /sys/module/pnp/parameters/debug. We always show
3620 current resource usage; turning this on also shows
3621 possible settings and some assignment information.
3627 { on | off | curr | res | no-curr | no-res }
3630 [ISAPNP] Exclude IRQs for the autoconfiguration
3633 [ISAPNP] Exclude DMAs for the autoconfiguration
3635 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration
3636 Ranges are in pairs (I/O port base and size).
3639 [ISAPNP] Exclude memory regions for the
3641 Ranges are in pairs (memory base and size).
3643 ports= [IP_VS_FTP] IPVS ftp helper module
3645 Up to 8 (IP_VS_APP_MAX_PORTS) ports
3647 Format: <port>,<port>....
3649 powersave=off [PPC] This option disables power saving features.
3650 It specifically disables cpuidle and sets the
3651 platform machine description specific power_save
3652 function to NULL. On Idle the CPU just reduces
3655 ppc_strict_facility_enable
3656 [PPC] This option catches any kernel floating point,
3657 Altivec, VSX and SPE outside of regions specifically
3658 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()).
3659 There is some performance impact when enabling this.
3663 Disable Hardware Transactional Memory
3665 print-fatal-signals=
3666 [KNL] debug: print fatal signals
3668 If enabled, warn about various signal handling
3669 related application anomalies: too many signals,
3670 too many POSIX.1 timers, fatal signals causing a
3673 If you hit the warning due to signal overflow,
3674 you might want to try "ulimit -i unlimited".
3678 printk.always_kmsg_dump=
3679 Trigger kmsg_dump for cases other than kernel oops or
3681 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3684 printk.devkmsg={on,off,ratelimit}
3685 Control writing to /dev/kmsg.
3686 on - unlimited logging to /dev/kmsg from userspace
3687 off - logging to /dev/kmsg disabled
3688 ratelimit - ratelimit the logging
3691 printk.time= Show timing data prefixed to each printk message line
3692 Format: <bool> (1/Y/y=enable, 0/N/n=disable)
3694 processor.max_cstate= [HW,ACPI]
3695 Limit processor to maximum C-state
3696 max_cstate=9 overrides any DMI blacklist limit.
3698 processor.nocst [HW,ACPI]
3699 Ignore the _CST method to determine C-states,
3700 instead using the legacy FADT method
3702 profile= [KNL] Enable kernel profiling via /proc/profile
3703 Format: [<profiletype>,]<number>
3704 Param: <profiletype>: "schedule", "sleep", or "kvm"
3705 [defaults to kernel profiling]
3706 Param: "schedule" - profile schedule points.
3707 Param: "sleep" - profile D-state sleeping (millisecs).
3708 Requires CONFIG_SCHEDSTATS
3709 Param: "kvm" - profile VM exits.
3710 Param: <number> - step/bucket size as a power of 2 for
3711 statistical time based profiling.
3713 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk
3715 See Documentation/admin-guide/blockdev/ramdisk.rst.
3717 psi= [KNL] Enable or disable pressure stall information
3721 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
3722 probe for; one of (bare|imps|exps|lifebook|any).
3723 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
3725 psmouse.resetafter= [HW,MOUSE]
3726 Try to reset the device after so many bad packets
3729 [HW,MOUSE] Set desired mouse resolution, in dpi.
3730 psmouse.smartscroll=
3731 [HW,MOUSE] Controls Logitech smartscroll autorepeat.
3732 0 = disabled, 1 = enabled (default).
3734 pstore.backend= Specify the name of the pstore backend to use
3737 See Documentation/admin-guide/blockdev/paride.rst.
3739 pti= [X86_64] Control Page Table Isolation of user and
3740 kernel address spaces. Disabling this feature
3741 removes hardening, but improves performance of
3742 system calls and interrupts.
3744 on - unconditionally enable
3745 off - unconditionally disable
3746 auto - kernel detects whether your CPU model is
3747 vulnerable to issues that PTI mitigates
3749 Not specifying this option is equivalent to pti=auto.
3752 Equivalent to pti=off
3755 [KNL] Number of legacy pty's. Overwrites compiled-in
3758 quiet [KNL] Disable most log messages
3763 See Documentation/admin-guide/md.rst.
3765 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
3766 See Documentation/admin-guide/blockdev/ramdisk.rst.
3768 random.trust_cpu={on,off}
3769 [KNL] Enable or disable trusting the use of the
3770 CPU's random number generator (if available) to
3771 fully seed the kernel's CRNG. Default is controlled
3772 by CONFIG_RANDOM_TRUST_CPU.
3774 ras=option[,option,...] [KNL] RAS-specific options
3777 Disable the Correctable Errors Collector,
3778 see CONFIG_RAS_CEC help text.
3781 The argument is a cpu list, as described above,
3782 except that the string "all" can be used to
3783 specify every CPU on the system.
3785 In kernels built with CONFIG_RCU_NOCB_CPU=y, set
3786 the specified list of CPUs to be no-callback CPUs.
3787 Invocation of these CPUs' RCU callbacks will be
3788 offloaded to "rcuox/N" kthreads created for that
3789 purpose, where "x" is "p" for RCU-preempt, and
3790 "s" for RCU-sched, and "N" is the CPU number.
3791 This reduces OS jitter on the offloaded CPUs,
3792 which can be useful for HPC and real-time
3793 workloads. It can also improve energy efficiency
3794 for asymmetric multiprocessors.
3797 Rather than requiring that offloaded CPUs
3798 (specified by rcu_nocbs= above) explicitly
3799 awaken the corresponding "rcuoN" kthreads,
3800 make these kthreads poll for callbacks.
3801 This improves the real-time response for the
3802 offloaded CPUs by relieving them of the need to
3803 wake up the corresponding kthread, but degrades
3804 energy efficiency by requiring that the kthreads
3805 periodically wake up to do the polling.
3807 rcutree.blimit= [KNL]
3808 Set maximum number of finished RCU callbacks to
3809 process in one batch.
3811 rcutree.dump_tree= [KNL]
3812 Dump the structure of the rcu_node combining tree
3813 out at early boot. This is used for diagnostic
3814 purposes, to verify correct tree setup.
3816 rcutree.gp_cleanup_delay= [KNL]
3817 Set the number of jiffies to delay each step of
3818 RCU grace-period cleanup.
3820 rcutree.gp_init_delay= [KNL]
3821 Set the number of jiffies to delay each step of
3822 RCU grace-period initialization.
3824 rcutree.gp_preinit_delay= [KNL]
3825 Set the number of jiffies to delay each step of
3826 RCU grace-period pre-initialization, that is,
3827 the propagation of recent CPU-hotplug changes up
3828 the rcu_node combining tree.
3830 rcutree.use_softirq= [KNL]
3831 If set to zero, move all RCU_SOFTIRQ processing to
3832 per-CPU rcuc kthreads. Defaults to a non-zero
3833 value, meaning that RCU_SOFTIRQ is used by default.
3834 Specify rcutree.use_softirq=0 to use rcuc kthreads.
3836 rcutree.rcu_fanout_exact= [KNL]
3837 Disable autobalancing of the rcu_node combining
3838 tree. This is used by rcutorture, and might
3839 possibly be useful for architectures having high
3840 cache-to-cache transfer latencies.
3842 rcutree.rcu_fanout_leaf= [KNL]
3843 Change the number of CPUs assigned to each
3844 leaf rcu_node structure. Useful for very
3845 large systems, which will choose the value 64,
3846 and for NUMA systems with large remote-access
3847 latencies, which will choose a value aligned
3848 with the appropriate hardware boundaries.
3850 rcutree.jiffies_till_first_fqs= [KNL]
3851 Set delay from grace-period initialization to
3852 first attempt to force quiescent states.
3853 Units are jiffies, minimum value is zero,
3854 and maximum value is HZ.
3856 rcutree.jiffies_till_next_fqs= [KNL]
3857 Set delay between subsequent attempts to force
3858 quiescent states. Units are jiffies, minimum
3859 value is one, and maximum value is HZ.
3861 rcutree.jiffies_till_sched_qs= [KNL]
3862 Set required age in jiffies for a
3863 given grace period before RCU starts
3864 soliciting quiescent-state help from
3865 rcu_note_context_switch() and cond_resched().
3866 If not specified, the kernel will calculate
3867 a value based on the most recent settings
3868 of rcutree.jiffies_till_first_fqs
3869 and rcutree.jiffies_till_next_fqs.
3870 This calculated value may be viewed in
3871 rcutree.jiffies_to_sched_qs. Any attempt to set
3872 rcutree.jiffies_to_sched_qs will be cheerfully
3875 rcutree.kthread_prio= [KNL,BOOT]
3876 Set the SCHED_FIFO priority of the RCU per-CPU
3877 kthreads (rcuc/N). This value is also used for
3878 the priority of the RCU boost threads (rcub/N)
3879 and for the RCU grace-period kthreads (rcu_bh,
3880 rcu_preempt, and rcu_sched). If RCU_BOOST is
3881 set, valid values are 1-99 and the default is 1
3882 (the least-favored priority). Otherwise, when
3883 RCU_BOOST is not set, valid values are 0-99 and
3884 the default is zero (non-realtime operation).
3886 rcutree.rcu_nocb_gp_stride= [KNL]
3887 Set the number of NOCB callback kthreads in
3888 each group, which defaults to the square root
3889 of the number of CPUs. Larger numbers reduce
3890 the wakeup overhead on the global grace-period
3891 kthread, but increases that same overhead on
3892 each group's NOCB grace-period kthread.
3894 rcutree.qhimark= [KNL]
3895 Set threshold of queued RCU callbacks beyond which
3896 batch limiting is disabled.
3898 rcutree.qlowmark= [KNL]
3899 Set threshold of queued RCU callbacks below which
3900 batch limiting is re-enabled.
3902 rcutree.rcu_idle_gp_delay= [KNL]
3903 Set wakeup interval for idle CPUs that have
3904 RCU callbacks (RCU_FAST_NO_HZ=y).
3906 rcutree.rcu_idle_lazy_gp_delay= [KNL]
3907 Set wakeup interval for idle CPUs that have
3908 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y).
3909 Lazy RCU callbacks are those which RCU can
3910 prove do nothing more than free memory.
3912 rcutree.rcu_kick_kthreads= [KNL]
3913 Cause the grace-period kthread to get an extra
3914 wake_up() if it sleeps three times longer than
3915 it should at force-quiescent-state time.
3916 This wake_up() will be accompanied by a
3917 WARN_ONCE() splat and an ftrace_dump().
3919 rcutree.sysrq_rcu= [KNL]
3920 Commandeer a sysrq key to dump out Tree RCU's
3921 rcu_node tree with an eye towards determining
3922 why a new grace period has not yet started.
3924 rcuperf.gp_async= [KNL]
3925 Measure performance of asynchronous
3926 grace-period primitives such as call_rcu().
3928 rcuperf.gp_async_max= [KNL]
3929 Specify the maximum number of outstanding
3930 callbacks per writer thread. When a writer
3931 thread exceeds this limit, it invokes the
3932 corresponding flavor of rcu_barrier() to allow
3933 previously posted callbacks to drain.
3935 rcuperf.gp_exp= [KNL]
3936 Measure performance of expedited synchronous
3937 grace-period primitives.
3939 rcuperf.holdoff= [KNL]
3940 Set test-start holdoff period. The purpose of
3941 this parameter is to delay the start of the
3942 test until boot completes in order to avoid
3945 rcuperf.nreaders= [KNL]
3946 Set number of RCU readers. The value -1 selects
3947 N, where N is the number of CPUs. A value
3948 "n" less than -1 selects N-n+1, where N is again
3949 the number of CPUs. For example, -2 selects N
3950 (the number of CPUs), -3 selects N+1, and so on.
3951 A value of "n" less than or equal to -N selects
3954 rcuperf.nwriters= [KNL]
3955 Set number of RCU writers. The values operate
3956 the same as for rcuperf.nreaders.
3957 N, where N is the number of CPUs
3959 rcuperf.perf_type= [KNL]
3960 Specify the RCU implementation to test.
3962 rcuperf.shutdown= [KNL]
3963 Shut the system down after performance tests
3964 complete. This is useful for hands-off automated
3967 rcuperf.verbose= [KNL]
3968 Enable additional printk() statements.
3970 rcuperf.writer_holdoff= [KNL]
3971 Write-side holdoff between grace periods,
3972 in microseconds. The default of zero says
3975 rcutorture.fqs_duration= [KNL]
3976 Set duration of force_quiescent_state bursts
3979 rcutorture.fqs_holdoff= [KNL]
3980 Set holdoff time within force_quiescent_state bursts
3983 rcutorture.fqs_stutter= [KNL]
3984 Set wait time between force_quiescent_state bursts
3987 rcutorture.fwd_progress= [KNL]
3988 Enable RCU grace-period forward-progress testing
3989 for the types of RCU supporting this notion.
3991 rcutorture.fwd_progress_div= [KNL]
3992 Specify the fraction of a CPU-stall-warning
3993 period to do tight-loop forward-progress testing.
3995 rcutorture.fwd_progress_holdoff= [KNL]
3996 Number of seconds to wait between successive
3997 forward-progress tests.
3999 rcutorture.fwd_progress_need_resched= [KNL]
4000 Enclose cond_resched() calls within checks for
4001 need_resched() during tight-loop forward-progress
4004 rcutorture.gp_cond= [KNL]
4005 Use conditional/asynchronous update-side
4006 primitives, if available.
4008 rcutorture.gp_exp= [KNL]
4009 Use expedited update-side primitives, if available.
4011 rcutorture.gp_normal= [KNL]
4012 Use normal (non-expedited) asynchronous
4013 update-side primitives, if available.
4015 rcutorture.gp_sync= [KNL]
4016 Use normal (non-expedited) synchronous
4017 update-side primitives, if available. If all
4018 of rcutorture.gp_cond=, rcutorture.gp_exp=,
4019 rcutorture.gp_normal=, and rcutorture.gp_sync=
4020 are zero, rcutorture acts as if is interpreted
4021 they are all non-zero.
4023 rcutorture.n_barrier_cbs= [KNL]
4024 Set callbacks/threads for rcu_barrier() testing.
4026 rcutorture.nfakewriters= [KNL]
4027 Set number of concurrent RCU writers. These just
4028 stress RCU, they don't participate in the actual
4029 test, hence the "fake".
4031 rcutorture.nreaders= [KNL]
4032 Set number of RCU readers. The value -1 selects
4033 N-1, where N is the number of CPUs. A value
4034 "n" less than -1 selects N-n-2, where N is again
4035 the number of CPUs. For example, -2 selects N
4036 (the number of CPUs), -3 selects N+1, and so on.
4038 rcutorture.object_debug= [KNL]
4039 Enable debug-object double-call_rcu() testing.
4041 rcutorture.onoff_holdoff= [KNL]
4042 Set time (s) after boot for CPU-hotplug testing.
4044 rcutorture.onoff_interval= [KNL]
4045 Set time (jiffies) between CPU-hotplug operations,
4046 or zero to disable CPU-hotplug testing.
4048 rcutorture.shuffle_interval= [KNL]
4049 Set task-shuffle interval (s). Shuffling tasks
4050 allows some CPUs to go into dyntick-idle mode
4051 during the rcutorture test.
4053 rcutorture.shutdown_secs= [KNL]
4054 Set time (s) after boot system shutdown. This
4055 is useful for hands-off automated testing.
4057 rcutorture.stall_cpu= [KNL]
4058 Duration of CPU stall (s) to test RCU CPU stall
4059 warnings, zero to disable.
4061 rcutorture.stall_cpu_holdoff= [KNL]
4062 Time to wait (s) after boot before inducing stall.
4064 rcutorture.stall_cpu_irqsoff= [KNL]
4065 Disable interrupts while stalling if set.
4067 rcutorture.stat_interval= [KNL]
4068 Time (s) between statistics printk()s.
4070 rcutorture.stutter= [KNL]
4071 Time (s) to stutter testing, for example, specifying
4072 five seconds causes the test to run for five seconds,
4073 wait for five seconds, and so on. This tests RCU's
4074 ability to transition abruptly to and from idle.
4076 rcutorture.test_boost= [KNL]
4077 Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
4078 "Maybe" means test if the RCU implementation
4079 under test support RCU priority boosting.
4081 rcutorture.test_boost_duration= [KNL]
4082 Duration (s) of each individual boost test.
4084 rcutorture.test_boost_interval= [KNL]
4085 Interval (s) between each boost test.
4087 rcutorture.test_no_idle_hz= [KNL]
4088 Test RCU's dyntick-idle handling. See also the
4089 rcutorture.shuffle_interval parameter.
4091 rcutorture.torture_type= [KNL]
4092 Specify the RCU implementation to test.
4094 rcutorture.verbose= [KNL]
4095 Enable additional printk() statements.
4097 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL]
4098 Dump ftrace buffer after reporting RCU CPU
4101 rcupdate.rcu_cpu_stall_suppress= [KNL]
4102 Suppress RCU CPU stall warning messages.
4104 rcupdate.rcu_cpu_stall_timeout= [KNL]
4105 Set timeout for RCU CPU stall warning messages.
4107 rcupdate.rcu_expedited= [KNL]
4108 Use expedited grace-period primitives, for
4109 example, synchronize_rcu_expedited() instead
4110 of synchronize_rcu(). This reduces latency,
4111 but can increase CPU utilization, degrade
4112 real-time latency, and degrade energy efficiency.
4113 No effect on CONFIG_TINY_RCU kernels.
4115 rcupdate.rcu_normal= [KNL]
4116 Use only normal grace-period primitives,
4117 for example, synchronize_rcu() instead of
4118 synchronize_rcu_expedited(). This improves
4119 real-time latency, CPU utilization, and
4120 energy efficiency, but can expose users to
4121 increased grace-period latency. This parameter
4122 overrides rcupdate.rcu_expedited. No effect on
4123 CONFIG_TINY_RCU kernels.
4125 rcupdate.rcu_normal_after_boot= [KNL]
4126 Once boot has completed (that is, after
4127 rcu_end_inkernel_boot() has been invoked), use
4128 only normal grace-period primitives. No effect
4129 on CONFIG_TINY_RCU kernels.
4131 rcupdate.rcu_task_stall_timeout= [KNL]
4132 Set timeout in jiffies for RCU task stall warning
4133 messages. Disable with a value less than or equal
4136 rcupdate.rcu_self_test= [KNL]
4137 Run the RCU early boot self tests
4141 Run specified binary instead of /init from the ramdisk,
4142 used for early userspace startup. See initrd.
4145 force - Override the decision by the kernel to hide the
4146 advertisement of RDRAND support (this affects
4147 certain AMD processors because of buggy BIOS
4148 support, specifically around the suspend/resume
4152 Turn on/off individual RDT features. List is:
4153 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp,
4155 E.g. to turn on cmt and turn off mba use:
4159 Format (x86 or x86_64):
4160 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \
4162 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \
4164 Where reboot_mode is one of warm (soft) or cold (hard) or gpio
4165 (prefix with 'panic_' to set mode for panic
4167 reboot_type is one of bios, acpi, kbd, triple, efi, or pci,
4168 reboot_force is either force or not specified,
4169 reboot_cpu is s[mp]#### with #### being the processor
4170 to be used for rebooting.
4173 [KNL, SMP] Set scheduler's default relax_domain_level.
4174 See Documentation/admin-guide/cgroup-v1/cpusets.rst.
4176 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory
4177 Format: <base1>,<size1>[,<base2>,<size2>,...]
4178 Reserve I/O ports or memory so the kernel won't use
4179 them. If <base> is less than 0x10000, the region
4180 is assumed to be I/O ports; otherwise it is memory.
4182 reservetop= [X86-32]
4184 Reserves a hole at the top of the kernel virtual
4189 Set the amount of memory to reserve for BIOS at
4190 the bottom of the address space.
4192 reset_devices [KNL] Force drivers to reset the underlying device
4193 during initialization.
4196 Specify the partition device for software suspend
4198 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>}
4200 resume_offset= [SWSUSP]
4201 Specify the offset from the beginning of the partition
4202 given by "resume=" at which the swap header is located,
4203 in <PAGE_SIZE> units (needed only for swap files).
4204 See Documentation/power/swsusp-and-swap-files.rst
4206 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to
4207 read the resume files
4209 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up.
4210 Useful for devices that are detected asynchronously
4211 (e.g. USB and MMC devices).
4213 hibernate= [HIBERNATION]
4214 noresume Don't check if there's a hibernation image
4215 present during boot.
4216 nocompress Don't compress/decompress hibernation images.
4217 no Disable hibernation and resume.
4218 protect_image Turn on image protection during restoration
4219 (that will set all pages holding image data
4220 during restoration read-only).
4222 retain_initrd [RAM] Keep initrd memory after extraction
4224 rfkill.default_state=
4225 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm,
4226 etc. communication is blocked by default.
4229 rfkill.master_switch_mode=
4230 0 The "airplane mode" button does nothing.
4231 1 The "airplane mode" button toggles between everything
4232 blocked and the previous configuration.
4233 2 The "airplane mode" button toggles between everything
4234 blocked and everything unblocked.
4236 rhash_entries= [KNL,NET]
4237 Set number of hash buckets for route cache
4240 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported
4243 ro [KNL] Mount root device read-only on boot
4246 on Mark read-only kernel memory as read-only (default).
4247 off Leave read-only kernel memory writable for debugging.
4250 Enable the uart passthrough on the designated usb port
4251 on Rockchip SoCs. When active, the signals of the
4252 debug-uart get routed to the D+ and D- pins of the usb
4253 port and the regular usb controller gets disabled.
4255 root= [KNL] Root filesystem
4256 See name_to_dev_t comment in init/do_mounts.c.
4258 rootdelay= [KNL] Delay (in seconds) to pause before attempting to
4259 mount the root filesystem
4261 rootflags= [KNL] Set root filesystem mount option string
4263 rootfstype= [KNL] Set root filesystem type
4265 rootwait [KNL] Wait (indefinitely) for root device to show up.
4266 Useful for devices that are detected asynchronously
4267 (e.g. USB and MMC devices).
4269 rproc_mem=nn[KMG][@address]
4270 [KNL,ARM,CMA] Remoteproc physical memory block.
4271 Memory area to be used by remote processor image,
4274 rw [KNL] Mount root device read-write on boot
4276 S [KNL] Run init in single mode
4278 s390_iommu= [HW,S390]
4279 Set s390 IOTLB flushing mode
4281 With strict flushing every unmap operation will result in
4282 an IOTLB flush. Default is lazy flushing before reuse,
4286 See drivers/net/irda/sa1100_ir.c.
4288 sbni= [NET] Granch SBNI12 leased line adapter
4290 sched_debug [KNL] Enables verbose scheduler debug messages.
4292 schedstats= [KNL,X86] Enable or disable scheduled statistics.
4293 Allowed values are enable and disable. This feature
4294 incurs a small amount of overhead in the scheduler
4295 but is useful for debugging and performance tuning.
4297 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate
4298 xtime_lock contention on larger systems, and/or RCU lock
4299 contention on all systems with CONFIG_MAXSMP set.
4300 Format: { "0" | "1" }
4301 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1"
4303 Note: increases power consumption, thus should only be
4304 enabled if running jitter sensitive (HPC/RT) workloads.
4306 security= [SECURITY] Choose a legacy "major" security module to
4307 enable at boot. This has been deprecated by the
4310 selinux= [SELINUX] Disable or enable SELinux at boot time.
4311 Format: { "0" | "1" }
4312 See security/selinux/Kconfig help text.
4315 Default value is set via kernel config option.
4316 If enabled at boot time, /selinux/disable can be used
4317 later to disable prior to initial policy load.
4319 apparmor= [APPARMOR] Disable or enable AppArmor at boot time
4320 Format: { "0" | "1" }
4321 See security/apparmor/Kconfig help text
4324 Default value is set via kernel config option.
4326 serialnumber [BUGS=X86-32]
4329 Maximal number of shapers.
4337 Disable merging of slabs with similar size. May be
4338 necessary if there is some reason to distinguish
4339 allocs to different slabs, especially in hardened
4340 environments where the risk of heap overflows and
4341 layout control by attackers can usually be
4342 frustrated by disabling merging. This will reduce
4343 most of the exposure of a heap attack to a single
4344 cache (risks via metadata attacks are mostly
4345 unchanged). Debug options disable merging on their
4347 For more information see Documentation/vm/slub.rst.
4349 slab_max_order= [MM, SLAB]
4350 Determines the maximum allowed order for slabs.
4351 A high setting may cause OOMs due to memory
4352 fragmentation. Defaults to 1 for systems with
4353 more than 32MB of RAM, 0 otherwise.
4355 slub_debug[=options[,slabs]] [MM, SLUB]
4356 Enabling slub_debug allows one to determine the
4357 culprit if slab objects become corrupted. Enabling
4358 slub_debug can create guard zones around objects and
4359 may poison objects when not in use. Also tracks the
4360 last alloc / free. For more information see
4361 Documentation/vm/slub.rst.
4363 slub_memcg_sysfs= [MM, SLUB]
4364 Determines whether to enable sysfs directories for
4365 memory cgroup sub-caches. 1 to enable, 0 to disable.
4366 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON.
4367 Enabling this can lead to a very high number of debug
4368 directories and files being created under
4371 slub_max_order= [MM, SLUB]
4372 Determines the maximum allowed order for slabs.
4373 A high setting may cause OOMs due to memory
4374 fragmentation. For more information see
4375 Documentation/vm/slub.rst.
4377 slub_min_objects= [MM, SLUB]
4378 The minimum number of objects per slab. SLUB will
4379 increase the slab order up to slub_max_order to
4380 generate a sufficiently large slab able to contain
4381 the number of objects indicated. The higher the number
4382 of objects the smaller the overhead of tracking slabs
4383 and the less frequently locks need to be acquired.
4384 For more information see Documentation/vm/slub.rst.
4386 slub_min_order= [MM, SLUB]
4387 Determines the minimum page order for slabs. Must be
4388 lower than slub_max_order.
4389 For more information see Documentation/vm/slub.rst.
4391 slub_nomerge [MM, SLUB]
4392 Same with slab_nomerge. This is supported for legacy.
4393 See slab_nomerge for more information.
4396 Format: <io1>[,<io2>[,...,<io8>]]
4398 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices
4399 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port
4400 smsc-ircc2.ircc_sir= [HW] SIR base I/O port
4401 smsc-ircc2.ircc_fir= [HW] FIR base I/O port
4402 smsc-ircc2.ircc_irq= [HW] IRQ line
4403 smsc-ircc2.ircc_dma= [HW] DMA channel
4404 smsc-ircc2.ircc_transceiver= [HW] Transceiver type:
4405 0: Toshiba Satellite 1800 (GP data pin select)
4406 1: Fast pin select (default)
4409 smt [KNL,S390] Set the maximum number of threads (logical
4410 CPUs) to use per physical CPU on systems capable of
4411 symmetric multithreading (SMT). Will be capped to the
4412 actual hardware limit.
4414 Default: -1 (no limit)
4417 [KNL] Should the soft-lockup detector generate panics.
4420 A nonzero value instructs the soft-lockup detector
4421 to panic the machine when a soft-lockup occurs. This
4422 is also controlled by CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC
4423 which is the respective build-time switch to that
4426 softlockup_all_cpu_backtrace=
4427 [KNL] Should the soft-lockup detector generate
4428 backtraces on all cpus.
4431 sonypi.*= [HW] Sony Programmable I/O Control Device driver
4432 See Documentation/admin-guide/laptops/sonypi.rst
4434 spectre_v2= [X86] Control mitigation of Spectre variant 2
4435 (indirect branch speculation) vulnerability.
4436 The default operation protects the kernel from
4439 on - unconditionally enable, implies
4441 off - unconditionally disable, implies
4443 auto - kernel detects whether your CPU model is
4446 Selecting 'on' will, and 'auto' may, choose a
4447 mitigation method at run time according to the
4448 CPU, the available microcode, the setting of the
4449 CONFIG_RETPOLINE configuration option, and the
4450 compiler with which the kernel was built.
4452 Selecting 'on' will also enable the mitigation
4453 against user space to user space task attacks.
4455 Selecting 'off' will disable both the kernel and
4456 the user space protections.
4458 Specific mitigations can also be selected manually:
4460 retpoline - replace indirect branches
4461 retpoline,generic - google's original retpoline
4462 retpoline,amd - AMD-specific minimal thunk
4464 Not specifying this option is equivalent to
4468 [X86] Control mitigation of Spectre variant 2
4469 (indirect branch speculation) vulnerability between
4472 on - Unconditionally enable mitigations. Is
4473 enforced by spectre_v2=on
4475 off - Unconditionally disable mitigations. Is
4476 enforced by spectre_v2=off
4478 prctl - Indirect branch speculation is enabled,
4479 but mitigation can be enabled via prctl
4480 per thread. The mitigation control state
4481 is inherited on fork.
4484 - Like "prctl" above, but only STIBP is
4485 controlled per thread. IBPB is issued
4486 always when switching between different user
4490 - Same as "prctl" above, but all seccomp
4491 threads will enable the mitigation unless
4492 they explicitly opt out.
4495 - Like "seccomp" above, but only STIBP is
4496 controlled per thread. IBPB is issued
4497 always when switching between different
4498 user space processes.
4500 auto - Kernel selects the mitigation depending on
4501 the available CPU features and vulnerability.
4504 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl"
4506 Not specifying this option is equivalent to
4507 spectre_v2_user=auto.
4509 spec_store_bypass_disable=
4510 [HW] Control Speculative Store Bypass (SSB) Disable mitigation
4511 (Speculative Store Bypass vulnerability)
4513 Certain CPUs are vulnerable to an exploit against a
4514 a common industry wide performance optimization known
4515 as "Speculative Store Bypass" in which recent stores
4516 to the same memory location may not be observed by
4517 later loads during speculative execution. The idea
4518 is that such stores are unlikely and that they can
4519 be detected prior to instruction retirement at the
4520 end of a particular speculation execution window.
4522 In vulnerable processors, the speculatively forwarded
4523 store can be used in a cache side channel attack, for
4524 example to read memory to which the attacker does not
4525 directly have access (e.g. inside sandboxed code).
4527 This parameter controls whether the Speculative Store
4528 Bypass optimization is used.
4530 On x86 the options are:
4532 on - Unconditionally disable Speculative Store Bypass
4533 off - Unconditionally enable Speculative Store Bypass
4534 auto - Kernel detects whether the CPU model contains an
4535 implementation of Speculative Store Bypass and
4536 picks the most appropriate mitigation. If the
4537 CPU is not vulnerable, "off" is selected. If the
4538 CPU is vulnerable the default mitigation is
4539 architecture and Kconfig dependent. See below.
4540 prctl - Control Speculative Store Bypass per thread
4541 via prctl. Speculative Store Bypass is enabled
4542 for a process by default. The state of the control
4543 is inherited on fork.
4544 seccomp - Same as "prctl" above, but all seccomp threads
4545 will disable SSB unless they explicitly opt out.
4547 Default mitigations:
4548 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl"
4550 On powerpc the options are:
4552 on,auto - On Power8 and Power9 insert a store-forwarding
4553 barrier on kernel entry and exit. On Power7
4554 perform a software flush on kernel entry and
4558 Not specifying this option is equivalent to
4559 spec_store_bypass_disable=auto.
4561 spia_io_base= [HW,MTD]
4566 srcutree.counter_wrap_check [KNL]
4567 Specifies how frequently to check for
4568 grace-period sequence counter wrap for the
4569 srcu_data structure's ->srcu_gp_seq_needed field.
4570 The greater the number of bits set in this kernel
4571 parameter, the less frequently counter wrap will
4572 be checked for. Note that the bottom two bits
4575 srcutree.exp_holdoff [KNL]
4576 Specifies how many nanoseconds must elapse
4577 since the end of the last SRCU grace period for
4578 a given srcu_struct until the next normal SRCU
4579 grace period will be considered for automatic
4580 expediting. Set to zero to disable automatic
4584 Speculative Store Bypass Disable control
4586 On CPUs that are vulnerable to the Speculative
4587 Store Bypass vulnerability and offer a
4588 firmware based mitigation, this parameter
4589 indicates how the mitigation should be used:
4591 force-on: Unconditionally enable mitigation for
4592 for both kernel and userspace
4593 force-off: Unconditionally disable mitigation for
4594 for both kernel and userspace
4595 kernel: Always enable mitigation in the
4596 kernel, and offer a prctl interface
4597 to allow userspace to register its
4598 interest in being mitigated too.
4600 stack_guard_gap= [MM]
4601 override the default stack gap protection. The value
4602 is in page units and it defines how many pages prior
4603 to (for stacks growing down) resp. after (for stacks
4604 growing up) the main stack are reserved for no other
4605 mapping. Default value is 256 pages.
4608 Enabled the stack tracer on boot up.
4610 stacktrace_filter=[function-list]
4611 [FTRACE] Limit the functions that the stack tracer
4612 will trace at boot up. function-list is a comma separated
4613 list of functions. This list can be changed at run
4614 time by the stack_trace_filter file in the debugfs
4615 tracing directory. Note, this enables stack tracing
4616 and the stacktrace above is not needed.
4620 Set the STI (builtin display/keyboard on the HP-PARISC
4621 machines) console (graphic card) which should be used
4622 as the initial boot-console.
4623 See also comment in drivers/video/console/sticore.c.
4626 See comment in drivers/video/console/sticore.c.
4629 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]]
4631 sunrpc.min_resvport=
4632 sunrpc.max_resvport=
4634 SunRPC servers often require that client requests
4635 originate from a privileged port (i.e. a port in the
4636 range 0 < portnr < 1024).
4637 An administrator who wishes to reserve some of these
4638 ports for other uses may adjust the range that the
4639 kernel's sunrpc client considers to be privileged
4640 using these two parameters to set the minimum and
4641 maximum port values.
4643 sunrpc.svc_rpc_per_connection_limit=
4645 Limit the number of requests that the server will
4646 process in parallel from a single connection.
4647 The default value is 0 (no limit).
4651 Control how the NFS server code allocates CPUs to
4652 service thread pools. Depending on how many NICs
4653 you have and where their interrupts are bound, this
4654 option will affect which CPUs will do NFS serving.
4655 Note: this parameter cannot be changed while the
4656 NFS server is running.
4658 auto the server chooses an appropriate mode
4659 automatically using heuristics
4660 global a single global pool contains all CPUs
4661 percpu one pool for each CPU
4662 pernode one pool for each NUMA node (equivalent
4663 to global on non-NUMA machines)
4665 sunrpc.tcp_slot_table_entries=
4666 sunrpc.udp_slot_table_entries=
4668 Sets the upper limit on the number of simultaneous
4669 RPC calls that can be sent from the client to a
4670 server. Increasing these values may allow you to
4671 improve throughput, but will also increase the
4672 amount of memory reserved for use by the client.
4674 suspend.pm_test_delay=
4676 Sets the number of seconds to remain in a suspend test
4677 mode before resuming the system (see
4678 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG
4679 is set. Default value is 5.
4682 Format: { on | off | y | n | 1 | 0 }
4683 This parameter controls use of the Protected
4684 Execution Facility on pSeries.
4687 [KNL] Enable accounting of swap in memory resource
4688 controller if no parameter or 1 is given or disable
4689 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst)
4691 swiotlb= [ARM,IA-64,PPC,MIPS,X86]
4692 Format: { <int> | force | noforce }
4693 <int> -- Number of I/O TLB slabs
4694 force -- force using of bounce buffers even if they
4695 wouldn't be automatically used by the kernel
4696 noforce -- Never use bounce buffers (for debugging)
4700 sysfs.deprecated=0|1 [KNL]
4701 Enable/disable old style sysfs layout for old udev
4702 on older distributions. When this option is enabled
4703 very new udev will not work anymore. When this option
4704 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled)
4705 in older udev will not work anymore.
4706 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in
4707 the kernel configuration.
4709 sysrq_always_enabled
4711 Ignore sysrq setting - this boot parameter will
4712 neutralize any effect of /proc/sys/kernel/sysrq.
4713 Useful for debugging.
4715 tcpmhash_entries= [KNL,NET]
4716 Set the number of tcp_metrics_hash slots.
4717 Default value is 8192 or 16384 depending on total
4718 ram pages. This is used to specify the TCP metrics
4719 cache size. See Documentation/networking/ip-sysctl.txt
4720 "tcp_no_metrics_save" section for more details.
4724 test_suspend= [SUSPEND][,N]
4725 Specify "mem" (for Suspend-to-RAM) or "standby" (for
4726 standby suspend) or "freeze" (for suspend type freeze)
4727 as the system sleep state during system startup with
4728 the optional capability to repeat N number of times.
4729 The system is woken from this state using a
4730 wakeup-capable RTC alarm.
4732 thash_entries= [KNL,NET]
4733 Set number of hash buckets for TCP connection
4735 thermal.act= [HW,ACPI]
4736 -1: disable all active trip points in all thermal zones
4737 <degrees C>: override all lowest active trip points
4739 thermal.crt= [HW,ACPI]
4740 -1: disable all critical trip points in all thermal zones
4741 <degrees C>: override all critical trip points
4743 thermal.nocrt= [HW,ACPI]
4744 Set to disable actions on ACPI thermal zone
4745 critical and hot trip points.
4747 thermal.off= [HW,ACPI]
4748 1: disable ACPI thermal control
4750 thermal.psv= [HW,ACPI]
4751 -1: disable all passive trip points
4752 <degrees C>: override all passive trip points to this
4755 thermal.tzp= [HW,ACPI]
4756 Specify global default ACPI thermal zone polling rate
4757 <deci-seconds>: poll all this frequency
4758 0: no polling (default)
4761 Force threading of all interrupt handlers except those
4762 marked explicitly IRQF_NO_THREAD.
4766 Specify if the kernel should make use of the cpu
4767 topology information if the hardware supports this.
4768 The scheduler will make use of this information and
4769 e.g. base its process migration decisions on it.
4772 topology_updates= [KNL, PPC, NUMA]
4774 Specify if the kernel should ignore (off)
4775 topology updates sent by the hypervisor to this
4780 tpm_suspend_pcr=[HW,TPM]
4781 Format: integer pcr id
4782 Specify that at suspend time, the tpm driver
4783 should extend the specified pcr with zeros,
4784 as a workaround for some chips which fail to
4785 flush the last written pcr on TPM_SaveState.
4786 This will guarantee that all the other pcrs
4789 trace_buf_size=nn[KMG]
4790 [FTRACE] will set tracing buffer size on each cpu.
4792 trace_event=[event-list]
4793 [FTRACE] Set and start specified trace events in order
4794 to facilitate early boot debugging. The event-list is a
4795 comma separated list of trace events to enable. See
4796 also Documentation/trace/events.rst
4798 trace_options=[option-list]
4799 [FTRACE] Enable or disable tracer options at boot.
4800 The option-list is a comma delimited list of options
4801 that can be enabled or disabled just as if you were
4802 to echo the option name into
4804 /sys/kernel/debug/tracing/trace_options
4806 For example, to enable stacktrace option (to dump the
4807 stack trace of each event), add to the command line:
4809 trace_options=stacktrace
4811 See also Documentation/trace/ftrace.rst "trace options"
4815 Have the tracepoints sent to printk as well as the
4816 tracing ring buffer. This is useful for early boot up
4817 where the system hangs or reboots and does not give the
4818 option for reading the tracing buffer or performing a
4819 ftrace_dump_on_oops.
4821 To turn off having tracepoints sent to printk,
4822 echo 0 > /proc/sys/kernel/tracepoint_printk
4823 Note, echoing 1 into this file without the
4824 tracepoint_printk kernel cmdline option has no effect.
4828 Having tracepoints sent to printk() and activating high
4829 frequency tracepoints such as irq or sched, can cause
4830 the system to live lock.
4833 [FTRACE] enable this option to disable tracing when a
4834 warning is hit. This turns off "tracing_on". Tracing can
4835 be enabled again by echoing '1' into the "tracing_on"
4836 file located in /sys/kernel/debug/tracing/
4838 This option is useful, as it disables the trace before
4839 the WARNING dump is called, which prevents the trace to
4840 be filled with content caused by the warning output.
4842 This option can also be set at run time via the sysctl
4843 option: kernel/traceoff_on_warning
4845 transparent_hugepage=
4847 Format: [always|madvise|never]
4848 Can be used to control the default behavior of the system
4849 with respect to transparent hugepages.
4850 See Documentation/admin-guide/mm/transhuge.rst
4853 tsc= Disable clocksource stability checks for TSC.
4855 [x86] reliable: mark tsc clocksource as reliable, this
4856 disables clocksource verification at runtime, as well
4857 as the stability checks done at bootup. Used to enable
4858 high-resolution timer mode on older hardware, and in
4859 virtualized environment.
4860 [x86] noirqtime: Do not use TSC to do irq accounting.
4861 Used to run time disable IRQ_TIME_ACCOUNTING on any
4862 platforms where RDTSC is slow and this accounting
4864 [x86] unstable: mark the TSC clocksource as unstable, this
4865 marks the TSC unconditionally unstable at bootup and
4866 avoids any further wobbles once the TSC watchdog notices.
4867 [x86] nowatchdog: disable clocksource watchdog. Used
4868 in situations with strict latency requirements (where
4869 interruptions from clocksource watchdog are not
4872 tsx= [X86] Control Transactional Synchronization
4873 Extensions (TSX) feature in Intel processors that
4874 support TSX control.
4876 This parameter controls the TSX feature. The options are:
4878 on - Enable TSX on the system. Although there are
4879 mitigations for all known security vulnerabilities,
4880 TSX has been known to be an accelerator for
4881 several previous speculation-related CVEs, and
4882 so there may be unknown security risks associated
4883 with leaving it enabled.
4885 off - Disable TSX on the system. (Note that this
4886 option takes effect only on newer CPUs which are
4887 not vulnerable to MDS, i.e., have
4888 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get
4889 the new IA32_TSX_CTRL MSR through a microcode
4890 update. This new MSR allows for the reliable
4891 deactivation of the TSX functionality.)
4893 auto - Disable TSX if X86_BUG_TAA is present,
4894 otherwise enable TSX on the system.
4896 Not specifying this option is equivalent to tsx=off.
4898 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
4901 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async
4902 Abort (TAA) vulnerability.
4904 Similar to Micro-architectural Data Sampling (MDS)
4905 certain CPUs that support Transactional
4906 Synchronization Extensions (TSX) are vulnerable to an
4907 exploit against CPU internal buffers which can forward
4908 information to a disclosure gadget under certain
4911 In vulnerable processors, the speculatively forwarded
4912 data can be used in a cache side channel attack, to
4913 access data to which the attacker does not have direct
4916 This parameter controls the TAA mitigation. The
4919 full - Enable TAA mitigation on vulnerable CPUs
4922 full,nosmt - Enable TAA mitigation and disable SMT on
4923 vulnerable CPUs. If TSX is disabled, SMT
4924 is not disabled because CPU is not
4925 vulnerable to cross-thread TAA attacks.
4926 off - Unconditionally disable TAA mitigation
4928 Not specifying this option is equivalent to
4929 tsx_async_abort=full. On CPUs which are MDS affected
4930 and deploy MDS mitigation, TAA mitigation is not
4931 required and doesn't provide any additional
4935 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst
4937 turbografx.map[2|3]= [HW,JOY]
4938 TurboGraFX parallel port interface
4940 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7>
4941 See also Documentation/input/devices/joystick-parport.rst
4943 udbg-immortal [PPC] When debugging early kernel crashes that
4944 happen after console_init() and before a proper
4945 console driver takes over, this boot options might
4946 help "seeing" what's going on.
4948 uhash_entries= [KNL,NET]
4949 Set number of hash buckets for UDP/UDP-Lite connections
4952 [USB] Ignore overcurrent events (default N).
4953 Some badly-designed motherboards generate lots of
4954 bogus events, for ports that aren't wired to
4955 anything. Set this parameter to avoid log spamming.
4956 Note that genuine overcurrent events won't be
4960 [X86] Cause panic on unknown NMI.
4962 usbcore.authorized_default=
4963 [USB] Default USB device authorization:
4964 (default -1 = authorized except for wireless USB,
4965 0 = not authorized, 1 = authorized, 2 = authorized
4966 if device connected to internal port)
4968 usbcore.autosuspend=
4969 [USB] The autosuspend time delay (in seconds) used
4970 for newly-detected USB devices (default 2). This
4971 is the time required before an idle device will be
4972 autosuspended. Devices for which the delay is set
4973 to a negative value won't be autosuspended at all.
4975 usbcore.usbfs_snoop=
4976 [USB] Set to log all usbfs traffic (default 0 = off).
4978 usbcore.usbfs_snoop_max=
4979 [USB] Maximum number of bytes to snoop in each URB
4982 usbcore.blinkenlights=
4983 [USB] Set to cycle leds on hubs (default 0 = off).
4985 usbcore.old_scheme_first=
4986 [USB] Start with the old device initialization
4987 scheme, applies only to low and full-speed devices
4990 usbcore.usbfs_memory_mb=
4991 [USB] Memory limit (in MB) for buffers allocated by
4992 usbfs (default = 16, 0 = max = 2047).
4994 usbcore.use_both_schemes=
4995 [USB] Try the other device initialization scheme
4996 if the first one fails (default 1 = enabled).
4998 usbcore.initial_descriptor_timeout=
4999 [USB] Specifies timeout for the initial 64-byte
5000 USB_REQ_GET_DESCRIPTOR request in milliseconds
5001 (default 5000 = 5.0 seconds).
5003 usbcore.nousb [USB] Disable the USB subsystem
5006 [USB] A list of quirk entries to augment the built-in
5007 usb core quirk list. List entries are separated by
5008 commas. Each entry has the form
5009 VendorID:ProductID:Flags. The IDs are 4-digit hex
5010 numbers and Flags is a set of letters. Each letter
5011 will change the built-in quirk; setting it if it is
5012 clear and clearing it if it is set. The letters have
5013 the following meanings:
5014 a = USB_QUIRK_STRING_FETCH_255 (string
5015 descriptors must not be fetched using
5017 b = USB_QUIRK_RESET_RESUME (device can't resume
5018 correctly so reset it instead);
5019 c = USB_QUIRK_NO_SET_INTF (device can't handle
5020 Set-Interface requests);
5021 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't
5022 handle its Configuration or Interface
5024 e = USB_QUIRK_RESET (device can't be reset
5025 (e.g morph devices), don't use reset);
5026 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has
5027 more interface descriptions than the
5028 bNumInterfaces count, and can't handle
5029 talking to these interfaces);
5030 g = USB_QUIRK_DELAY_INIT (device needs a pause
5031 during initialization, after we read
5032 the device descriptor);
5033 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For
5034 high speed and super speed interrupt
5035 endpoints, the USB 2.0 and USB 3.0 spec
5036 require the interval in microframes (1
5037 microframe = 125 microseconds) to be
5038 calculated as interval = 2 ^
5040 Devices with this quirk report their
5041 bInterval as the result of this
5042 calculation instead of the exponent
5043 variable used in the calculation);
5044 i = USB_QUIRK_DEVICE_QUALIFIER (device can't
5045 handle device_qualifier descriptor
5047 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device
5048 generates spurious wakeup, ignore
5049 remote wakeup capability);
5050 k = USB_QUIRK_NO_LPM (device can't handle Link
5052 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL
5053 (Device reports its bInterval as linear
5054 frames instead of the USB 2.0
5056 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs
5057 to be disconnected before suspend to
5058 prevent spurious wakeup);
5059 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a
5060 pause after every control message);
5061 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra
5062 delay after resetting its port);
5063 Example: quirks=0781:5580:bk,0a5c:5834:gij
5066 [USBHID] The interval which mice are to be polled at.
5069 [USBHID] The interval which joysticks are to be polled at.
5072 [USBHID] The interval which keyboards are to be polled at.
5074 usb-storage.delay_use=
5075 [UMS] The delay in seconds before a new device is
5076 scanned for Logical Units (default 1).
5079 [UMS] A list of quirks entries to supplement or
5080 override the built-in unusual_devs list. List
5081 entries are separated by commas. Each entry has
5082 the form VID:PID:Flags where VID and PID are Vendor
5083 and Product ID values (4-digit hex numbers) and
5084 Flags is a set of characters, each corresponding
5085 to a common usb-storage quirk flag as follows:
5086 a = SANE_SENSE (collect more than 18 bytes
5088 b = BAD_SENSE (don't collect more than 18
5089 bytes of sense data);
5090 c = FIX_CAPACITY (decrease the reported
5091 device capacity by one sector);
5092 d = NO_READ_DISC_INFO (don't use
5093 READ_DISC_INFO command);
5094 e = NO_READ_CAPACITY_16 (don't use
5095 READ_CAPACITY_16 command);
5096 f = NO_REPORT_OPCODES (don't use report opcodes
5098 g = MAX_SECTORS_240 (don't transfer more than
5099 240 sectors at a time, uas only);
5100 h = CAPACITY_HEURISTICS (decrease the
5101 reported device capacity by one
5102 sector if the number is odd);
5103 i = IGNORE_DEVICE (don't bind to this
5105 j = NO_REPORT_LUNS (don't use report luns
5107 l = NOT_LOCKABLE (don't try to lock and
5108 unlock ejectable media);
5109 m = MAX_SECTORS_64 (don't transfer more
5110 than 64 sectors = 32 KB at a time);
5111 n = INITIAL_READ10 (force a retry of the
5112 initial READ(10) command);
5113 o = CAPACITY_OK (accept the capacity
5114 reported by the device);
5115 p = WRITE_CACHE (the device cache is ON
5117 r = IGNORE_RESIDUE (the device reports
5118 bogus residue values);
5119 s = SINGLE_LUN (the device has only one
5121 t = NO_ATA_1X (don't allow ATA(12) and ATA(16)
5122 commands, uas only);
5123 u = IGNORE_UAS (don't bind to the uas driver);
5124 w = NO_WP_DETECT (don't test whether the
5125 medium is write-protected).
5126 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE
5127 even if the device claims no cache)
5128 Example: quirks=0419:aaf5:rl,0421:0433:rc
5130 user_debug= [KNL,ARM]
5132 See arch/arm/Kconfig.debug help text.
5133 1 - undefined instruction events
5135 4 - invalid data aborts
5138 Example: user_debug=31
5141 [X86] Flags controlling user PTE allocations.
5143 nohigh = do not allocate PTE pages in
5144 HIGHMEM regardless of setting
5148 On X86_32, this is an alias for vdso32=. Otherwise:
5150 vdso=1: enable VDSO (the default)
5151 vdso=0: disable VDSO mapping
5153 vdso32= [X86] Control the 32-bit vDSO
5154 vdso32=1: enable 32-bit VDSO
5155 vdso32=0 or vdso32=2: disable 32-bit VDSO
5157 See the help text for CONFIG_COMPAT_VDSO for more
5158 details. If CONFIG_COMPAT_VDSO is set, the default is
5159 vdso32=0; otherwise, the default is vdso32=1.
5161 For compatibility with older kernels, vdso32=2 is an
5164 Try vdso32=0 if you encounter an error that says:
5165 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed!
5168 vector=percpu: enable percpu vector domain
5170 video= [FB] Frame buffer configuration
5171 See Documentation/fb/modedb.rst.
5173 video.brightness_switch_enabled= [0,1]
5174 If set to 1, on receiving an ACPI notify event
5175 generated by hotkey, video driver will adjust brightness
5176 level and then send out the event to user space through
5177 the allocated input device; If set to 0, video driver
5178 will only send out the event without touching backlight
5183 [VMMIO] Memory mapped virtio (platform) device.
5185 <size>@<baseaddr>:<irq>[:<id>]
5187 <size> := size (can use standard suffixes
5189 <baseaddr> := physical base address
5190 <irq> := interrupt number (as passed to
5192 <id> := (optional) platform device id
5194 virtio_mmio.device=1K@0x100b0000:48:7
5196 Can be used multiple times for multiple devices.
5198 vga= [BOOT,X86-32] Select a particular video mode
5199 See Documentation/x86/boot.rst and
5200 Documentation/admin-guide/svga.rst.
5201 Use vga=ask for menu.
5202 This is actually a boot loader parameter; the value is
5203 passed to the kernel using a special protocol.
5205 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y.
5206 May slow down system boot speed, especially when
5207 enabled on systems with a large amount of memory.
5208 All options are enabled by default, and this
5209 interface is meant to allow for selectively
5210 enabling or disabling specific virtual memory
5213 Available options are:
5214 P Enable page structure init time poisoning
5215 - Disable all of the above options
5217 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact
5218 size of <nn>. This can be used to increase the
5219 minimum size (128MB on x86). It can also be used to
5220 decrease the size and leave more room for directly
5223 vmcp_cma=nn[MG] [KNL,S390]
5224 Sets the memory size reserved for contiguous memory
5225 allocations for the vmcp device driver.
5227 vmhalt= [KNL,S390] Perform z/VM CP command after system halt.
5230 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic.
5233 vmpoff= [KNL,S390] Perform z/VM CP command after power off.
5237 Controls the behavior of vsyscalls (i.e. calls to
5238 fixed addresses of 0xffffffffff600x00 from legacy
5239 code). Most statically-linked binaries and older
5240 versions of glibc use these calls. Because these
5241 functions are at fixed addresses, they make nice
5242 targets for exploits that can control RIP.
5244 emulate [default] Vsyscalls turn into traps and are
5245 emulated reasonably safely. The vsyscall
5248 xonly Vsyscalls turn into traps and are
5249 emulated reasonably safely. The vsyscall
5250 page is not readable.
5252 none Vsyscalls don't work at all. This makes
5253 them quite hard to use for exploits but
5254 might break your system.
5256 vt.color= [VT] Default text color.
5257 Format: 0xYX, X = foreground, Y = background.
5258 Default: 0x07 = light gray on black.
5260 vt.cur_default= [VT] Default cursor shape.
5261 Format: 0xCCBBAA, where AA, BB, and CC are the same as
5262 the parameters of the <Esc>[?A;B;Cc escape sequence;
5263 see VGA-softcursor.txt. Default: 2 = underline.
5265 vt.default_blu= [VT]
5266 Format: <blue0>,<blue1>,<blue2>,...,<blue15>
5267 Change the default blue palette of the console.
5268 This is a 16-member array composed of values
5271 vt.default_grn= [VT]
5272 Format: <green0>,<green1>,<green2>,...,<green15>
5273 Change the default green palette of the console.
5274 This is a 16-member array composed of values
5277 vt.default_red= [VT]
5278 Format: <red0>,<red1>,<red2>,...,<red15>
5279 Change the default red palette of the console.
5280 This is a 16-member array composed of values
5286 Set system-wide default UTF-8 mode for all tty's.
5287 Default is 1, i.e. UTF-8 mode is enabled for all
5288 newly opened terminals.
5290 vt.global_cursor_default=
5293 Set system-wide default for whether a cursor
5294 is shown on new VTs. Default is -1,
5295 i.e. cursors will be created by default unless
5296 overridden by individual drivers. 0 will hide
5297 cursors, 1 will display them.
5299 vt.italic= [VT] Default color for italic text; 0-15.
5302 vt.underline= [VT] Default color for underlined text; 0-15.
5305 watchdog timers [HW,WDT] For information on watchdog timers,
5306 see Documentation/watchdog/watchdog-parameters.rst
5307 or other driver-specific files in the
5308 Documentation/watchdog/ directory.
5312 Set the hard lockup detector stall duration
5313 threshold in seconds. The soft lockup detector
5314 threshold is set to twice the value. A value of 0
5315 disables both lockup detectors. Default is 10
5318 workqueue.watchdog_thresh=
5319 If CONFIG_WQ_WATCHDOG is configured, workqueue can
5320 warn stall conditions and dump internal state to
5321 help debugging. 0 disables workqueue stall
5322 detection; otherwise, it's the stall threshold
5323 duration in seconds. The default value is 30 and
5324 it can be updated at runtime by writing to the
5325 corresponding sysfs file.
5327 workqueue.disable_numa
5328 By default, all work items queued to unbound
5329 workqueues are affine to the NUMA nodes they're
5330 issued on, which results in better behavior in
5331 general. If NUMA affinity needs to be disabled for
5332 whatever reason, this option can be used. Note
5333 that this also can be controlled per-workqueue for
5334 workqueues visible under /sys/bus/workqueue/.
5336 workqueue.power_efficient
5337 Per-cpu workqueues are generally preferred because
5338 they show better performance thanks to cache
5339 locality; unfortunately, per-cpu workqueues tend to
5340 be more power hungry than unbound workqueues.
5342 Enabling this makes the per-cpu workqueues which
5343 were observed to contribute significantly to power
5344 consumption unbound, leading to measurably lower
5345 power usage at the cost of small performance
5348 The default value of this parameter is determined by
5349 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT.
5351 workqueue.debug_force_rr_cpu
5352 Workqueue used to implicitly guarantee that work
5353 items queued without explicit CPU specified are put
5354 on the local CPU. This guarantee is no longer true
5355 and while local CPU is still preferred work items
5356 may be put on foreign CPUs. This debug option
5357 forces round-robin CPU selection to flush out
5358 usages which depend on the now broken guarantee.
5359 When enabled, memory and cache locality will be
5362 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of
5363 default x2apic cluster mode on platforms
5366 x86_intel_mid_timer= [X86-32,APBT]
5367 Choose timer option for x86 Intel MID platform.
5368 Two valid options are apbt timer only and lapic timer
5369 plus one apbt timer for broadcast timer.
5370 x86_intel_mid_timer=apbt_only | lapic_and_apbt
5372 xen_512gb_limit [KNL,X86-64,XEN]
5373 Restricts the kernel running paravirtualized under Xen
5374 to use only up to 512 GB of RAM. The reason to do so is
5375 crash analysis tools and Xen tools for doing domain
5376 save/restore/migration must be enabled to handle larger
5379 xen_emul_unplug= [HW,X86,XEN]
5380 Unplug Xen emulated devices
5381 Format: [unplug0,][unplug1]
5382 ide-disks -- unplug primary master IDE devices
5383 aux-ide-disks -- unplug non-primary-master IDE devices
5384 nics -- unplug network devices
5385 all -- unplug all emulated devices (NICs and IDE disks)
5386 unnecessary -- unplugging emulated devices is
5387 unnecessary even if the host did not respond to
5389 never -- do not unplug even if version check succeeds
5391 xen_legacy_crash [X86,XEN]
5392 Crash from Xen panic notifier, without executing late
5393 panic() code such as dumping handler.
5395 xen_nopvspin [X86,XEN]
5396 Disables the ticketlock slowpath using Xen PV
5400 Disables the PV optimizations forcing the HVM guest to
5401 run as generic HVM guest with no PV drivers.
5402 This option is obsoleted by the "nopv" option, which
5403 has equivalent effect for XEN platform.
5405 xen_scrub_pages= [XEN]
5406 Boolean option to control scrubbing pages before giving them back
5407 to Xen, for use by other domains. Can be also changed at runtime
5408 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages.
5409 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT.
5411 xen_timer_slop= [X86-64,XEN]
5412 Set the timer slop (in nanoseconds) for the virtual Xen
5413 timers (default is 100000). This adjusts the minimum
5414 delta of virtualized Xen timers, where lower values
5415 improve timer resolution at the expense of processing
5416 more timer interrupts.
5418 nopv= [X86,XEN,KVM,HYPER_V,VMWARE]
5419 Disables the PV optimizations forcing the guest to run
5420 as generic guest with no PV drivers. Currently support
5421 XEN HVM, KVM, HYPER_V and VMWARE guest.
5423 xirc2ps_cs= [NET,PCMCIA]
5425 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]]
5428 By default on POWER9 and above, the kernel will
5429 natively use the XIVE interrupt controller. This option
5430 allows the fallback firmware mode to be used:
5432 off Fallback to firmware control of XIVE interrupt
5433 controller on both pseries and powernv
5434 platforms. Only useful on POWER9 and above.
5436 xhci-hcd.quirks [USB,KNL]
5437 A hex value specifying bitmask with supplemental xhci
5438 host controller quirks. Meaning of each bit can be
5439 consulted in header drivers/usb/host/xhci.h.
5442 Format: { early | on | rw | ro | off }
5443 Controls if xmon debugger is enabled. Default is off.
5444 Passing only "xmon" is equivalent to "xmon=early".
5445 early Call xmon as early as possible on boot; xmon
5446 debugger is called from setup_arch().
5447 on xmon debugger hooks will be installed so xmon
5448 is only called on a kernel crash. Default mode,
5449 i.e. either "ro" or "rw" mode, is controlled
5450 with CONFIG_XMON_DEFAULT_RO_MODE.
5451 rw xmon debugger hooks will be installed so xmon
5452 is called only on a kernel crash, mode is write,
5453 meaning SPR registers, memory and, other data
5454 can be written using xmon commands.
5455 ro same as "rw" option above but SPR registers,
5456 memory, and other data can't be written using
5458 off xmon is disabled.